Norwegian Petroleum Society (NPF), Special Publication No.

Structural and
Tectonic Modelling
and its Application to
Petroleum Geology
Proceedings of Norwegian Petroleum Society Workshop,
18-20 October 1989f Stavanger, Norway

Edited by

R.M. Larsen
Statoil, P.O. Box300, 4001 Stavanger, Norway

H.Brekke
Norwegian Petroleum Directorate, P.O. Box600, 4001 Stavanger, Norway

B.T. Larsen
Statoil, P.O. Box300, 4001 Stavanger, Norway

and

E.Talleraas
AISNorskeShell, P.O. Box40, 4056Tananger, Norway

ELSEVIER
Amsterdam-London- New York-Tokyo 1992
ELSEVIER SCIENCE PUBLISHERS B.V.
Sara Burgerhartstraat25
P.O. Box 211, 100CAE Amsterdam, The Netherlands

ISBN 0-444-886079

1992 Elsevier Science Publishers B.V. All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means,
electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher,
Elsevier Science Publishers B.V, Copyright and Permissions Department, P.O. Box 521, 1000 AM Amsterdam,
The Netherlands.

Special regulations for readers in the U.S.A. -This publication has been registered with the Copyright Clearance Center
Inc. (CCC); Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies
of parts of this publication maybe made in the U.S.A. All other copyright questions, including photocopying outside of
the U.S.A., should be referred to the publisher.

No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products
liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained
in the material herein.

This book is printed on acid-free paper.

Printed In The Netherlands

 V

Preface

A conference entitled "Structural and Tectonic Modelling and Its Application to Petroleum Geol-
ogy" was arranged by the Norwegian Petroleum Society in Stavanger, Norway, and took place on the
18th to 20th of October, 1989. At the conference, more than 80 papers and posters were presented.
Thirty-six of these papers are contained in this volume.
Structural geology and tectonic models have implications on all scales and at several phases during
the exploration for and exploitation of hydrocarbons. Large-scale models define the basic concepts of
regional deformation, which influences subsidence history and, hence, basin modelling and maturation
studies. Semi-regional and local structural studies are necessary to understand fault geometry, timing,
migration paths and hydrocarbon traps. Fractures, from composite faults to simple joints, can act as
seals and enhance reservoir communication.
A complete structural analysis includes interpretation of seismic data, geodynamic modelling, geo-
metric analysis and studies of fracture systems down to micro-scale. Analyses should construct realistic
three-dimensional models in order to unravel structural development in both time and space. The aim
of this international conference was to present examples of structural and tectonic models at all scales.
Emphasis was put on models for prediction and their applications and limitations in petroleum geol-
ogy
The conference was organized in three parts: each topic having a keynote speaker. The three parts
were Regional Tectonics and Basin History, The Application of Tectonic and Structural Aanafysis to Hy-
drocarbon Accumulation and Structural Geology on Reservoir and Field Scale. In addition, two special
sessions were arranged; "Tertiary Uplift/Barents Sea" and "Fault Processes and Geometries".
The meeting was initiated by the Geology and Geophysics Committee of the Norwegian Petroleum
Society. The organization committee consisted of A. Andresen, H. Brekke, R.H. Gabrielsen, S. Hans-
lien, B.T. Larsen, R.M. Larsen, E. Tklleraas, E. Holter and K. Haugnaess from NPF
The Norwegian Petroleum Society has decided to introduce a special publication series of which
this is the first. This volume, therefore, is the Norwegian Petroleum Society, Special Publication No.
1. It is recommended that this volume is referred to in one of the following forms, as appropriate:
Larsen, R.M., Brekke, H., Larsen, B.T. and Tklleraas, E. (Editors), 1992. Structural and Tectonic
Modelling and its Application in Petroleum Geology. Norwegian Petroleum Society, Special Publi-
cation No. 1. Amsterdam, Elsevier.
Jensen, L.N. and S0rensen, K 1992. Tectonic framework and kalokinesis of the Nordkapp Basin,
western Barents Sa. In: R.M. Larsen, H. Brekke, B.T. Larsen and E. Tklleraas (Editors), Structural
and Tectonic Modelling and its Application in Petroleum Geology. Norwegian Petroleum Society,
Special Publication No. 1. Amsterdam, Elsevier, pp. 109-120.
The symposium in Stavanger was the ninth in a generally biennial series of meetings organized by
the Norwegian Petroleum Society (Norsk Petroleumsforening, NPF) comprising:

Jurassic Northern North Sea Symposium, Stavanger (1975)

o Mesozoic Northern North Sea Symposium, Oslo (1977)
Norwegian Sea Symposium, TYoms0 (1979)
The Sedimentation of the North Sea reservoir rocks, Geilo (1980)
Norwegian Symposium on Exploration, Bergen (1981)
North European Margin Symposium, TVondheim (1983)
Habitat of Hydrocarbons Symposium, Stavanger (1985)
o Correlation in Hydrocarbon Exploration, Bergen (1988)
VI Preface

Information on the first five symposia and their publication can be obtained from: NPF, P.O. Box
1897 - Vika, 0124 Oslo, Norway.
The proceedings of the next three symposia, plus those of the "Petroleum Geochemistry in the Ex-
ploration of the Norwegian Continental Shelf" and an "atlas" of hydrocarbon discoveries, containing
a full description of 35 of Norway's major oil and gasfieldsand finds, "Geology of the Norwegian Oil
and Gas Fields", have been published by and can be obtained from Graham and TVotman Ltd., Stirling
House, 66 Wilton Road, London SW1V IDE, England.
The committee would like to acknowledge the administrative support of T.H. Hult, Statoil.

The Editorial Committe

September 1991

ACKNOWLEDGEMENTS
The editing committee wish to thank the following reviewers for their thoughtful comments and
suggestions.

Andresen, A. Koestler, A.G. Pallesen, S.

Beach, A. Larsen, B.T. Roberts, A M .
Dangerfield, J. Larsen, R.M. Sales, J.K.
Dengo, C.A. Lippard, S. Skarpnes, O.
Ehrenberg, S. Michelsen, O. Skogseid, J.
Eldholm, O. Milnes, A.G. S0rensen, S.
Faleide, J.I. Myhre, A M . Thon, A
Forth, Ph. Nash, D.F. T0rudbakken, B.
Fasrseth, R.B. Needham, C. Vejbaek, O.V
Gabrielsen, R.H. Nelson, E.B. Welbon, A I .
Gowers, M.B. N0ttvedt, A. Wiggins, B.D.
Graham, R.H. Olaussen, S. Yielding, G.
Jensen, L.N. Olsen, T Young, R.
Knipe, R. Oprinsen, F.
VI Preface

Information on the first five symposia and their publication can be obtained from: NPF, P.O. Box
1897 - Vika, 0124 Oslo, Norway.
The proceedings of the next three symposia, plus those of the "Petroleum Geochemistry in the Ex-
ploration of the Norwegian Continental Shelf" and an "atlas" of hydrocarbon discoveries, containing
a full description of 35 of Norway's major oil and gasfieldsand finds, "Geology of the Norwegian Oil
and Gas Fields", have been published by and can be obtained from Graham and TVotman Ltd., Stirling
House, 66 Wilton Road, London SW1V IDE, England.
The committee would like to acknowledge the administrative support of T.H. Hult, Statoil.

The Editorial Committe

September 1991

ACKNOWLEDGEMENTS
The editing committee wish to thank the following reviewers for their thoughtful comments and
suggestions.

Andresen, A. Koestler, A.G. Pallesen, S.

Beach, A. Larsen, B.T. Roberts, A M .
Dangerfield, J. Larsen, R.M. Sales, J.K.
Dengo, C.A. Lippard, S. Skarpnes, O.
Ehrenberg, S. Michelsen, O. Skogseid, J.
Eldholm, O. Milnes, A.G. S0rensen, S.
Faleide, J.I. Myhre, A M . Thon, A
Forth, Ph. Nash, D.F. T0rudbakken, B.
Fasrseth, R.B. Needham, C. Vejbaek, O.V
Gabrielsen, R.H. Nelson, E.B. Welbon, A I .
Gowers, M.B. N0ttvedt, A. Wiggins, B.D.
Graham, R.H. Olaussen, S. Yielding, G.
Jensen, L.N. Olsen, T Young, R.
Knipe, R. Oprinsen, F.
 VII

List of Contributors

A. ANDRESEN University of Oslo, Department of Geology, P.O. Box 1047, Blindem, 0316 Oslo 3,
Norway

N. BARTON Norges GeotekniskInstitutt, P.O. Box 40 Tsen, 0801 Oslo 1, Norway

A. BEACH Alastair Beach Associates, 11 Royal Exchange Square, Glasgow Gl 3AJ, Scotland,
U.K.

D. BERING Geology Institute, University of Bergen, Bergen, Norway

Present address: Norwegian Petroleum Directorate, P.O. Box 600, N-4001 Stavanger

K. BJ0RLYKKE Department of Geology, University of Oslo, P.O. Box 1047 Blindem, N-0316 Oslo 3,
Norway

A. BROWN BPNorway Limited UA., P.O. Box 197, N4033Foms, Norway

R.W.H. BUTLER Department of Earth Sciences, The University, Leeds, LS2 9JT, U.K.

B.P.R. CALINE Koninklijke/Shell Exploratie en Produktie Laboratorium, Volmerlaan 6, 2288 GD

Rijswijk ZH, The Netherlands

J. DANGERFIELD Phillips Petroleum Company Norway, UtenlandskAksjeselskap, P.O. Box 220, N-4056
Tan anger, Norway

R.L. DART Branch of Geologic Risk Assessment, U.S. Geological Survey, Box 25046, MS 966,
Denver Federal Center, Denver, CO 80225, U.S.A.

M. DeBUYL Western Geophysical, 10001 Richmond Avenue, Houston, TX 77252, USA.

CA. DENGO Exxon Exploration Company, P.O. Box 4279, Houston, TX 77210-4279, USA.

J.B. DUNHAM Unocal International Division, P.O. Box 7600, Los Angeles, CA 90051, U.S.A.

J.E. ILIFFE Department of Geological Sciences, University of South Carolina, Columbia, SC

29208, USA.

H. FARRELL Phillips Petroleum Company Norway, UtenlandskAksjeselskap, P.O. Box 220, N-4056
Tananger, Norway

P.S. FEATHERSTONE Koninklijke/Shell Exploratie en Produktie Laboratorium, Vhnerlaan 6, Rijswijk, The

Netherlands

W. FJELDSKAAR Rogaland Research Institute, P.O. Box 2503, Ullandhaug N-4004 Stavanger, Norway

R.H. GABRIELSEN Geological Institute, Department A, University of Bergen, Allegaten 41, 5007 Bergen,
Norway
Present address: Norsk Hydro A/S, Forskningssenteret, N-5020 Bergen, Norway

K. GERDES BP Exploration, Britannic House, Moore Lane, London EC2Y 9BU, UK.
Present address: BP Norway Limited U.A., P.O. Box 197, N-4033 Foms, Norway

P. HAREMO University of Oslo, Department of Geology, P.O. Box 1047, Blindem, 0316 Oslo 3,
Norway
Present address: Norsk Hydro AJS, Drammensveiem 264, P.O. Box 200, N-1321
Stabekk, Norway
VIII List of Contributors

L. HERMANS Koninklijke IShell Exploratie en Produktie Laboratorium, Volmerlaan 6, 2288 GD

Rijswijk, The Netherlands

R.G. HICKMAN Unocal Science &. Technology Division, 5460 East La Palma Avenue, P.O. Box 68076,
Anaheim, CA 92817-8076, U.S.A.

J. HURST BP Exploration, London, U.K.

R. JEANS BP Exploration, London, U.K.

L.N. JENSEN StatoilA/S, P.O. Box 300, N-4001 Stavanger, Norway
A. KJEMPERUD READ Production Geology Services A/S, P.O. Box 145, N-1300 Sandvika, Norway
Present address: PETEC, P.O. Box 63, N-3478Ncernes, Norway
O.S. KL0VJAN Norsk Hydro A/S, Harstad, P.O. Box 31, 9401 Harstad, Norway

I. KNIGHT StatoilA/S, P.O. Box 300, 4001 Stavanger, Norway

Present address: Phillips Petroleum Company Norway, Utenlandsk Aksjeselskap, P.O.
Box 220, N-4056 Tananger, Norway

R.J. KNIPE Department of Earth Sciences, The University of Leeds, Leeds, LS2 9JT, UK.

A.G. KOESTLER Geo-ReconA/S, Bernhard Herres vei 3, 0376 Oslo 3, Norway

J.A. KORSTGRD Geologisk Institut, Aarhus Universitet, DK-8000rhus C, Denmark

VB. LARSEN StatoilA/S, Stavanger, Norway
S.E. LAUBACH Bureau of Economic Geology, University of Texas at Austin, University Station Box X,
Austin, TX 78713-7508, U.S.A.
EK. LEHNER Koninklijke/Shell Exploratie en Produktie Laboratonum, Voltnerlaan 6, Rijswijk, The
Netherlands
I. LERCHE Department of Geological Sciences University of South Carolina, Columbia, SC 29208,
U.S.A.
S. LIPPARD Continental Shelf and Petroleum Technology Research (IKU), N-7034 Trondheim,
Norway
Present address: Department of Geology and Mineral Resources Engineering Norwe-
gian Institute of Technology (NTH), N-7034 Trondheim, Norway
G.LIU Continental Shelf and Petroleum Technology Research (IKU), N-7034 Trondheim,
Norway
S.E. LI VERA Koninklijke IShell Exploratie en Produktie Laboratorium, Volmerlaan 6, 2288 GD
Rijswijk ZH, The Netherlands
A. MAKURAT Norges GeotekniskInstitutt, P.O. Box 40 Tsen, 0801 Oslo 1, Norway
O. MICHELSEN Geological Survey of Denmark, Copenhagen, Denmark
Present address: Department of Geology, University of Aarhus, DK-8000 rhus C,
Denmark
A.G. MILNES Department of Earth Sciences, ETH, Zrich, Switzerland
Present address: Geological Institute, Department A, University of Bergen, Allegt 41,
N-5007 Bergen, Norway
A.W. MITCHELL BP Norway Limited UA., P.O. Box 197, N-4033 Fonts, Norway
T.E. MOGENSEN Aarhus University, rhus, Denmark
K. MONSEN Norges Geoteknisk Institutt, P.O. Box 40 Lsen, 0801 Oslo 1, Norway
H. MORIZOT Elf Aquitaine Norge A/S, Stavanger, Norway
K. NAKAYAMA Department of Geological Sciences University of South Carolina, Columbia, SC 29208,
U.S.A.
LR. NILSSEN BP Norway Limited UA., P.O. Box 197, N-4033 Torus, Norway
List of Contributors IX

B. NYLAND Norwegian Petroleum Directorate, P.O. Box 600, N-4001 Stavanger, Norway

N.E. ODLING Continental Shelf and Petroleum Technology Research Institute (IKU), NHkon Mag-
nusson'sgt. IB, P.O. Box 1883, larlesletta, 7001 Trondheim
Present address: Bergen Environmental Centre, IBM, Thorm0hlensgt. 55, Bergen,
N-5008, Norway
T. PEDERSEN Department of Geology, University of Oslo, Oslo, Norway
A. RASMUSSEN Norsk Hydro A IS, Harstad, P.O. Box 31, 9401 Harstad, Norway
R.P. RATTEY BP Exploration, Glasgow, U.K.
S J. REYNOLDS Arizona Geological Survey, Tucson, AZ 85719, USA.
F. RIIS Norwegian Petroleum Directorate, P.O. Box 600, N-4001 Stavanger, Norway
A. ROBERTS Badley Ashton and Associates Limited, Winceby House, Winceby, Home as tie, Lin-
colnshire, LN9 6PB, UK.
K G . R0SSLAND EssoNorgeA/S, P.O. Box 60, N-4033Torus, Norway
Present address: StatoilA/S, UND-LS, ST-THDl, P.O. Box 300, N-4001 Stavanger,
Norway
M.G. ROWAN Alastair Beach Associates, 11 Royal Exchange Square, Glasgow Gl 3AI, Scotland,
UK
W SASSI Koninklijke IShell Exploratie en Produktie Laboratorium, Volmerlaan 6, 2288 GD
Rijswijk ZH, The Netherlands
Present address: Institut Geologie-Geochimie, Institut Frangais du Petrole, 14 Avenue
de Bois Preau, B.P 311, 92506 Rueil-Malmaison Cedex, France
WJ. SCHMIDT Conoco Norway Inc., Norway
Present address: Conoco, 600 N Dairy Ash ford, Houston, TX 77079, U.S.A.
J. SKAGEN StatoilA/S, P.O. Box 40, N-9401 Harstad
O. SKARPNES StatoilA/S, P.O. Box 40, N-9401 Harstad
Present address: StatoilA/S, U&PINT, DT-FHB3, P.O. Box 300, N-4001 Stavanger,
Norway
J. SKOGSEID Department of Geology, University of Oslo, P.O. Box 1047 Bndern, 0316 Oslo 3,
Norway
S. SKOTTHEIM Elf Aquitaine Norge A/S, Stavanger, Norway
KS0RENSEN StatoilA/S, St. Annce Plads 13, DK-1298K0benhavnK, Denmark
S. S0RENSEN Elf Aquitaine Norge AIS, Stavanger, Norway
Present address: Norex A IS, Luragaarden, Standnes N-4300, Norway
I J . STEWART BP Norway Limited UA., P.O. Box 197, N4033 Forus, Norway
Present address: BP Exploration Inc., P.O. Box 4587, Houston, TX 77210, USA.
T ST0LAN Norsk Hydro A IS, P.O. Box 31, N-9401 Harstad, Norway
A. STORLI StatoilA/S, Bergen, Norway
P.T. SVELA BP Norway Limited UA., P.O. Box 197, N-4033 Forus, Norway
E. SVERDRUP Saga Petroleum A/S, Kj0rbovn. 16, P.O. Box 490, N-1301 Sandvika, Norway
H.S. SWOLFS U.S. Geological Survey, Box 25046, MS 966, Denver Federal Center, Denver, CO 80225,
USA.
Geologisk Institut, Aarhus Universitet, DK-8000 Arhus C, Denmark, and Department
R.O. THOMSEN
of Geological Sciences, University of South Carolina, Columbia, SC 29208, USA.
Present address: Saga Petroleum A/S, P.O. Box 490, N-1301 Sandvika, Norway
Koninklijke /Shell Exploratie en Produktie Laboratonum, Vobnerlaan 6, Rijswijk, TJxe
A.D. van KUYK Netherlands
X List of Contributors

I.R. VANN BP Exploration, London, U.K.

O.V. VEJB^EK Geological Survey of Denmark, Thoravej 8, DK-2400 K0benhavn, Denmark

Present address: Amoco Production Company, Houston, TX 77253-3092, U.S.A.

B.C. VENDEVILLE Bureau of Economic Geology, University of Texas at Austin, Austin, TX 78713-7508,
U.S.A.

T. VORREN University of Troms0, P.O. Box 3085, N-9001 Troms0

A.I. WELBON Department of Geology, University of Oslo, P.O. Box 1047 Bndern, 0316 Oslo 3,
Norway
Present address: Department of Geological Sciences, Queens University, Union Street,
Kingston, Ont. K7L 3N6, Canada

G. YIELDING Badley Ashton and Associates Limited, Winceby House, Winceby, Florncastle, Lin-
colnshire, LN9 6PB, U.K.

R. YOUNG BP Exploration, London, UK.

Present address: BP Norway Limited U.A., Forusbeen 35, P.O. Box 197, N-4033 Forus,
Norway
 Geodynamic modelling of the Danish Central Trough
O.V. Vejbaek

A geodynamic model is applied to the post-Middle Jurassic tectonic development of the northern and middle Danish Central
Trough. The model uses stretching and phase transformations as subsidence mechanisms. It uses a definition of the lower
boundary below the lithosphere in terms of heat flow that contrasts with other models.
Following the Middle Jurassic uplift and erosion of major parts of the Central Trough, an important stretching phase caused
rapid subsidence, centred in the Tail End Graben area. The subsidence was controlled by the boundary fault separating the
Ringk0bing Fyn High and the Tail End Graben-Salt Dome Province. Local initial stretching factors are calculated to be up to
2. Due to an asymmetric model, the relatively large stretching factors are interpreted as having affected mainly the crust locally,
leaving the lithospheric mantle almost unaffected by stretching, as indicated by the assumed presence of a detachment at the
deep crustal continuation of the master fault, and by backstripping. As the tectonic activity continued, faulting spread to the
west to the Feda Graben, the Gertrud Graben and the Heno Plateau.
The modelling shows that heat flow reached its peak in the deepest parts of the Tail End Graben area during the main
Middle to Late Jurassic rifting phase. This was later reversed leading to a concentration of relatively high heat flow in relatively
high areas. This is explained by a more shallow location of anomalous heat in the lithosphere and, therefore, a shorter decay
constant in the Graben areas, than in the relatively high areas to the west.
It is shown that stretching, thermal subsidence and the metamorphic mechanism incorporated in the model cannot account
for the Cenozoic portion of the subsidence pattern. It is suggested that a gabbro-eclogite phase transition may explain the
Cenozoic subsidence, with only negligible effect on surface heat flow.

Introduction Watts (1989) has also challenged the stretching hy-

pothesis. They found that additional mechanisms
The purpose of this paper is to present a quan- must be active such as gabbro-eclogite phase tran-
titative analysis of the Middle Jurassic to Cenozoic sition, deep crustal dewatering and anomalous as-
subsidence pattern of the Danish Central Trough. thenospheric convection. This is especially true in
This is carried out on the basis of the geodynamic the central parts of the middle and southern North
model proposed by Vejbaek (1989). Sea, where a linear to accelerating tectonic subsi-
The purpose of modelling the evolution of a par- dence during the Cretaceous, Tertiary and Quater-
ticular basin is to increase the understanding of the nary is observed. Thermal subsidence alone can thus
interplay between the tectonism, as interpreted from be rejected for that period.
seismic data, and the responsible lithospheric pro- The concept that metamorphism may play an im-
cesses. It is also to constrain heat flow variation portant role in the formation of sedimentary basins
through time used for modelling maturation of or- has also been proposed in the form of a one-
ganic matter. Geodynamic modelling has its main dimensional model (Vejbaek, 1989). This model dif-
origin in the work of McKenzie (1978), which was fers from the original concept of McKenzie (1978) by
later applied to the North Sea area (e.g., Sclater including heat production in the lithosphere, phase
and Christie, 1980; Dewey, 1982; Barton and Wood, transformations and, most importantly, by having a
1984). The basic concept in this hypothesis is that heat flow condition below the lithosphere. The lower
sedimentary basins are formed by stretching alone. boundary of the fixed frame used for the calculations
In the North Sea, this has been questioned, because is at a fixed depth equal to or below the base of
industrial reflection seismic data show too little fault- the lithosphere and differs from the frequently used
ing to explain crustal thinning by stretching (e.g., temperature definition. The heat flow condition ac-
Ziegler, 1982, p. 90). A recent regional North Sea counts for a longer duration of anomalous heat flow
quantitative subsidence analysis study by Thorne and as required by data (Vejbaek, 1989, 1990). The model

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 1-17. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
2 O.VVejbcek

handles thermal perturbations induced by one or sev- Triassic. During the Cretaceous all of the structural
eral phases of stretching, variations in asthenospheric highs became submerged and the individual sub-
heat flow or sill intrusions. basins coalesced into a regionally subsiding basin.

General geology of the Danish Central Subsidence history

Trough The Danish Central Trough probably started to
subside in the Early Permian (e.g., Ziegler, 1975;
The main structural units of the Danish Central Michelsen, 1982; Gowers and Saeb0e, 1985). The ex-
Trough, as defined by Andersen et al. (1982) and tent of Permian deposits below the central part of the
Gowers and Saeb0e (1985) and revised by M0ller Danish Central TYough is limited. Only the northeast-
(1986) and Vejbask (1986), are summarized in Fig. 1. ern part of the Danish Central TYough in the south-
In the western part of the region, the Late Jurassic ern S0gne Basin and the Salt Dome Basin, show Zech-
Grensen Nose and the Outer Rough Basin separate stein salt structures. The only connection between the
th~ Tnge High and the Mads High from the Mid northern and southern Zechstein basins in the Central
North Sea High, located in the adjacent British sec- Trough area seems to have been through the Tkil End
tor (Gowers and Saeb0e, 1985). The Heno Plateau Graben and the S0gne Basin. However, the deeper
represents an intermediate block in which relatively parts of the Tkil End Graben, which is mainly filled
thin Upper Jurassic deposits are unconformably with Mesozoic sediments, cannot be resolved seismi-
overlain by Lower and Upper Cretaceous sediments. cally. Hence, it remains uncertain whether or not there
The Feda Graben contains up to 2.5 km of Upper has actually been any connection between the north-
Jurassic shales (M0ller, 1986). It is characterized ern and southern Zechstein basins through the Cen-
by continuous sedimentation from Late Jurassic to tral Trough. If a connection existed between the two
Late Cretaceous times. This trough is separated from basins, it probably was of a limited extent, since no
the Gertrud Graben by the narrow Gert Ridge and salt structures can be detected in the central part of
it extends northwards into Norwegian waters. The the region.
Gertrud Graben probably has a subsidence history Triassic sedimentation in the Central Trough is in-
similar to that of the Feda Graben, but it contains dicated by more than 1400 m of Triassic sediments
considerably thinner Upper Jurassic deposits. in the U-l well (Jacobsen, 1982). This is a minimum
The Gertrud Graben and the Heno Plateau are thickness, since the well did not penetrate the en-
separated from the TM1 End Graben by the Arne- tire Triassic succession, and the top of the section is
Elin Graben. The latter has a flower-type structural mildly truncated by the Fjerritslev Formation. Where
geometry and is suggested to have been formed by the overlying Lower Jurassic Fjerritslev Formation is
Late Jurassic dip-slip faulting with significant NNW- present, the unconformity is weak and the youngest
SSE left-lateral strike-slip components (M0ller, 1986; Triassic belongs to the Lower Rhaetian. Where the
Vejbaek, 1986). The Tail End Graben came into Fjerritslev Formation has been removed during the
existence during the Permian(?) to Triassic and rapid Middle Jurassic uplift, the Triassic may be more
subsidence during the Late Jurassic resulted in the severely truncated or even completely removed, as is
accumulation of more than 4 km of shale-dominated seen in the western part of the Central Trough area.
sediments. How much of the Central Trough was actively sub-
It has been suggested that strike-slip movements siding as a graben feature is uncertain due to these
along a conjugate set of WNW-ESE right-lateral and later erosion events. Data from the V-l well sug-
NNW-SSE left-lateral fault systems also influenced gest that the Ringk0bing Fyn High was covered by
the rapid Late Jurassic subsidence of the pull-apart sediments before the Middle Jurassic uplift (M0ller,
basins of the Feda and Tkil End Grabens (Gowers 1986). Similarly, the absence of Triassic sediments
and Saeb0e, 1985). However, faulting was probably in wells such as P-l may indicate Middle Jurassic
dominated by dip-slip. The Mandal High and the erosion rather than a primary distribution of Trias-
associated S0gne Basin represent an intermediate sic sediments. The present distribution of THassic
fault block between the Ringk0bing Fyn High and sediments is not well known due to limited well con-
the Tail End Graben. The eastern margin of the Tkil trol and uncertain seismic correlation across major
End Graben is formed by a major basement block of structural features. Triassic sediments are likely to
the stable Ringk0bing Fyn High. be present in the entire Salt Dome Basin, Tail End
The Tail End Graben is bounded to the south by Graben and S0gne Basin. To the west, the THassic is
the Salt Dome Province. Here the structural frame- thin or absent due to erosion, and Middle and Upper
work becomes strongly overprinted by successive dia- Jurassic sediments rest with angular unconformity on
pirism of the Zechstein salt, initiated during the Late Palaeozoic strata. Possible Triassic sediments in the
Geodynamic modelling of the Danish Central Trough 3

Fig. 1. Structural outline of the Danish Central Trough with location of the wells and profiles used in the modelling (based on M0ller,
1987). A-A' and B-B' are profiles shown in Figs. 8 and 10, respectively.

Feda Graben are probably truncated to the south. characterized by tectonic stability, with rather uni-
A minor unconformity separates the Triassic from form and possibly widespread deposition of shales
the Lower Jurassic in the Salt Dome Province (An- in a much-expanded basin. This situation was re-
dersen et al., 1982). However, this minor event prob- versed in the Middle Jurassic, when renewed tectonic
ably represents an episode in a decreasingly active activity caused updoming and erosion of the cen-
tectonic environment. This may also be indicated by tral part of the North Sea. This was accompanied
reconstructions of tectonic subsidence elsewhere in by marked narrowing of depositional areas to fault-
the North Sea (Sclater and Christie, 1980; Wood bound basins and deposition of mainly coarse elastics
and Barton, 1983). Hence, the Early Jurassic was (Eynon, 1981). As noted above, this tectonism re-
4 O.VVejbcek

suited in extensive erosion to the west, whereas very

AGE SYSTEM/
DANISH CENTRAL TROUGH
little material was removed in the vicinity of the
CEN-6UNIT
Ringk0bing Fyn High. During this period the gen- PLIO-I
eral post-THassic structural framework of the Central CENE
CEN-5UNIT
Ttough evolved. Since the magnitude of the erosion
is hard to assess with sufficient precision for quantita-
tive modelling, the ensuing modelling attempt begins CEN-4UNIT
with the post-Middle Jurassic.
During the Late Jurassic, the rate of tectonic
subsidence increased locally as wrenching gave rise
to considerable lateral variations in subsidence rates.
CEN-3UNIT
This initiated the major Upper Jurassic depocentres
such as the Feda and Tkil End Grabens (Gowers and
Saeb0e, 1985; M0ller, 1986). During the Middle to
Late Jurassic and Early Cretaceous, the depositional
area gradually expanded towards the west.
The structural framework of the Lower Cretaceous
is very similar to the Upper Jurassic, with exten- jjprjTj^^
Maastrichtian
sional block faulting continuing under the influence
of wrenching. The importance of the WNW-ESE
orientated right-lateral wrench system suggested by
Gowers and Sa^b0e (1985) gradually ceased, but the
conjugated left-lateral NNW-SSE system continued
to be active during the Early Cretaceous (Vejbaek,
1986). The rate of subsidence decreased markedly
and the Lower Cretaceous deposits became sepa-
rated more into local depocentres, partly as a result
of decreased sediment supply and partly due to the
accentuation of ridges and structural highs separat-
ing the subbasins. At this time, the Tail End Graben
was separated from the S0gne Basin due to uplift
of the Mandal High Complex. Also, the connection
between the Tkil End Graben and the Salt Dome
Province became restricted due to the initiation of
the low local east-west trending Pollerne Ridge.
Block faulting gradually ceased from Hauterivian
and onwards and gave way to regional subsidence,
interrupted by periods of inversion (Vejbaek and
Andersen, 1987). Inversion produced strong reverse
faulting. Throws exceed 600 m locally in conjunction
with mild dextral strike-slip faulting along reactivated
NNW-SSE striking fault systems. During the latest
Cretaceous times the inversion was dominated by
flexing rather than faulting.
Inversion ceased in the Tertiary and regional sub-
sidence turned the North Sea Basin into the large
north-west European sedimentary basin seen today.
Deposition rates changed gradually from moderate
to high through the Tertiary and Quaternary with
shifting depocentres, but with very little variation in Fig. 2. General stratigraphy of the post Triassic sediments of
the Danish Central Trough. Absolute ages are from Haq et al.
total thickness. The total thickness of the Cenozoic
(1987). The figure is modified from Michelsen et al. (1981) and
within the Danish Central Ttough varies from about Jensen et al. (1986).
2 km in the southern part to about 3 km in the north-
ern part (Day et al., 1981; Vejbaek and Andersen,
1987).
Geodynamic modelling of the Danish Central Trough 5

Rock Unit
Cen Unit 6 1.7 My /.ll.m/My] C r u s t a l spreading axis

Cen Unit 5 8.3My Yield stress

Cen U n i t U 17. My
Cen U n i t 3 22. My
Cen U n i t 1& 2 12. My
Chalk Unit 5 & 6 10-Myl
C h a l k Unit 3 &U 13.My
Chalk Unit 1& 2 12.My
K0dby Formation .My
Sola Formation 15. My
Lower lithosphenc spreading axis
Tuxen Formation . My|
Valhall Formation 11. My app. 100Km.
Farsund Formation 5.My 600m/My|
Lola Formation 7.My Fig. 4. Tentative model to unify the modelling results (modified
f>
O.m/My 50.m/My 100.m/My 150m/My from Braun and Beaumont, 1989). A detachment level develops
Fig. 3. Estimated maximum rates of deposition for major rock as temperature increases during stretching (dotted line in yield
units. Thicknesses are not decompacted. The durations noted stress diagram). This may cause stretching to be transferred to a
are based on the ages cited in Fig. 2. pre-existing weakness zone, represented here by the main fault
of the Tail End Graben. Lithospheric stretching factors will be
low at that location, as opposed to the spreading axis area.
The average rates of deposition for the major
stratigraphic units (Fig. 2) are shown in Fig. 3. Pe-
riods of relatively vigorous tectonic activity are indi- two layers are stretched differentially (cf. Coward,
cated by the high depositional rates, as water depths 1986). Although the regionally averaged stretching
are assumed to have remained below a few hun- factor should be the same for both layers, the axis
dred metres. High deposition rates are recorded for of maximum stretching offsets considerably from one
the Lola and Farsund Formations and the youngest layer to another. The two zones of maximum ex-
part of the Cenozoic. The idea of McKenzie (1978), tension are accommodated by a detachment zone in
that short episodic events of rapid subsidence are the deep crust (Fig. 4). The two main steps in the
caused by rifting and account for the bulk of the exercise, backstripping and the modelling itself, are
sediments, and that they are separated by longer briefly summarized below. Detailed discussions may
periods of thermal subsidence, seems to be justified be found in Ykler et al. (1978) and Vejbaek (1989,
(Fig. 3). However, as will be discussed later, the Late 1990) with regard to backstripping and subsidence
Cenozoic subsidence is not a rifting period, although modelling, respectively.
subsidence rates compare with those recorded by
deposition of the syn-rift Late Jurassic Farsund For- Backstripping
mation. Backstripping was performed using the procedure
The base of the modelled succession is taken as described by Ykler et al. (1978) which, among
the Middle Jurassic, because the amount of preced- other things, accounts for the effects of overpres-
ing erosion is considered ambiguous, thus rendering suring on the burial history. It is a finite difference
quantitative reconstruction of tectonic subsidence method based on the principles of conservation of
unreliable. mass, momentum and energy. Compaction is con-
sidered to be governed by fluid flow influenced by
Geodynamic modelling thermo-mechanical changes. The reason for using
this method, instead of the more simple backstrip-
Modelling is divided into a study of the subsidence ping procedure based on exponential decrease of
history and an assessment of what may have caused porosity with depth, is that substantial overpressuring
it. The subsidence is suggested to be controlled by is seen in the Central Trough and the Viking Graben
three major factors: crustal thinning, phase trans- (e.g., Chiarelli and Duffaud, 1980; Thorne and Watts,
formations and the heat budget of the lithosphere. 1989). It causes significant breakdown of the sim-
Modifications of subsidence patterns by variations ple empirical porosity-depth relationship, that would
in fiexural rigidity are disregarded. Quantification cause serious errors if it were not considered. Ac-
is done mainly by fitting the one-dimensional back- cording to the Ykler (1978) model, overpressuring
stripped tectonic burial graphs to tectonic subsidence occurs mainly as a consequence and during periods
curves. The stretching factors are also constrained of high deposition rates. For instance, if high de-
by evidence seen in industrial seismic data. In con- position rates during stretching were not considered
trast to the original McKenzie (1978) model, in they might lead to overpressuring and thus underes-
which the crust and immediately subjacent litho- timated initial subsidence.
spheric mantle are stretched by the same factor, The backstripping is followed by tectonic back-
an asymmetric model is used here, in which the stripping. Tectonic backstripping is referred to here
6 O.VVejbcek

as a separate procedure whereby the effects of sed- the low velocity layer of the asthenospheric man-
iment loading are subtracted to yield the tectonic tle below Phanerozoic tectonic provinces (Calcagnile
subsidence under the assumption of isostatic equi- and Scarpa, 1985). In the present study, the astheno-
librium. It is represented as waterloaded basin sub- spheric heat flow has been allowed to vary slightly in
sidence (modified from Steckler and Watts, 1978; order to give minor adjustments to the rate of de-
Vejbaek, 1990) : cay of the thermal anomaly responsible for thermal
subsidence.
= Zpm(l~aTi)-M + 5_ Pm(l-Q^i) The phase transformation mechanism included in
pm(l-aTi)-p pm(l-aTi)-pw the model is very similar to those proposed by Falvey
where Y is water loaded basement elevation, pm is and Middleton (1981) and Sandiford and Powell
mantle density, pc is average crustal density, pw is (1986). It assumes that stretching causes a release of
water density, Tx is the mantle temperature at the volatiles in the crust that increase its density, pro-
base of the lithosphere, a is the thermal expansion vided the volatiles escape. The theoretical pressure-
coefficient, Z is decompacted sediment thickness, W& temperature path that rocks in the vicinity of the
is palaeo-water depth and AS\ is palaeo-sealevel. Moho follow during stretching and subsequent cool-
The total mass of the sediments (M) is given by: ing is shown in Fig. 5. During stretching, the rocks
first experience a pronounced pressure drop without
any significant fall in temperature. During thermal
M = Pw / (')'+ 5(,)[1-(,)]' subsidence temperatures slowly decrease. A slight
Jo Jo
pressure increase owing to sediment loading occurs,
where (') is the porosity-depth function and ps(zf) but is not included in the model. These strong
is the density-depth function of the sediment matrix. pressure/temperature changes are assumed to cre-
ate possibilities for a granulite-type metamorphism
The geodynamic model exemplified by the serpentine-forsterite transition.
The model that was used for geodynamic mod- This transition may not be a typical granulite fa-
elling (Vejbaek, 1989) differs from the original con- des process, but is chosen because densities honour
cept of McKenzie (1978) mainly by including heat the data and fit very well into the simple two layer
production in the lithosphere, allowing subsidence to description of the system. The traditional eclogite/
be enhanced by phase transformations and by sup- gabbro system (represented by the eclogite-garnet
plying a flow condition at the lower boundary below granulite transition) is unable to contribute to sub-
the lithosphere. The heat flow definition at the base sidence during stretching, because a volume of rock
of the model frame differs from the frequently used
temperature definition. The model handles thermal
perturbations caused by one or several phases of
GPa
stretching, variations in asthenospheric heat flow and
sill intrusions.
Three heat sources are assumed to govern the
steady-state conditions prior to stretching: in the
crust, in the lithospheric mantle and a heat flow
from the asthenospheric into the lithosphere. By
including these sources, it is possible to fit modelled
surface heat flow to observed surface heat flow and
to match present lithospheric thickness, as well as
honour constraints imposed by burial history. The
latter causes a fourth thermal contribution, in the
300 400
form of a relatively high transient component in Temperature
present surface heat flow. Fig. 5. Pressure-temperature diagram, showing the theoretical
Here the traditional temperature definition at the PIT time development in the vicinity of the Moho. One GPa
corresponds roughly to 30 km. Pressure decreases represent
base of the lithosphere is replaced by a heat flow periods of stretching (A to B), temperature decreases indicate
condition at a fixed depth. This is because the former periods with decay of the thermal anomaly (B to C). The
leads to large variations in asthenospheric heat flow former period is relatively fast, the latter is relatively slow. The
(Vejbaek, 1989, 1990). The heat flow condition used garnet-granulite to eclogite (relevant for the gabbro to eclogite
transition) and the serpentine to forsterite phase boundaries
also accounts for a longer duration of raised tran- (Kitahara et al., 1966) representing amphibolite to granulite
sient surface heat flow that is better able to honour transformations are also shown. G&R = Green and Ringwood
the data and may also explain the prominence of (1971); I&K = Ito and Kennedy (1971).
Geodynamic modelling of the Danish Central Trough 7

will move from the high density eclogite field to the TABLE 1
low density garnet granulite field during stretching Pertinent parameters common for all of the modelled wells. The
(Fig. 5). However, this phase transition may become crustal thicknesses etc. refer to the initial conditions before the
post-Middle Jurassic stretching commenced
important long after stretching. This may be of par-
ticular importance if gabbroic rocks are intruded Parameter Symbol Value
during, and as a consequence of, stretching. Crust cz 35.542 km
Lithosphere U 11.019 km
Heat production crust A cz 6.13 x 1 0 - 7 W/m 3
Heat production lithosphere \ 4.00 x 1 0 " 8 W/m 3
Application Asthenopheric heat flow 0.0325 W/m 2
Oast
Mantle density Pm 3330 kg/m 3
The evaluation of the post-Middle Jurassic sub- Crustal density Pc 2800 kg/m 3
sidence history of the Danish Central TYough is Water density Pw 1000 kg/m 3
Expansion coefficient a 3.28 x 10" 5 K " 1
based on analysis of seven wells (Fig. 1). Two of Thermal conductivity K 3.24 W/m/K
these (PW1 and PW3) are pseudo-wells, constructed Thermal diifusivity /v 8.0 x 1 0 " 7 m 2 ^ - 1
with the aid of seismic data and using rock types,
porosities, etc. derived by extrapolation from nearby
wells. The location of the seven wells are chosen The stretching history as obtained from the simu-
as representative of the central part of each sub- lations is summarized in Fig. 6. The tectonic activity
basin in order to minimize effects relating to higher began in the Callovian along the boundary faults
orders of dimension and, therefore, cannot be mon- of the Tail End Graben and Salt Dome Province
itored by the one-dimensional model. These effects trend, where large stretching factors are applied, ini-
include lateral heat flow (Cochran, 1983; Steckler, tial stretching factors are modelled to be 2.1, 1.6
1986; Moretti and Chenet, 1987). Data pertinent for and 1.4 for the PW3, the PW1 and the E-l loca-
the backstripping, such as palaeo-water depths, rock tions, respectively. These are crustal stretching fac-
types, thermal and hydraulic conductivities, etc., are tors, whereas the immediately subjacent lithospheric
given by Jensen (1990). mantle is thinned only by factors of 1.15 for the
In order to ensure comparability and mapability PW1 and PW3 locations and 1.03 for the E-l loca-
of the modelling results, the starting conditions for tion. The difference in stretching between crust and
each simulation are the same (Table 1). Present-day lithospheric mantle is necessary in order to match
crustal thicknesses follow those of Barton and Wood
(1984). According to their interpretations, based on
refraction seismic data, the crustal thickness exclud- Early Cret Late Cret. Tertiary
JrynelLo Fa. Va. USola cqChalk
ing sediments may be less than 10 km in central areas WELLS Fm pm|Fm |Fm. n Fm. pjGroup

of the adjacent Norwegian Central Trough.

The heat production parameters for the crust and (SS/3
lithospheric mantle and the average asthenospheric
heat flow are fitted to data in the E-l well to yield

present-day heat flow values (Madsen, 1975; Jensen,

1990). Madsen (1975) estimated the heat flow to be
in the order of 59 mW/m2 for the Danish Central
Trough area, but this value was not corrected for EL ^^~

palaeoclimatic effects and may thus be around 10 r^

I ~
mW/m2 higher (e.g., Balling, 1979). It is important to (3=2.0

note that the temperatures of the sediments within (3--1.5

(3 = 1.0
the Central TYough are unlikely to be in steady-
state equilibrium with regard to basement heat flow.
This is due to effects from compaction and fluid
flow caused by the high deposition rates in the Fig. 6. Diagram showing the timing, amount and the duration
Quaternary (Fig. 3), which are likely to perturb the of stretching events. All stretching factors are as shown for
the Sten-1 well. When the lithospheric mantle has not been
temperature profile. Results may, therefore, not be stretched by an equal amount as the crust, the mantle stretching
exactly comparable to present heat flow estimates factor is indicated in the block. The encircled values indicate the
obtained from wells, even if the modelling is correct introduction of volcanic sills simulated by a raised temperature
of 1400C. The first number indicates the depth to the top, the
with respect to heat flow at top basement level.
second the thickness of the sill in kilometers. Lo. Fm. = Lola
However, it is suggested that values between 60 and Formation; Fa. Fm. = Farsund Formation; Va. Fm. = Valhall
70 mW/m2 are acceptable. Formation; TX = Tuxen Formation; RB = R0dby Formation
8 O.VVejbcek

The 2-11-1 well, The Feda Graben The STEN-1 well, The Feda Graben

Tertiary . Cretaceous Jurassic Tertiary , Cretaceous . Jurassic

INeogenel Palaeogene I Late I Early | Late I Middle [ E a r l y INeogenelPalaeogene I Late I Early I Late I Middle [ E a r l y
100 120
M.y .B.R M y B.P

The MONA-1 well, The Gertrud Graben The W-1 well, The Heno Plateau

1000-

2000-

3000-

4000-

5000-
Tertiary . Cretaceous Jurassic Tertiary Cretaceous i Jurassic
NeogenelPalaeogene I Late I Early I Late | Middle | Early
6000 HNeogenel Palaeogene Late I | Early I Late I Middle | Early
100 120 U0 100 120 UO
M.y. B.P. M. y. B . P.

The P W 3 w e l l , The Tail End Graben The PW1 well, The Salt Dome Province

Tertiary . Cretaceous Jurassic Tertiary Cretaceous i Jurassic

iNeogenel Palaeogene I Late I Early I Late [ M i d d l e [Early Neogenel Palaeogene Late [ Early I Late [ Middle I Early
120 UO
M .y . B . P.

The E-1 well, The Salt Dome Province

- B a c k s t r i p p e d t e c t o n i c subsidence
-Modelled tectonic subsidence

Tertiary , Cretaceous Jurassic

NeogenelPalaeoqene I Late | Early I Late [ M i d d l e [Early

Fig. 7. Backstripping and the best fitting modelled tectonic subsidence curves of the seven wells (Fig. 1). All curves represent
water-loaded basement movements where effects of sediment loading have been subtracted. The actual sediment accumulation is about
twice as large. The general mis-match between the Cenozoic portion of the modelled and the backstripped curves is intentional,
because no mechanism in the model can provide a satisfactory explanation of that part of the backstripped subsidence curve. See text
for discussion.

the backstripped curves, which display remarkably Although stretching clearly has occurred as seen
little thermal subsidence following stretching. This in seismic data (Fig. 8), the tectonism has given rise
is particularly evident in the case of the E-1 well to surprisingly little anomalous heat build-up in the
(Fig. 7). lithosphere during the stretching. The small stretch-
Geodynamic modelling of the Danish Central Trough 9

W
OD ,0.0

1.0 1.0
E
- - 2.0I

s
'<-/~^yT-v,' r,\'^i^v v'NO i r r -i
-/r,.r'/r5^/^/3T/x'r,/">^/-T/lt^,vvO.^H
Heno Plateau | Ringkabing Fyn High

aij Pleistocene I I Middle & Upper Jurassic

| Tertiary B Middle Jurassic
I; ;| Upper Cretaceous and Danian [ j I Upper Palaeozoic- Triassic
yiliil Lower Cretaceous k ' / / i Basement
Fig. 8. Profile across the Tail End Graben (Fig. 1). Note the zone with an increased dip in the central to eastern part of the graben
(RD?), which may be equivalent to the reverse drag of Groshong (1989). The divergence of reflectors in the Upper Jurassic succession
into the graben reflects the syn-depositional block-rotation. AE. G = Arne-Elin Graben.

ing factors for the lithospheric mantle accounts for crease modelled thereby also enhances modelled
this and requires the presence of a detachment level phase transformations. This activity is simulated
in the vicinity of the Moho (cf. Coward, 1986). The as sill intrusions with a temperature of 1400C
apparent block rotation of the Tail End Graben along (Fig. 6). The most important locations for intru-
the boundary fault to the Ringk0bing Fyn High may sions were under the Upper Jurassic depocentres
suggest a deep detachment for the fault, possibly via at the PW3, PW1 and E-l locations, where density
a listric fault geometry (Fig. 8). The fault-controlled increases are seen caused by the intrusion simu-
localized Late Jurassic subsidence is also reflected lation equivalent to thinning factors of 3, 1.4 and
in the isopach map for the Upper Jurassic, which 1.5, respectively. Subsidence originating from phase
shows thicknesses in excess of 3.5 km adjacent to the changes modelled for the N2/11-1, the Sten-1 and
boundary fault between the Tkil End Graben and the the Mona-1 locations only equals thinning factors of
Ringk0bing Fyn High (Fig. 9). 1.05, 1.05 and 1.02, respectively, but has no effect at
The tectonic activity gradually spread to the west the W-l location.
during the Late Oxfordian, culminating around the The relatively few stretching events needed at the
Kimmeridgian-Volgian transition. This is simulated PW1, PW3 and E-l locations compared to the other
as several stretching events with maximum factors locations may be caused by the fact that they are
of 1.25 at the Mona-1 and W-l locations and 1.2 located in the main Upper Jurassic depocentres. As
at the N2/11-1 and the Sten-1 locations. Although the later events at the other locations are relatively
stretching in these locations (except for the W-l small, they may have been masked completely by
location) also is concentrated in the crust, the differ- compaction at the PW1, PW3 and E-l locations,
entiation is less pronounced with factors of 1.1 to 1.2 thus preventing their detection in the subsidence
for the lithospheric mantle. The migration of exten- curves. However, in the Hauterivian and the Late
sion westward is also evidenced by the progressively Cretaceous interval, stretching of up to 1.3 is applied
younger sediments on top of the basal Middle Juras- at the PW1 and PW3 locations, whereas extension
sic unconformity westwards indicating later onset of ceased in all other areas. These areas continued
subsidence (Johannesen, 1988). to subside, whereas compressional fault activity af-
In addition to the initial stretching, simulation fected most of the other locations (cf. Vejbask and
of intrusive activity is required to enable a bet- Andersen, 1987).
ter fit to the initial part of the thermal subsidence The modelled curves fit well for the pre-Cenozoic.
curves following main rifting. The temperature in- However, the Cenozoic portion of the curves fits only
10 O.V.Vejbcek

THICKNESS OF SEISMIC
SEQUENCE CORRESPONDING T
FARSUND FORMATION
Legend:
Normal / Reverse fault
Salt structure
> 3000 m
2250-3000m
1500 -2250m
750 - 1500 m
<750m

10 20 Km

Fig. 9. Isopach map of the Upper Jurassic (based on M0ller, 1986).

for the Palaeogene of the PW1 well (Fig. 7). Thus, the basis of the seismic data (Fig. 8). The North
the Cenozoic period is characterized by increas- Sea region formed a regionally subsiding basin al-
ing subsidence rates despite the absence of faulting most devoid of any faults during the Cenozoic (e.g.,
(Figs. 3 and 8), as has also been pointed out by Ziegler, 1982, pp. 89). None of the mechanisms ac-
Thorne and Watts (1989). As thermal subsidence is counted for in the geodynamic model can explain
expected to decrease with time corresponding to the this subsidence pattern. Consequently, no attempt is
decay of the underlying thermal anomaly, this pre- made to fit these parts of the curves. However, as
cludes the possibility that the Cenozoic was governed stretching probably is the most efficient way of per-
only by thermal subsidence following the preced- turbing the heat flow, the anomalous Cenozoic subsi-
ing stretching. Stretching cannot be postulated on dence pattern probably had affected the heat flow
Geodynamic modelling of the Danish Central Trough 11

L-Average Density ( g / c m 3 ) . 2.603

+ Field data
Model respons
Distance (Km)
'- ':'.
<"> lvV:.'V:V' ' '"-
' ''' '''-'

-I 1 1 1 1 1 I I I I I I i i I i I
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0

Legend:
Densities
;:;:j:;:ij:j:::::::] 2.05 g/cnr? Pleistocene 2.45g/cm Mesozoic
3
;:!?;;:;:;] 2.19 g/cm Tertiary 2.63g/cnrf Upper Palaeozoic - Middle
Jurassic
3 3
I 2.37 g/cm Upper Cretaceous and P A V / ^ . I 2.76 g/cm Crystalline Crust
. Da n i a n
I 2.53 g/cm 3 Upper Jurassic- Lower 3.32 g/cm 3 Mantle
Cretaceous

Fig. 10. Tentative gravity modelling across the Tail End Graben using the two-dimensional foreward modelling scheme of Talwani et al.
(1959) (Fig. 1). Gravity data are from Michelsen (1982). Sediment densities are based on Uldall and Balling (1989).

only slightly, although the mechanism remains an W-l, 21 km at Sten-1 and 22 km at N2/11-1. These
enigma. thicknesses are generally somewhat larger than the
Total stretching, as indicated by the modelling has 10-15 km of Barton and Wood (1984), except for
a maximum of 3.15 at PW3, decreasing from there, PW3, which may be regarded as a local extreme. The
to 2.4 at PW1, 1.6 at E-l and Sten-1, 1.8 at Mona-1 discrepancy between Barton and Wood's values and
and W-l and 1.5 at N2/11-1. Enhancement of sub- the values in the other wells is suggested to reflect
sidence by simulation of phase transformations is the present inability to model the Cenozoic subsi-
only significant at PW3, PW1 and E-l as mentioned dence. The rather large crustal thickness variations
above. These factors have resulted in crustal thick- seems to be in agreement with gravity data (Fig. 10).
nesses excluding sediments of only 4 km at PW3, Mapping of Lg-wave attenuation transmitted across
9 km at PW1, 15 km at E-l, 20 km at Mona-1 and the Central Trough area has also been suggested to
12 O.VVejbcek

The 2-11-1 well, The Feda Graben The STEN-1 well, The Feda Graben
KO.OO- KO.oo-|

120.00- 120.00-

100.00- 100.00-

* 80.00

E 60.00

/.0.00-

Tertiary Cretaceous ( Jurassic Tertiary , Cretaceous Jurassic

[NeoqenelPalaeoqene I Late I Early I Late I Middle [ E a r l y |Neoqene|Palaeoqene I Late I Early | Late I Middle I Early
ioo
M.y B.P

The MONA-1 well, The Gertrud Graben The W - 1 well, The Heno Plateau
U0.00- uo.oo-
120.00 -I 120.00-

Tertiary Cretaceous Jurassic Tertiary Cretaceous Jurassic

NeoqenelPalaeogene I Late I _ Early | Late I Middle ^Early NeoqenelPalaeoqene I Late I Early I Late I Middle I Early

The PW3 well, The Tail End Graben The PW1 well, The Salt Dome Province

J
uo.oo- uo.oo-
120.00 120.00-1

| Tertiary Cretaceous Jurassic Tertiary . Cretaceous Jurassic

NeoqenelPalaeoqene I Late I Early Late I Middle I Early NeoqenelPalaeoqene 1 Late 1 Early | Late I Middle | Early
100 120
< y B P

The E-1 well, The Salt Dome Province

U0.00

120.00 -Surface heat flow

--Asthenospheric heat flow
100 00-1

80 00-1

E 60CO-

40.00-1

2000-1
Tertiary , Cretaceous Jurassic
0.00 [NeoqenelPalaeoqene | Late 1 Early | Late | Middle I Early
0

Fig. 11. Calculated heat flow for the seven wells (Fig. 1). Compare with Fig. 7.

be indicative of large horizontal gradients in crustal the calculated situation at 131, 96, 60 and 0 mil-
thickness (Kennett et al, 1985). lion years. Machine contouring is used to avoid bias
from ill-defined structural constraints on heat flow
Heat flow development patterns. Map A (Fig. 12) depicts heat flow near
The calculated heat flow history is shown as plots the beginning of the Ryazanian (131 million years).
of heat flow versus time (Fig. 11) and is summa- The Tail End Graben area still had a relatively high
rized in contoured heat flow maps (Fig. 12) depicting heat flow, although activity had abated according to
Geodynamic modelling of the Danish Central Trough 13

E 330' EC EC30* E 3"30' EC E430'

_N2/11-1 J^ind\i A 1 Wn^ V

* T /^

3fiten -1(V
K a c-
VS.

0
\ litxvn

i \ 1 \ ] \ <svrr ILJA^A^ -JA-

W-1
\u\rHJ \ V

_ \ i I ^ \ I if r- \
/ i7i\^\\F\VTr

J
A 131.M.y.B.P (Beginning of Ryazanian) B 96.M.y.B.P. (Beginning of Late
La V I
Cretaceous)

E 330' EC 30' E 330' C E430'

mrt&i / lirrd^T /
__N2/11M ___N2/11-1

T
j^3^
steten-1 in
J^Vbten ^1 0s
vhv \ 1 \
in

[W 1\
PW3\ pyov
k/ W-1 h^i \KT

f-
V. o
x^L / \ ,
TN ^
W\
* (\
in
y |

\ N
(" \
\ 1r&
I PW1 M
j-BC-1

I \< I \i I I
C 60. M.y.B. P. (End of Daman ) D 0. M.y.B.P. ( Present \

Fig. 12. Maps showing calculated heat flow in latest Jurassic (131 million years) end of Early Cretaceous (96 million years) end of
Danian (Early Paleocene, 60 million years) and at present. The rifting activity is reflected in the heat flow maximum in the Tail End
Graben on the first map. Subsequently, the heat flow decayed. At present the highest heat flow is found in the Heno Plateau area.

the calculation. The differences are probably overes- Apart from the southern part of the area, which
timated, because an anomaly of this short duration also has relatively high values, the northern and the
(less than 5 million years, Fig. 11) is likely to be western parts of the area display heatflowvalues that
subdued by the effects of the sediment cover and are above steady-state conditions. Even though this
lateral heat flow, not included in the model (cf. Lu- situation represents maximum heat flow at E-l, it is
cazeau and Douaran, 1985; Steckler, 1986; Moretti below that of the nearby PW1 and PW3 locations,
and Chenet, 1987). The strong gradient at the PW3 apparently because of the pronounced difference
location should cause significant lateral heat transfer between crustal and lithospheric stretching at E-l. If
and some error in the estimate of magnitude and most of the stretching is in the crust, only a small
duration of heat flow. thermal anomaly will result and it will decay faster
14 O.VVejbcek

than when the whole lithosphere is stretched. This Yield stress ( - )()
correlates with the end of the Hot Shale deposition in 5 loop
the Farsund Formation, but effects on the maturation
patterns should be minimal, as no overburden had
accumulated to blanket in the heat.
Map B shows the beginning of the Late Creta-
ceous, when faulting had decreased. However, at this
time heat flow was close to its peak at W-l, although
it never reached values as high as in other places. At
this time, the present-day heat flow pattern started
to develop and maps C and D have similar trends,
but lower values. However, some areas cooled faster
than others. The area at W-l and Mona-1 cooled
only slightly during that period, whereas the PW3
location especially, and to some extent the PW1 lo-
cation, cooled much more. The E-l location seems
to have remained fairly cold through the whole 131 Fig. 13. Theoretical yield stress diagram for the lithosphere
million-year period. This is explained by asymmet- based on Lynch and Morgan, 1987). The diagram shows the
calculated distribution of the horizontal deviatoric stress needed
ric extension, that causes low anomalous heat flow to cause non-elastic extension. The presence and size of the yield
and different cooling trends relative to other loca- stress minima are strongly dependent on ambient temperatures
tions. This pattern is due to both the amount and and, thus, heat flow along with other things such as strain rate.
depth of anomalous heat. Map D reflects the present
remaining heat anomaly and differences in crustal
thickness, because little anomalous heat is left and sphere may give rise to yield stress minima, because
the steady-state heat production potential in the the temperature at which ductility increases drasti-
model is dependent on the crustal thickness. cally (i.e., the rocks approach their melting point)
The present lithospheric thicknesses calculated, as is different for different rock types. A yield stress
defined by the depth to the 1330C isotherm, reflect minimum may occur, if the lithosphere is heated,
the amount of transient heat left in the lithosphere and disappear again as it cools. The resulting decou-
and are consequently roughly inversely correlated pling level (i.e., at the base of the crust) makes the
with present surface heat flow. In decreasing order lithosphere mechanically discontinuous and renders
thicknesses are 95.2 km at N2/11-1, 87.7 km at uniform stretching unlikely. As syn-rift subsidence
Sten-1, 86.3 km at E-l, 81.5 km at PW1, 81 km at at the Upper Jurassic depocentre in the Tail End
Mona-1 and 78 km at PW3. A thickness of 77.4 km Graben clearly was accompanied by major faulting
at W-l coincides with the present maximum surface along the boundary fault to the Ringk0bing Fyn High
heat flow. A significant transient component is still (Figs. 8 and 9), it is likely that movement along a
present in the surface heat flow, as the approximate possible detachment level was intimately connected
ultimate steady-state heat flow is estimated to vary to movements along the fault. It is difficult to ascer-
between 43 mW/m2 at PW3 and 50 mW/m2 at tain whether the main fault is listric or planar, owing
N2/11-1 (cf. Fig. 12). This variation is dependent to absence of direct fault plane images in the seismic
on stretching factors alone, since conservation of data and to uncertainties in velocities for the neces-
heat producing elements is assumed during phase sary depth conversion. However, the apparent block
transformations. rotation of the Tkil End Graben along the bound-
ary fault to the Ringk0bing Fyn High may suggest a
Discussion deep detachment. Analysis of recent earthquake data
shows that active faults are all approximately planer
The notion that stretching in the crust locally and evidence for detachments in a ductile lower crust
may exceed stretching in the lithospheric mantle is is also sometimes present (Jackson, 1987). Jackson
based on the conspicuous absence of thermal post- (1987) suggests that rotation in the brittle upper
rift subsidence, best seen in the case of the E-l crust is mainly due to domino-type fault plane rota-
well. This differential stretching, accommodated by tion by continuing extension. However, whether the
a detachment level, may be supported by theoretical fault plane is rotating or is curved, both may be
studies of tensile strength profiles of the lithosphere accommodated by a ductile lower crust.
(Fig. 13). According to Lynch and Morgan (1987), According to Groshong (1989) the depth to a
the rock composition change with depth in the litho- detachment assuming a planar fault geometry may
Geodynamic modelling of the Danish Central Trough 15

faulting, the latter with no faulting. Faulting ceased

in earliest Tertiary, apart from local faulting at salt
diapirs. As stretching, therefore, can be ruled out as
a cause for Cenozoic subsidence, it is suggested that
some metamorphic process leading to a density in-
crease and additional subsidence is active. Because it
cannot be induced by stretching, the mechanism must
Fig. 14. Sketch showing the concept of reverse drag that may be different from that incorporated in this study. The
be used to determine the depth of a detachment level (based
mechanism of the present study is mainly sensitive
on Groshong, 1989). The abrupt change in the fault plane dip
requires an equally abrupt change in the ductility. Compare with to pressure drops and temperature increases, which
Fig. 8. take place during convective uplift. However, the
mechanism causing Cenozoic subsidence is suggested
to be a self-supporting interaction between sediment
be estimated from the distance between the reverse loading and a pressure-induced phase transforma-
drag section in the hanging wall and the main fault tion. A thermal drive for the transformation does
and the dip of the fault (Fig. 14). This model seems not seem to be a viable mechanism, because the
to be applicable to the Tkil End Graben profile, as immediate response, as seen in the case of the Qua-
a reverse drag section seems to be present in the ternary subsidence and deposition, cannot possibly
hanging wall (Fig. 8). Unfortunately, the exact geom- be thermally driven as a consequence of the insu-
etry of the fault is ill-constrained, for reasons stated lating effect of sediments. It takes a very long time
above. However, the depth to a possible detachment before the insulating effect is felt at depths where
level below the Tkil End Graben is likely to be lo- possible metamorphic reactions occur. It is suggested
cated within the first second of two-way reflection that the mechanism is driven mainly by pressure,
time below the top of the basement, corresponding because this will provide an immediate response to
to at least 2 km. The calculated very thin crust below an extra load, as may be the case with the Qua-
the Tkil End Graben may, therefore, be supported by ternary loading. An obvious mechanism may be the
this interpretation of the fault and seems to honour well-known eclogite/garnet-granulite system, which is
gravity data (Fig. 10). pressure-sensitive (e.g., O'Connell and Wasserburg,
The conclusive tentative model for the middle to 1972; Haxby et al., 1976). As seen in Fig. 5, an extra
northern Danish Central TYough may be similar to load may cause portions of the lower crust to enter
the sketch on Fig. 4, which is based on theoreti- the eclogite field and thus lead to a density increase.
cal studies by Braun and Beaumont (1989). They However, the lithosphere should be close to ther-
showed, by application of a two-dimensional finite mal steady-state to allow the pressure mechanism to
element model, that extension in the lithospheric be active (Fig. 5), so some cooling is required to
mantle may be transferred laterally within about 100 make the system sensitive to loading. To incorporate
km to a pre-existing weakness zone in the crust, the eclogite/garnet-granulite system in the model, a
provided a detachment level is also developed. This much larger complexity of the equations is required
results in different stretching factors for the litho- as sediments have to be directly included in the
spheric mantle as compared to the crust at points calculations. This is regarded as extremely difficult
along a line perpendicular to strike (Fig. 4), although with the semi-analytical approach used in the theory
the regional stretching values are the same. If the of the present model and, therefore, has not been
boundary fault between the Tkil End Graben and attempted.
the Ringk0bing Fyn High corresponds to this pre-
existing crustal weakness zone, then this model may
explain the large localized stretching factors applied Conclusion
to the crust as well as the small stretching factors
applied to the lithospheric mantle at PW3, PW1 and The post-Middle Jurassic tectonic development
E-l. started with major faulting along the boundary fault
A major problem is the Cenozoic subsidence, between the Tkil End Graben and the Ringk0bing
which cannot be explained by the geodynamic model. Fyn High, thus giving rise to the well-defined de-
The problem is particularly conspicuous when com- pocentres of the Farsund Formation. Southwards
parison is made between the deposition rates of the towards the Salt Dome Province this depocentre di-
Upper Jurassic Farsund Formation and the Quater- vides, because stretching was accommodated by two
nary Cen Unit 6 (Fig. 3), which are comparable. The major faults. This early faulting was much stronger
former is a syn-rift deposit associated with strong than the later activity. The resulting heat flow also
16 O.VVejbcek

was at its peak during this period, with a maximum Day, G.A., Cooper, B.A., Andersen, C , Burgers. W.F.J.,
in the Tail End Graben. R0nnevik, H.C. and Sch0neich, H., 1981. Regional seis-
mic structure maps of the North Sea. In: L.V. Illing and G.D.
The tectonism gradually spread to the west during Hobson (Editors), Petroleum Geology of the Continental
the Upper Jurassic, leading to a later maximum heat Shelf of North-West Europe. Institute of Petroleum/Heyden,
flow in this region, but with a lower level than in London, pp. 76-84.
the Tkil End Graben. On the basis of the tectonic Dewey, J.F., 1982. Plate tectonics and the evolution of the British
history, it is concluded that much of the variation Isles. J. Geol. Soc. London, 139: 371-412.
Eynon, G., 1981. Basin development and sedimentation in
in subsidence patterns in the Danish Central Though the Middle Jurassic of the northern North Sea. In: L.V.
is generated by differential stretching between crust Illing and G.D. Hobson (Editors), Petroleum Geology of
and lithospheric mantle. However, during the Creta- the Continental Shelf of North-West Europe. Institute of
ceous, stretching gradually abated and gave way to Petroleum/Heyden, London, pp. 196-204.
regional subsidence. This subsidence could not be Falvey, D.A. and Middleton, M.F., 1981. Passive continental mar-
gins: evidence for a pre-breakup deep crustal metamorphic
modelled, as the suggested pressure-induced density subsidence mechanism. Proc. 26th Int. Geology Congr., Ge-
increases from gabbro-eclogite metamorphism is not ology of Continental Margins Symp., Paris, 7-17 July, 1980,
represented in the model of the present study. Oceanol. Acta, 4: 103-114.
Gowers M.B. and Saeb0e, A., 1985. On the structural evolution
of the Central Trough in the Norwegian and Danish sectors
Acknowledgements of the North Sea. Mar. Pet. Geol., 2: 298-318.
Green, D.H. and Ringwood, A.E., 1972. A comparison of recent
The STATOIL Group and the Danish Research experimental data on the gabbro-garnet granulite-eclogite
transition, J. Geol., 80: 277-288.
Academy are thanked for financial support of the Groshong, R.H., Jr., 1989. Half-graben structures: balanced
research project upon which this paper is based. The models of extensional fault-bend folds. Geol. Soc. Am. Bull.,
working team of the Danish Modelling Group is 101: 96-101.
also thanked for making their studies of the subsi- Haq, B.U., Hardenbol, J. and Vail, PR., 1987. Chronology of
dence history using the Ykler model available for fluctuating sea levels since the Triassic (250 million years ago
to present). Science, 235: 1156-1167.
the geodynamic studies of this paper. Comments and Haxby, W.F., Turcotte, D.L. and Bird, J.M., 1976. Thermal and
suggestions of Claus Andersen, Finn Surlyk, John mechanical evolution of the Michigan Basin. Tectonophysics,
Sales (who acted as referee) and one anonymous 36: 57-75.
reviewer have been helpful and are gratefully ac- Ito, K. and Kennedy, G.C., 1971. An experimental study of the
knowledged. Finally I thank Kaj Jensen-J0rgensen gabbro-garnet granulite-eclogite transition. In: J.G. Heacock
(Editor), Am. Geophys. Union, Geophys. Monogr., 14, pp.
(Geokon) for drafting the illustrations. 303-314.
Jackson, J.A., 1987. Active normal faulting and crustal extension.
In: M.P. Coward, J.F. Dewey and PL. Hancock (Editors),
References Continental Extensional Tectonics. Geol. Soc. London, Spec.
Publ., 28: 3-17.
Andersen, C , Olsen, J.C., Michelsen, O. and Nygaard, E., Jacobsen, F., 1982. Triassic. In: O. Michelsen (Editor), Geology
1982. Structural outline and development. In: O. Michelsen of the Danish Central Graben. Danm. Geol. Unders., Ser. B,
(Editor), Geology of the Danish Central Graben. Danm. 8: 32-36.
Geol. Unders., Ser. B, 8: 9-26. Jensen, P.K., 1990. Analysis of the temperature field around salt
Balling, N.P., 1979. Subsurface temperatures and heat flow esti- diapirs. Geothermics, 19(3): 273-283
mates in Denmark. In: V. Cermak and L. Rybach (Editors), Johannesen, P., 1988. Upper Jurassic sandstones Part 1. De-
Terrestrial Heat Flow in Europe. Springer-Verlag, Berlin, pp. positional environments and basin development, CENBAS
161-171. Sub-project B. Reservoir rocks and depositional models.
Barton, P. and Wood, R., 1984. Tectonic evolution of the North Danm. Geol. Unders., Internal Rep.
Sea Basin: crustal stretching and subsidence. Geophys. J. R. Kennett, B.L.N., Gregersen, S., Mykkeltveit, S. and Newmark,
Astron. Soc, 79: 987-1022. R., 1985. Mapping of crustal heterogeneity in the North Sea
Braun, J. and Beaumont, C , 1989. Dynamical models of the role basin via the propagation of Lg-waves. Geophys. J. R. Astron.
of crustal shear zones in asymmetric continental extension. Soc, 83: 299-306.
Earth Planet. Sei. Lett., 93: 405-423. Kitahara, S. Takenouchi, S. and Kennedy, G.C., 1966. Phase
Calcagnile, G. and Scarpa, R., 1985. Deep structures of the Eu- relations in the system M g O - S i 0 2 - H 2 0 at high temperatures
ropean Mediterranean area from seismological data. Tectono- and pressures, Am. J. Sei., 264: 223-233.
physics, 118: 93-111. Lucazeau, F. and Le Douaran, S., 1985. The blanketing effect of
Chiarelli, A. and Duffaud, F., 1980. Pressure origin and distribu- sediments in basins formed by extension: a numerical model.
tion in Jurassic of Viking Basin (United Kingdom-Norway). Application to the Gulf of Lion and Viking graben. Earth
Am. Assoc. Pet. Geol., Bull., 64: 1245-1266. Planet. Sei. Lett., 74: 92-102.
Cochran, J.R., 1983. Effects of finite rifting times on the Lynch, H.D. and Morgan, P., 1987. The tensile strength of
development of sedimentary basins, Earth Planet. Sei. Lett., the lithosphere and the localization of extension. In: M.P.
66: 289-302. Coward, J.F. Dewey and PL. Hancock (Editors), Continental
Coward, M.P., 1986. Heterogeneous stretching, simple shear and Extensional Tectonics. Geol. Soc. London, Spec. Publ., 28:
basin development. Earth Planet. Sei. Lett., 80: 325-336. 53-65.
Geodynamic modelling of the Danish Central Trough 17

Madsen, L., 1975. Approximate geothermal gradients in Den- 64: 49-60.

mark and the Danish North Sea Sector. Danm. Geol. Unders., Thorne, J.A. and Watts, A.B., 1989. Quantitative analysis of
rbog, 1974: 5-16. North Sea subsidence. Am. Assoc. Pet. Geol., Bull., 73: 8&-
McKenzie, D.P., 1978. Some remarks on the development of 116.
sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32. Uldall, A. and Balling N., 1989. Basin analysis of the Dan-
Michelsen, O. (Editor), 1982. Geology of the Danish Central ish Central Graben. CENBAS Sub-project F: Petrophysics,
Graben. Danm. Geol. Unders., Ser. B, 8, 133 pp. Department of Geology, Geophysical Laboratory, Aarhus
Moretti, I. and Chenet, P.Y., 1987. The evolution of the Suez University, 95 pp.
rift. A combination of stretching and secondary convection, Vejbaek, O.V., 1986. Seismic stratigraphy of the Lower Creta-
Tectonophysics, 133: 229-234. ceous in the Danish Central Trough. Danm. Geol. Unders.,
M0ller, J.J. 1986. Seismic structural mapping of the Middle and Ser. A, 11, 46 pp.
Upper Jurassic in the Danish Central Trough. Danm. Geol. Vejbaek, O.V. and Andersen, C , 1987. Cretaceous-Early Tertiary
Unders., Ser. A, 13, 37 pp. inversion tectonism in the Danish Central Trough. Tectono-
O'Connell, R.J. and Wasserburg, J.G., 1972. Dynamics of sub- physics, 137: 221-238.
mergence and uplift of a sedimentary basin underlain by Vejbask, O.V, 1989. Effects of asthenospheric heat flow in basin
a phase-change boundary. Rev. Geophys. Space Phys., 10: modelling exemplified with the Danish Basin. Earth Planet.
335-368. Sei. Lett., 95: 97-114.
Sandiford, M. and Powell, R., 1986. Deep crustal metamorphism Vejbaek, O.V, 1990. Geodynamic modelling of the Danish Basin.
during crustal extension. Modern and ancient examples. Earth In: B. Pinet and C. Bois (Editors), The Potential of Deep Seis-
Planet. Sei. Lett., 79: 151-158. mic Profiling for Hydrocarbon Exploration. Institut Francais
Sclater, J.G. and Christie, P.A.F., 1980. Continental stretching. du P6trole, 1989, Aries, Edition Technip, 10 pp.
An explanation of the post-mid-Cretaceous subsidence of the Wood, R. and Barton P., 1983. Crustal thinning and subsidence
Central North Sea basin. J. Geophys. Res., 85: 3711-3739. in the North Sea. Nature, 302: 134-136.
Steckler, M.S., 1986. Uplift and extension of the Gulf of Ykler, M.A., Cornford, C. and Weite, D., 1978. One-
Suez, indications of induced mantle convection. Nature, 317: dimensional model to simulate geologic, hydrodynamic and
135-139. thermodynamic development of a sedimentary basin. Geol.
Steckler, M.S. and Watts, A.B., 1978. Subsidence of the Atlantic Rundsch., 67: 960-979.
type continental margin oif New York. Earth Planet. Sei. Ziegler, P.A., 1975. Geologic evolution of North Sea and its
Lett., 41: 1-13. tectonic framework. Am. Assoc. Pet. Geol., Bull., 59: 1073-
Talwani, M., Worzel, J.L. and Landisman, M., 1959. Rapid gravity 1097.
computations for two-dimensional bodies with application to Ziegler, P.A., 1982. Geological Atlas of Western and Central
the Mendocino Submarine Fracture Zone. J. Geophys. Res., Europe. Elsevier, Amsterdam, 130 pp. and encl.

O.V. V E J B ^ K Geological Survey of Denmark, Thoravej 8, DK-2400 K0benhavn, Denmark

Present address: Amoco Production Company, Houston, TX 77253-3092, USA
 19

A tectonostratigraphic analysis of the southeast Norwegian

North Sea Basin
S. S0rensen, H. Morizot and S. Skottheim

In the South East North Sea several wells have been drilled. Oil has been found in four of them. In order to better
understand the hydrocarbon distribution in the area, Elf Aquitaine Norge A/S made a regional study of the Egersund Basin. In
this paper the regional aspect with respect to structural history, depositional history and stratigraphy will be discussed.
The study includes a biostratigraphic update of 26 wells in general. The regional seismic study included the interpretation of
17300 km of seismic.
The area between 57N and 59N and 3 and 5 can be divided structurally into a mainly salt induced basin to the south of
58N, Egersund Basin sensu latio and a more fault controlled basin to the north of 58N, sta Graben.
The Late Jurassic period was dominated by an overall transgressive phase in the lower part and a regressive phase in the
upper part.
A compression during the transition from Early to Late Cretaceous has been observed and has a certain bearing on the Chalk
distribution of the area.
The seismic interpretation further allowed us to define the shift of basin axis through time and to define the importance of
the different faults through time and their geographical position. This, together with seismic stratigraphy, better age control and
seismic facies analysis increased our confidence as regards the mapping of the sediment distribution and facies shift through
time.
This presentation concludes with a set of paleogeographic maps constructed from well data, seismic facies analysis and the
structural mapping.

Introduction total of 17 300 km of seismic data were interpreted.

The sources were Maxipulse, Airgun or Watergun
The study was performed in order to evaluate and the coverage varied from 24-fold to 60-fold. The
the regional geological evolution of the area be- size of the seismic grid ranged from 5 x 5 km region-
tween 57N and 59N and 320 and 540. This ally to 1 x 1 km locally. A total of 27 wells were used
area includes several structural provinces, such as in the seismic calibration.
the sta Graben, Egersund Basin, Fiskebank Basin
and part of the Norwegian Danish Basin (Fig. 1).
Basement
During the study and in the text we have applied
the common term "South East North Sea" for these In the study area three wells (8/3-1, 10/5-1 and
structural provinces, and it is further abbreviated 18/11-1) reached metamorphic basement (Fig. 2).
to the "SENS study". The study includes a bios- Well 8/3-1 is located in the western part of the
tratigraphical update of several wells in addition study area on the southernmost part of the Sele High.
to seismic interpretation, geological modelling and The basement rocks in this well comprise schistose
reinterpretation. The methodology is a combination rocks of Silurian age (411 4 Ma) (Myhre, 1975).
of different geological and geophysical disciplines Well 10/5-1 is located on the Lista Nose structure. It
applied in order to better appreciate the sediment stopped in granites of indicated Pre-Cambrian age.
distribution in relationship to the tectonic evolution Well 18/11-1 reached a weathered and slightly folded
through time. quartzitic basement.

Data base Paleozoic

In this study, lines from seismic surveys with acqui- During the transition from Caledonian to Variscan
sition years ranging from 1977 to 1987 were used. A orogenesis, the stress pattern seems to have rotated

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 19-42. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
20 S. S0rensen, H. Monzot and S. Skottheim

3:0'

59 00'

57 00'

Fig. 1. Principal structural elements SENS (South East North Sea).

A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 21

WELL SUMMARY]
SOUTH EAST
NORTH SEA
22 S. S0rensen, H. Morizot and S. Skottheim

TABLE 1
Structural trends through time, with respect to geographical position: south-east North Sea

Time map Area

5630'-57N 57-5730'N 5730'58N 58-5830'N 5830'-59N
Basement D R DR S D (R) SD SD
Pre-Zechstein D K D(K) D(S) SR(K) S K R (S)
Top Triassic D K D (R) DK(R) S R (KD) D K R (S)
Top Middle Jurassic D K D (R) DK SR D KR
Top Kimmeridgian D(K) D(K) D(K) KDR KD
Base Cretaceous Unconformity D DK(R) D(KR) K D R (S) R K D (S).
Base Chalk D(K) DK KD KS R S

D = 130-150, Dinaric (Tornquist Trend); K = 150-170, Kattegat; R = 10-30, Rhenian (Oslo Trend); S = 35-55,
Skagerrak (Caledonian Trend).
(X) = Minor influence when trend between brackets.

clockwise from a NW-SE direction of compressional fault trends. However, the significance of the differ-
stress in Silurian times to a more N-S direction ent faults is rather uncertain, in the sense that the
of compressional stress in Eifelian-Givetian (Mid- throws on the faults are unknown at this time.
dle Devonian times). This clockwise rotation of the
stress field seems to have continued into a NE- Devonian
SW stress in Early Carboniferous times (Ziegler,
1988). While the NW-SE compressional direction The oldest sediments in the area might be of
represents the Caledonian orogeny, the N-S to NE- Devonian age. This age has been suggested for the
SW compressional direction represents the Variscan lowermost section penetrated in well 17/12-2. Its as-
orogeny. During the transition from Carboniferous sociation with the sandstones in the Hornelen Basin
to Permian times, the stress system changed from in Western Norway has been suggested (Jacobsen,
a predominantly compressional to a predominantly 1982).
extensional system. However, a Rotliegendes age for this sandstone
The fault pattern at basement level is dominated is favoured by these authors, based on regional
by three trends (Table 1). The area south of 58N interpretation.
is dominated by the 130-150N trend (Dinaric, Ta-
ble 1) initiated in Early Paleozoic times, which, as we Permian
shall see later, was reactivated during several later
periods. Rotliegendes
The second trend in the south, oriented 10- Clastic deposits of Rotliegendes age are likely to
30N, might be associated with an extension phase in be widely present in the SENS area, as indicated
Devonian-Permian times. This orientation is similar by their seismic character. At present only two wells
to the main trends of the Oslo Graben. The third (10/5-1 and 17/12-2) have reached the Rotliegendes
trend 150-170N is analogous to the orientation of series (Fig. 2).
the northern part of the Horn Graben. This fault Well 17/12-2 is located on the Sele High. As
trend was initiated prior to mid Carboniferous times. discussed in the paragraph on the Devonian, the
The NW-SE trends of 130-150N and 150-170N age dating is uncertain. However, based on the
are the main axes for the salt structures created in regional context it is believed that the rocks belong
Early THassic times. to the Rotliegendes suite. In well 10/5-1 only 7 m of
The area north of 58N is dominated by the 35- postulated Rotliegendes siltstones were penetrated
55N (Skagerrak or Caledonian) trend and the 130- before the well bottomed in Pre-Cambrian granite.
150N (Dinaric or Tbrnquist trend) at basement level, A northern Permian Basin probably existed over
indicating the Early Paleozoic initiation of these parts of or the whole SENS area, with connections
trends. On the Pre-Zechstein map (Fig. 3) the im- both to the southeast (Norwegian Danish Basin) and
portance of the 130-150N trend diminishes north maybe to the north into Stord Basin (Ziegler, 1988,
of 58N, and the NW-SE trending fault orientation S0rensen and Martinsen, 1987) (Fig. 1).
turns to 150-170N between 58N and 59N. The Rotliegendes rocks of Early Permian age were
The Carboniferous and Rotliegendes sediment dis- deposited in an arid climate with desert conditions.
tribution is assumed to follow the same controlling The arid conditions led to the deposition of aeolian
A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 23

3W 4'" 5 00' 6~00'

5800'

5700'

Faults ^ . Location of line

* - SiBGS 87-14
Salt domes piercing at Mid.Jur. level

Basin axis

Fig. 3. Fault orientation at Top Pre-Zechstein level.

24 S. S0rensen, H. Monzot and S. Skottheim

sands and sabkhas in structurally low areas. On the change in facies occurs (halite-carbonate). Between
Sele High (well 17/12-2) the continental deposits are the two fault patterns there was a fairly flat lying
mainly arkosic. or slightly westward tilted basin dominated by halite
During this period the sta Graben and Egersund deposits. Generally, the Late Permian experienced a
Basin already existed as sedimentary basins. The gentle subsidence in the central parts of the present
structural framework was initiated with a central basins. The subsidence axis was probably located in
basin flanked by the somewhat structurally elevated a N-S orientation east of the Sele High in the north
Stavanger Platform to the east and the Sele High to with the axis having a more NW-SE orientation to
the west. The boundary fault seen at Base Zechstein the south of 58N.
level along the eastern side of the Sele High was Ten of the wells in the SENS study area reached
in existence and governed the assumed alluvial fan or penetrated the Zechstein interval. These wells
deposits on the downthrown side of the Sele High represent all the different structural provinces. Of
(Fig. 3). A drainage area from the east is expected the ten wells, eight are in a definite salt facies
along the gently westerly dipping margins of the (8/3-1, 8/9-1, 9/4-1, 9/8-1, 10/8-1, 17/4-1, 17/11-1
Stavanger Platform. and 17/12-1), whilst two are in a marginal marine
Rotliegendes deposits are known to have fair to facies (10/5-1 and 17/12-2) (Fig. 2). The evaporitic
good reservoir characteristics, depending on the de- sequences (salts) are divided into four cycles for
gree of diagenesis. The facies range from alluvial, the North Sea (Rhys, 1974), however some authors
continental and fluvial to aeolian sandstones. Ac- (Ormaasen et al., 1980) claim that only three of
curately mapping this sequence is a problem. In the four cycles are present in the Northern Permian
places, it can be seismically observed as a transparent Basin.
(white) sequence below the pre-Zechstein marker, Zechstein deposits are believed to be restricted
whilst in other areas it cannot be mapped. to the west of the Stavanger Platform. If this is
the case, it has yet to be resolved whether the Sta-
Kupferschiefer equivalent vanger Platform was an area of non-deposition or an
This sequence is supposed to be widespread in the erosional feature during Zechstein times.
SENS area and blankets the Rotliegendes deposits, On the structural highs flanking the basin, we
being the beginning of the "marine" transgression find the marginal marine platform sediment, as in
at the end of Rotliegendes time. Generally, these well 10/5-1. The other well with Zechstein platform
deposits are dominated by organic-rich shales, which sediments is 17/12-2, located on the Sele High. This
can be of excellent source rock quality, but in re- well is located at a structurally higher position than
stricted quantity. the 17/11-1 which encountered salt deposits.
The Kupferschiefer is only drilled in the struc- The general thickness of the salt deposits in the
turally elevated areas, such as the Lista Nose (well SENS area is in the order of 1000-1500 m. As no
10/5-1) and on the southern side of Ling Graben northern limit of the deposits has been defined in the
(17/4-1), whilst it has been eroded or not deposited SENS area, the Zechstein evaporitic series should be
on the Sele High. In the structurally deeper parts of expected in the Stord Basin.
the basin the sequence has not been reached, but in
this interpretation it is assumed to be present. Triassic

Zechstein Triassic deposits have been penetrated in several

For the Zechstein deposits it seems that the fault wells. The thickest series encountered so far is in
activity was of minor importance during deposi- well 17/12-1 where more than 1500 m of TViassic
tion, i.e., that little synsedimentary faulting took rocks was penetrated. Even on the structural highs,
place. The area south of 58N is linked directly thick THassic series have been penetrated (e.g., well
to the Norwegian-Danish Basin, in what is termed 17/11-1) (Fig. 2).
the Northern Permian Basin (Ziegler, 1988). To the According to Vollset and Dore (1984), no revision
north of 58N, the Zechstein deposits seem locally to of the lithostratigraphy has been carried out. How-
be fault controlled. The deposits are thin on the east- ever, from the seismic data, it seems that the Triassic
ern side of the basin and do not extend beyond the sequence can be divided into three sequences. The
eastern boundary fault (Fig. 4), indicating that this lowermost part of the THassic series comprises con-
fault probably existed prior and during the Zechstein formable seismic reflectors on top of the underlying
deposition. On the western side the sediments cross (often Permian) sequences. The top of this series is
the eastern boundary faults of the Sele High. How- cut by an unconformity which we believe might be
ever, the sediment thickness is reduced and a clear the Hardegsen Unconformity.
A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 25

17/9-1 PROJ

__. .'T-rnrarrTj'iyiliy^

2.0-

m s m m w
^ ^ ^ ^ ^ ^ ^ " "
3.0-

X
___A ___-*"* """ ^ \ bHHl Shale ['.-'. ] Sandslon-2
4 0-
Sa
C = a Chalk [ ^ "

Fig. 4. Geoseismic section line SBGS 87-14.

The middle part of the THassic series contains a a somewhat more widespread distribution. During
fairly uniform series and the top of it is often marked the Middle THassic a marine transgression occurred
by a distinct seismic marker. The uppermost THassic from the south and southeast, epitomized by the
series is normally cut by an unconformity. distribution of the Muschelkalk. How far north the
The THassic period was affected by a global sea transgression can be traced is debatable, but the
level drop at the end of the Paleozoic era, chang- limestone stringers of probable Anisian age in well
ing the overall depositional environment from the 17/12-1 have been indicated to be of marine ori-
marine Zechstein deposits to the dominantly conti- gin (Ziegler, 1988). However, Jacobsen et al. (1982)
nental THassic deposits. The evolution of the THassic claim that these limestones are sabkha deposits. De-
in the area is related to the activity along the Torn- posits of lacustrine origin are found in wells 17/12-2
quist Tbisseyre Line, seen as the development of and 17/12-3.
the Horda-Egersund-North Danish system of THas- From in-house geochemical analysis of the THassic
sic half-grabens, sub-paralleling the southwest coast interval in well 17/11-2, a deltaic to lacustrine origin
of Norway and the west coast of Sweden (Ziegler, of the shaly sediments has been postulated (Fig. 6).
1988). Following these observations and the structural
The study area can be partly separated into two trends, we believe it likely that a fairly large lake
tectonic domains during THassic times, one severely existed periodically located on the eastern side of
affected by salt movement south of 58N (Egersund the Sele High, with sabkha deposits as seen in well
Basin) and one controlled by the westerly dipping 17/12-1 and somewhat more fluvial deposits on the
basement to the north of 58N (sta Graben) structurally elevated sections, like the Sele High (well
(Fig. 5). 17/11-2). The position of this lake is linked to the
For the whole TYiassic period the basin axes seem development of a half-graben east of the Sele High,
to be located on the east side of the Sele High going the sta Graben.
southwards where they split into several axes located During the Late THassic times the climate changed
between the active growing salt walls. A minor axis from arid to humid principally due to the northward
west of the Sele High in Block 17/10 and parts of plate movement of the area. In addition, there was a
17/11. strong rejuvenation of the tectonic regime, with dike
The distribution of Early THassic sediments gen- intrusions onshore (Sunnhordland) dated at 220 Ma
erally seems to be limited to the east by the Sta- (Faerseth et al., 1975) and volcanic activity offshore
vanger Platform (Fig. 4). The Middle THassic has (well 17/9-1).
26 S. S0rensen, H. Monzot and S. Skottheim

3*00' 400' 5O0' 600'

5roo

58c00'

5700'

Zechstein Salt Structures Structural High Areas

Faults 111111
X> Structural Low Areas
Fig. 5. Triassic basin configuration.
A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 27

5900*

5800

5700*

300' 400* 500 6

Alluvial
\\ Continental
ED Lacustrian
^ Alluvial fans
<> Dry well
Oil well

Fig. 6. Triassic palcogeography.

28 S. S0rensen, H. Monzot and S. Skottheim

The clastic input from the Fennoscandian hinter- (Fig. 8). Southeast of this, ther seems to be a new
land was probably significant. The increased chemical threshold before the Early Jurassic basin axis extends
weathering of the Pre-Cambrian feldspar-rich rocks into the Norwegian Danish Basin.
to the east due to the climate change led to an in- The extension of the Early Jurassic deposits is de-
creased development of clay. This may partly explain batable, but based on their seismic facies character,
the increased shale content in the Upper TYiassic they can be traced almost to the eastern edge of
(pre-Rhaetian) series in the SENS area. the SENS area in the northeast (Blocks 18/5 and
In the very Late TYiassic time (Rhaetian) the sea 18/6) and also into the area named Fiskebank Basin
level rose and the rate of subsidence in the SENS (Blocks; 8/3, 8/6 and 8/9).
area decreased as one entered the tectonically un-
eventful Early Jurassic period. The Late TYiassic Middle Jurassic
transgression led to the deposition of basal trans- The Middle Jurassic deposits range in age from
gressive sands termed the Gassum Formation of Aalenian through Callovian. However, from a se-
Rhaetian age. quence point of view the Callovian deposits belong
to the Upper Jurassic transgressive mega cycle. For
Jurassic the SENS study area, the Middle Jurassic can be split
into two main cycles.
Early Jurassic (a) The Aalenian through Bathonian generally flu-
The tectonic activity during the Early Jurassic vial deltaic deposits, with some marine influx from
was probably fairly limited. The period as such was the south. During periods of the Middle Jurassic a
characterized by a transgressive sequence (Vail and marine connection might have existed to the north
Todd, 1981). The thickest Lower Jurassic series is along the main basin axis. Several of the wells near
165 m thick (17/9-1) representing a time span of the axis contain marginal marine to nearshore sed-
some 20 Ma (Fig. 2). iments in the Middle Jurassic series (i.e., 17/9-1,
The Early Jurassic is termed the Fjerritslev For- 17/12-1, 17/12-3 and 18/10-1).
mation and it is dominated by shales, indicative of (b) The Callovian deposits dominated by marginal
shallow marine deposits. In this study, we have in- marine deposits (sands). During the Middle Jurassic
cluded the sandy facies encountered in wells 17/ the SENS area experienced a dextral movement
12-1 and 17/12-3, representing near coast to deltaic along the Tornquist-Teisseyre Line (Fig. 8). Activity
deposits in this formation and not in the Gassum along this trend around 58N continued westward to
Formation. These sandy facies attributed to the Fjer- the southeastern edge of the Sele High. This dextral
ritslev Formation are known in the Danish sector. movement led to the reactivation of the normal fault
In Early Jurassic times the evolutions of the flanking the western side of the sta Graben and
Egersund Basin and sta Graben were slightly dif- Egersund Basin (the eastern boundary fault of the
ferent (Fig. 6). A fluvial system with drainage from Sele High) with the result that the sta Graben to
the Stavanger Platform probably had its front in the the north of approximately 58N shows a gentle tilt
geographical centre (north-south) of block 17/12. to the west at pre-Zechstein level. In the area where
A marine connection existed between the northern the Tornquist-Teisseyre Line links the N-S normal
part and the southern part of the SENS area. In the faults (Block 17/12), a thick post-Early Jurassic basin
northern part, Fjerritslev shales where deposited in developed as seen on the isopach map Top Triassic to
an anoxic environment indicating restricted bottom Top Middle Jurassic (Fig. 9).
water circulation (Fig. 7). Furthermore, the area to the west of the study
The sandy deposits in well 18/10-1 are indicative of area (Montrose-Forties High) was a centre of domal
a fluvial to deltaic environment. In well 9/2-1 to the uplift, where the relief might have reached 2000-
south of the threshold the Fjerritslev Formation was 3000 m (Ziegler, 1988). This uplifted area acted
deposited in a shallow marine shelf environment. as a sediment source also in a southeast direction.
The main basin axis was on the east side of Generally, the whole North Sea underwent severe
the Sele High, extending from 59N southwards to tectonic activity involving the up-doming with associ-
5810'N along a fairly narrow band between 320'E ated volcanic activity, normal faulting (perpendicular
and 4. The shaly Fjerritslev Formation can be to wrench movements), wrenching and subsidence
traced on seismic data into the Stord Basin. To the (Ziegler, 1988). To the south of 58N, in Quadrants
south (south of 58N) the main Early Jurassic Basin 9 and 10, increased tectonic activity can be observed.
axis has a slight shift in orientation from N-S (north In addition to the effect of wrench movement and
of 58N) to NNW-SSE (south of 58N). The main normal faulting, this southern part of the SENS area
axes are located in Blocks 9/2, 9/5, 9/6 and 9/9 was affected by halokinetic movements.
A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 29

59c00*

5800'

5700'

300' 400* 500* 600

Normal fault

Extension of Fjerritslev Fm

Oil accumulation

Dry well

m Oil well

Fig. 7. Distribution of the Early Jurassic, Fjerritslev Formation,

30 S. S0rensen, H. Morizot and S. Skottheim

3*00' 5W 6 ; 00'

59W

5800*

57 c 00 4

\ym'\ Areas with thin E- and M- Jurassic deposits (onlaps & toplaps)
1 I E- and M- Jurassic thin to med. sedimentary basins (conformable reflectors)
E- and M- Jurassic thick sedimentary basins
Onlap limit
~ Erosional limit f'^'w Structural High Areas
Faults i % Structural Low Areas

Fig. 8. Early and Middle Jurassic basin configuration.

A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 31

LEG! NO:

>400m

40(b >200m

A -200m

No M5
s Fault Sketch at
' Top Triassic Level

Fig. 9. Isopach map between Top Triassic and Top Middle Jurassic (metres).
32 S. S0rensen, H. Morizot and S. Skottheim

At the Middle Jurassic level the main fault orienta- Late Jurassic
tion between 5630'N and 58N is 130-150N with a The second major extensional event in the Jurassic
slight change towards 150-170N in the SW-corner occurred at the beginning of Late Jurassic (approx-
of the study area (Tkble 1). The third trend of less imately 160 Ma Oxfordian). This event is well
importance is the 10-30N. observed in the Central and Viking Grabens, but
In the area between 58N and 59N, the western seems to be of fairly minor significance in the SENS
boundary fault orientated 10-30N has the largest area, its effects being some reactivation with a minor
throw, whilst that in the eastern Egersund boundary vertical throw along the existing faults. According
fault orientated 10-30N is very limited. In the to Ziegler (1988), some of the eastern parts of the
northeastern corner of the study area the throw is Egersund Basin were affected by compressional fore-
almost zero. land deformation, but the main parts of the Horda-
The Middle Jurassic Bryne Formation is wide- sta half-grabens became inactive. To the west of
spread in the study area, as can be documented by the SENS area the Vestland Ridge system probably
the 27 wells involved in the study, of which only four formed an archipelago during Oxfordian (Fig. 12).
(17/11-1, 17/10-1, 17/11-2 and 8/11-1) did not pene- This archipelago might have been a local source for
trate Bryne Formation. Whether the formations were more coarse clastic sediments in the westernmost
eroded or not deposited in these wells is not clear. part of the area, which is partly reflected in the
All these wells have encountered older stratigraphic lithology of the formations and the kerogen compo-
horizons. But it appears as if it has been eroded, par- sition of the Oxfordian to Kimmeridgian deposits.
ticularly in areas on the Sele High (17/10-1, 17/11-1 Rocks of a more (deep) marine origin in the central
and 17/11-2). parts of the basin gradually become more marginal
The Bryne Formation was deposited in fluvial marine deposits to the west. In the northeastern
deltaic system, with basinal connections to the north corner of the basin (Blocks 18/2, 18/6 and 18/9) de-
(Stord Basin) and to the south (Norwegian Danish posits of Oxfordian age are fairly thick indicating a
Basin). Marine incursions are observed as far north fairly strong subsidence and perhaps significant flu-
as the northern limit of Quadrant 9 (Fig. 10). The vial drainage from the hinterland to the east. In the
deposits are dominantly sands, silts and shales, with northwestern corner the marine Egersund Forma-
some volcanics and coal beds. The volcanic material tion is found in the Ling Graben (17/4-1), indicating
is basically found in well 17/9-1, where the Middle a marine connection at least as far west as this. It
Jurassic series is more than 500 m thick. is, however, likely that the Ling Graben was only
In Callovian times a global transgression started an embayment connected to the southeast at that
(Vail and Todd, 1981), which also affected the time.
SENS area, represented by the Late Middle Juras- The old threshold, more or less orientated east to
sic Sandnes Formation, which is a marginal marine west at approximately 58N, was still in existence, and
sandstone of limited distribution and thickness. The the thickness of Oxfordian sediments are reduced in
Sandnes Formation ranges in age from Callovian this area. To the southeast of 58N the deposits
to Middle Oxfordian (Vollset and Doro, 1984). A decrease in thickness. In the western part, on the
marine connection to the north is believed to have southern part of Sele High, some local Oxfordian
existed frequently during Callovian times (Fig. 11). basins occur (Blocks 17/10 and 17/11).
In the study area only 50% of the wells encoun- During Late Jurassic Kimmeridgian to Volgian
tered the Sandnes Formation. This is probably due to times, the area had a more direct contact to the west
non-deposition of this formation on the structurally (Central Graben), with the eastern platform edge
elevated areas. Wells 17/9-1 and 18/11-1 are believed between 4 and 420 and between 58N and 59N
to have been located in a more basinal position, (Fig. 13). The isopach map between Base Cretaceous
hence no marginal marine sandstone is encountered and Middle Jurassic clearly shows the effect of the
in the wells. We prefer to assign these Callovian age active faulting at the Middle Jurassic level in terms
deposits to the Egersund Formation. of sediment thickness, as a result of faulting and
The Middle Jurassic sand deposits are the main subsidence (Figs. 9 and 14).
reservoir rocks in the Egersund Basin and the sta At the Intra Upper Jurassic level (Top Tku Forma-
Graben. Wells 9/2-1 and 17/12-2 both tested oil from tion) the number of faults affecting this horizon is
the Sandnes Formation, whilst wells 17/12-1 and 18/ reduced compared with the map of the Top Middle
10-1 tested oil from the Bryne Formation. Jurassic level. Again, the overall dominating trends
are 130-150N and 150-170N with a 35-55N
influence being the northwestern boundary of the
Sele High (T^ble 1). During Kimmeridgian times, the
A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 33

5900

5800'

5700*

3W 4*00' 6
six)

CZJ Marine influx ft\ Fans

^M Periodically flooded <> Dry well
Marginal marine Oil well
{Delta, Marsh, Flood-plain)
i 1 Continental
I 1 (Deltaic)
Fig. 10. Bathonian-Bajocian paleogeography.
34 S. S0rensen, H. Monzot and S. Skottheim

5900(

CTAt4M2L H

5800*

57

300' 4W 500 600'

Marine B Bryne Fm.

Periodically flooded E Egersund Fm
Marginal Marine S Sandnes Fm.
(Fluvial. Deltaic. Flood-plain)
<> Dry well
Continental
# Oil well
Fig. 11. Callovian paleogeography.
A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 35

5900*

S800

5700'

300 400* 500' 600*

j j Shallow marine
1
* Inner shelf
Marginal marine
<> Dry well
Oil well

Fig. 12. Oxfordian paleogeography.

 S. S0rensen, H. Morizot and S. Skottheim

300# 400* 5C00 600*

59"00'

Structural High Areas

Structural Low Areas
Faults
Fig. 13. Late Jurassic basin configuration.
A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 37

i Gf ND:

>400m

400m> I ' -I >200m

c200m

X Fault Sketch
at Top Tau Fm.
Level

Fig. 14. Isopach map between Top Middle Jurassic and Top Tau (metres).

sea level rose with inundation of the Sele High and was extensive, probably covering most of the area.
probably covered most of the Vestland Ridge. The The Kimmeridgian (Tku Formation) can be observed
existence of local islands, however, is likely (Figs. 13 as an onlapping sequence. At this time a marine
and 15). The Egersund Basin and the sta Graben connection existed to the north (Stord Basin) and
continued to subside, this being epitomized by the to the west (Viking Graben and Central Graben).
significant increase in sediment thicknesses. On the This Late Jurassic ocean was connected to the Tethys
Stavanger Platform, the Kimmeridgian transgression through the Polish-Danish trough, and not through
38 S. S0rensen, H. Monzot and S. Skottheim

5900'

5800'

5700*

300* 400* 500' 600*

Marginal marine
Shallow marine
Restricted
Deep marine
Restricted
O Dry well
Oil well

Fig. 15. Kimmeridgian paleogeography.

A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 39

the Bohemian seaway, which came to an end at this The Sauda Formation. The type well in the Sauda
time (Ziegler, 1988). Formation is 9/4-3. It ranges in age from Late Kim-
Due to differential subsidence in the SENS area meridgian to Top Ryazanian. It shows a downlap
during Kimmeridgian times, the area was separated character on the Tku Formation in the northeastern
into several sub-basins. This led to very limited part of the study area. It was deposited in an open
bottom water circulation and, hence, anoxic bottom marine, generally low energy basinal environment
water conditions were developed in the basins. On (Fig. 16).
the margins of the basin, water circulation was better
due to shallow marine longshore currents. The effect Cretaceous
being reduced preservation of the organic material
and a less favourable kerogen composition in terms Early Cretaceous
of source rock. During the Early Cretaceous, the marine trans-
During Volgian times, the sea level dropped some- gressive phase continued with a fairly large subsi-
what, once more exposing parts of the Vestland dence in the area. At the Top Early Cretaceous level,
Ridge. Tb the east, this is seen as a regression the fault trends can be separated south and north of
on the Stavanger Platform with the prograding 58N. The southern part being dominated by 130-
sequences of the Sauda Formation. These pro- 150N and 150-170N with a minor 10-30N com-
grading sequences indicate the westernmost exten- ponent (Table 1), whereas north of 58N, the main
sion of the Late Jurassic shelf (Fig. 16). To the trends are 10-30N and 35-55N.
west of this limit, at approximately 420 (north The Early Cretaceous deposits are lithostrati-
of 58N), the depositional environment was open graphically divided into the Flekkefjord Forma-
marine, well oxygenated, probably bounded west- tion, Valhall Formation and R0dby Formation. The
wards by some islands located along the Vestland Flekkefjord Formation of Ryazanian age being a part
Ridge. of the Late Jurassic mega-cycle.
At this level, it becomes obvious that the Egersund
Basin and sta Graben ceased to exist as separate The Flekkefjord and Valhall Formations. The
structural features (sensu stricto) and should be Flekkefjord Formation ranges in age from Late
regarded as a part of the overall Late Jurassic North Volgian through Ryazanian. It was deposited in a
Sea basin. marine, low energy, basinal environment.
In the area there is no clear-cut evidence of an Near the base of the Valanginian there was a
unconformity between the Late Jurassic and Early significant drop in sea level, followed by an over-
Cretaceous, and the Early Cretaceous series should all transgressive sequence ending in Mid Barremian
be regarded as a part of the transgressive mega- time. The lithostratigraphic unit representative of
sequence initiated in early Callovian times. this time span is the Valhall Formation. The Val-
The Late Jurassic deposits range in age from hall Formation reached extensive thicknesses, as ob-
Base Oxfordian through Volgian. Lithostratigraphi- served even on the former structural highs (wells
cally speaking, they consist of the Egersund, Tku and 17/11-1, 17/11-2 and 17/12-2). A local depocentre
Sauda Formations. has been postulated to have developed in Block
17/10 during this time. It has further been specu-
The Egersund Formation. The formation ranges lated that the Sele High became inverted (Jacobsen,
in age from Base Oxfordian to Base Volgian. It was 1982), although we find no evidence of inversion of
deposited in an open marine, generally low energy the Sele High during Valanginian through Barremian
basinal environment throughout most of the area. times. However, inversion along the eastern side of
It is encountered in most of the wells in the study the Egersund Basin and sta Graben probably was
area, except for the ones on the Sele High where the initiated during this time. During the deposition of
formation probably was not deposited (Fig. 12). the Valhall Formation the basin subsidence kept pace
with the rate of sedimentation.
the Tau Formation. The Tku Formation ranges in
age from Base Kimmeridgian to Middle Volgian. It The R0dby Formation. The R0dby Formation
was deposited in an anaerobic marine environment ranges in age from Barremian to Base Cenomanian.
with high organic productivity and restricted bottom The lower part was deposited in a marine envi-
water circulation. It also represents the maximum of ronment in an overall regressive sequence (Jacobsen,
the Upper Jurassic transgressive mega-cycle in the 1982). From Mid to Late Albian times, the area had a
Egersund Basin (Fig. 15). shift of depositional system from a general regressive
phase to a Late Albian-Cenomanian transgressive
40 S. S0rensen, H. Monzot and S. Skottheim

5900

5700*

3W 400' 600*

fH3 Marginal marine

Shallow marine
Shelf
I 1 Deep marine
Dry well
Oil well
Fig. 16. Volgian paleogeography.
A tectonostratigraphic analysis of the southeast Norwegian North Sea Basin 41

mega-cycle (Vail and Todd, 1981). The formation is in the area there is no proof of the existence of either
fairly thin and is eroded or not deposited on the of these types of deposit, but they might be present
flank of the basin. in the southern part of the basin along the old faults
During the transition from Early to Late Cre- and salt structures, and along the eastern Egersund
taceous, compression starts to have an effect in the Basin and sta Graben basin margin faults, which
study area, with inversion accomplished by salt defor- are inverted features formed during Late Cretaceous
mation (i.e., reactivation of the old salt structures). to Early Tertiary times.
The compression phase was affecting the area peri-
odically, until Early Eocene times. Tertiary
For the whole of the Tertiary period, the area was
Late Cretaceous technically associated to the North Sea basin which
At the onset of the Late Cretaceous, the acceler- experienced thermal subsidence. Tertiary sediments
ation of the global sea floor spreading rate induced are found all over the SENS area, but they gradually
a major regional transgression inundating the whole thin out towards the Fennoscandian shield.
North Sea area. The Late Cretaceous deposits can During the Paleocene, the eastern Egersund Basin
be divided geographically into two zones along a and sta Graben were uplifted (Jacobsen, 1982),
NE-SW line north of Quadrant 25. The southern leading to the condensed sequences of Lista and Sele
part being the chalk province to which the SENS Formations.
area belongs, and the northern part being the clastic The boundary between Paleocene and Eocene is
province. represented by the volcanic ash-rich Balder Forma-
The southern province is lithologically dominated tion. This formation is clay dominated with a high
by chalk, chalky limestone and limestone deposits. content of tuifaceous material. In the SENS area, the
The depositional environment for the carbonate formation is assigned to the lowermost Eocene.
province was one of warm sea with water depths During Eocene times, sands were shed into the
in the range of 300-500 m (S0rensen et al., 1986). southeastern part of the Egersund Basin. The sand
The chalk group has been divided into five for- deposits have been termed lithostratigraphically the
mations ranging in age from Cenomanian through Fiskebank Formation (Deegan and Scull, 1977). The
Danian. The chalk in the Egersund Basin ranges in Late Eocene and Early Oligocene deposits are dom-
thickness from 60 m to almost 1000 m. inantly shaly. However, in the Early Chattian time a
The oldest formation, the Hidra Formation, ranges significant drop in sea level occured with the devel-
in age from Cenomanian to Turonian. This formation opment of Chattian sands along the eastern margins
generally has a "high" clay content (with respect to of the SENS area. This drop in sea level could be as-
pure chalk) and as such is not regarded as a potential sociated with the uplift of the Fennoscandian Shield
reservoir rock. to the east. One might infer that this uplift has con-
The Blod0ks Formation (Isaksen and Tonstad, tinued to the present times and is associated either
1989) of Turonian age is a thin shaly formation. with the non-deposition or the erosion of some of
It seems that the drop in sea level in Turonian the Tertiary deposits. During the Late Oligocene and
times might have led to erosion of the Blod0ks Miocene times the subsidence continued with the
and Hidra Formation on the somewhat structurally deposition of significant thicknesses of shales. How-
elevated areas like the Sele High. Neither of the ever, all these Upper Tertiary series thin towards the
two formations are present in the wells in the area east.
(17/11-1, 17/11-2 and 17/12-2).
The Hod Formation, (Coniacian to Late Cam-
panian) was deposited in a calcareous province with Conclusions
fairly limited tectonic activity. The boundary between
Hod and Tor Formations is often seen as a hard- From Permian to Mesozoic times, the study area
ground. The boundary between Tor (Maastrichtian) can be divided into several structural provinces. Both
and Ekofisk (Danian) Formations is also seen as a the Sele High and the Stavanger Platform are old
hardground in places. Both the Tor and Ekofisk For- structural highs. The major structural basins can
mations are represented by both allochthonous and be divided in two: the Egersund Basin, which was
autochthonous chalks. The allochthonous chalks are heavily salt affected from Mid THassic times, and the
found locally, and represent deposition due either to sta Graben, which actually is a half-graben. At the
faulting activity along pre-existing faults, or to local end of Mesozoic the study area ceased to exist as
halokinetic movements. These allochthonous chalks an independent tectonic province and became part
are known to be the best reservoir chalks. However, of the North Sea as a whole. The basinal part of the
42 S. S0rensen, H. Monzot and S. Skottheim

study area comprises sediments ranging in age from Myhre, L., 1975. Lithology, Well 8/3-1. Norwegian Petroleum
at least Early Permian to recent. Directorate, NPD Paper No. 1, 20 pp.
Ormaasen, D.E., Hamar, G.P., Jacobsen, K.H. and Skarpnes, O.,
1980. Permo-Triassic correlations in the North Sea area north
Acknowledgements of the Central Highs. In: The Sedimentation of the North
Sea Reservoir Rocks. NPF Meet. Geilo, 11-14 May, 1980.
We would like to thank Elf Aquitaine Norge A/S Institute of Petroleum, Heyden & Son, London.
for granting permission to publish this paper. Fur- Rhys, G.H. (Editor), 1974. A proposed standard lithostrati-
graphic nomenclature for the Southern North Sea and an
ther, we would like to thank the drafting department outline structural nomenclature for the whole of the (UK)
for the figures and Karin Nielsen who kindly typed North Sea. A report of the joint Oil Industry-Institute of
the manuscript. In addition, we would like to thank Geological Nomenclature. Rep. Inst. Inst. Geol. Sei. 74/8: 14
our colleagues in the Exploration Division for fruitful pp.
discussions during this study. S0rensen, S., Jones, M., Hardman, R.F.P., Leutz, W.K. and
Schwarz, P.H., 1986. Reservoir characteristics of high and
low productivity chalks from the Central North Sea. In:
References Habitat of Hydrocarbons on the Norwegian Continental
Shelf. Norwegian Petroleum Society, Graham and Trotman,
Deegan, C.E. and Scull, B.J., 1977. A proposed standard lithos- London, pp. 91-110.
tratigraphic nomenclature for the Central and Northern North S0rensen, S. and Martinsen, B., 1987. A paleogeographic recon-
Sea. Institute for Geological Sciences, London, Rep. 77/25. struction of the Rotliegendes Deposits in the Northeastern
Faerseth, R.B., Macintyre, R.M. and Naterstad, J., 1975. Meso- North Sea. In: J. Brook and K. Glennie (Editors), Petroleum
zoic alkaline dykes in the Sunnhordland region, western Geology of North-West Europe. Graham and Trotman, Lon-
Norway: ages, geochemistry and regional significance. Lithos don, pp. 497-508.
9, 331-345. Vail, P.R. and Todd, R., 1981. Northern North Sea Jurassic
Isaksen, D. and Tonstad, K. (Editors), 1989. A revised Cre- Unconformity, chronostratigraphy and sea level from seismic
taceous and Tertiary lithostratigraphic nomenclature for the stratigraphy. In: L.V. Illing and G.D. Hobsen (Editors),
Norwegian North Sea. Norw. Pet. Dir., Bull., 5, 59 pp. Petroleum Geology of the Continental Shelf of North-West
Jacobsen, H.K., Hamar G.P., Ormaasen, D.E. and Skarpnes, Europe. Norwegian Petroleum Society, Graham and Trotman,
O., 1980. Triassic facies in the North Sea Reservoir Rocks. London, pp. 216-235.
Norwegian Petroleum Society, Article XV. Vollset, J. and Dore, A.G., 1984. A revised Triassic and Jurassic
Jacobsen, V.W., 1982. Sedimentation and paleogeography. In: lithostratigraphic nomenclature for the Norwegian North Sea.
R . C Olsen and I.F. Strass (Editors), 1982. The Norwegian Norw. Pet. Dir., Bull., 3.
Danish Basin. Norwegian Petroleum Directorate, NPD Paper Ziegler, P.A., 1988. Evolution of the Arctic-North Atlantic and
No. 31, pp. 54-66. the Western Tethys. Am. Assoc. Pet. Geol., Mem., 43, 197 pp.

S. S0RENSEN Elf Aquitaine Norge A IS, Stavanger, Norway

Present address: NorexA/S, Luragaarden, Standnes N-4300, Norway
H. MORIZOT Elf Aquitaine Norge AIS, Stavanger, Norway
S. SKOTTHEIM Elf Aquitaine Norge A IS, Stavanger, Norway
 Tectonic modelling of the northern North Sea using
program RIFT

Stephen Lippard and Guojiang Liu

A one-dimensional pure shear lithospheric extension model has been applied to modelling the subsidence of the northern
Viking Graben using a depth converted and backstripped regional seismic section along 61N. Early rifting events are poorly
constrained by the seismic data, but the mid- to late Jurassic rifting event is clearly recognised and gives rise to block faulting
and the regional Base Cretaceous unconformity. Post-rifting thermal subsidence following this event resulted in the deposition
of a largely unfaulted Cretaceous and Cenozoic sequence of up to 5 km thick. A single phase, uniform stretching model can
account for the subsidence in the graben if paleobathymetries of up to 650 m occurred in the Cretaceous and early Tertiary. The
factors for the mid- to late Jurassic rifting event in the graben centre are modelled as 1.7 to 1.8 assuming that prior to this
event the crust was already thinned by earlier rifting to about 25 km. On the margins of the rift, most notably on the northern
Horda Platform, the subsidence curves depart from those predicted by the uniform stretching model. In particular, initial
subsidences are too small to account for the total subsidence. In this case, a non-uniform stretching model, with /?mantie > crust,
can be applied. This suggests that the zone of mantle extension is distributed over a wider area than the crustal extension, which
is more restricted to the graben centre. A heterogeneous stretching model, with both laterally and vertically varying factors,
is suggested. By interpolating the top of the basement beneath the graben from published data, an attempt has been made to
model the effects of earlier rifting phases. Two scenarios; late Permian-early Iriassic and late Triassic rifting, have been tested.
In both cases, the early rift phase can be modelled by factors of 1.3-1.5. This gives a total value for all rifting phases of
2.7 in the graben centre, which is in agreement with the present-day crustal thickness. This differs substantially from published
results which seem to have underestimated the total factors in the area.

Introduction accounted for by crustal stretching. He proposed in-

stead a model of sub-crustal erosion with thinning of
The northern part of the North Sea between 60 the lower crust by a "thermally active mantle plume".
and 62N, encompassing the northern part of the Later work in the Viking and Central Grabens and
Viking Graben and the flanking lumpen Spur and other parts of the North Sea rift system showed, how-
Horda Platform areas, has been the subject of con- ever, that the McKenzie model of pure-shear crustal
siderable study since the discovery of major oil and extension can be successfully applied (Christie and
gas fields in the area in the 1970s. In early de- Sclater, 1980; Wood and Barton, 1983; Barton and
scriptions of the geology, the dominantly extensional Wood, 1984; Giltner, 1987; Hellinger et al., 1989).
nature of the structures was recognised (Blair, 1975; Beach (1985, 1986) and Beach et al. (1987) in-
Bowen, 1975; Halstead, 1975). The first complete terpreted a deep reflection seismic profile across
geological cross-section across the area was pre- the northern Viking Graben and suggested a sim-
sented by Ziegler (1982), who showed tilted fault ple shear detachment model (Wernicke, 1985) to
blocks containing Jurassic and Tftassic sediments on account for the asymmetry of the structures. The
the flanks of a central trough. The fault blocks are model implies that the upper and lower crustal ex-
buried beneath an infill of Cretaceous and Cenozoic tensions are decoupled by a major crustal shear
sediments up to 5 km thick in the centre of the basin. zone dipping eastwards from the western flank under
Ziegler (1982) suggested that, although the main rift the Viking Graben and passing through the Moho
phase occurred in the Jurassic, the presence of thick beneath the Horda Platform.
TYiassic sediments suggested an earlier rift phase. He Beach et al. (1987) presented a generalised sub-
also noted that the amount of extension that could sidence curve for the Viking Graben which showed
be measured on the faults is relatively small and three distinct periods of increased subsidence in the
suggested that not all of the crustal thinning could be Triassic, late Jurassic and late Cretaceous. These

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 43-54. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
44 S. Lippard and G. Liu

events were related to three rift phases with factors suggest that the deep seismic data lend most support
of 1.5, 1.8-2.2 and 1.1, respectively, giving a total to a pure-shear rather than simple-shear model of
factor for all three phases of over 3.0. Giltner basin development.
(1987) applied the McKenzie model to the North Both Hollinger and Klemperer (1988) and Fichler
Viking Graben using Ziegler's cross-sections as a and Hospers (1991) have studied the NSPD-84 data
basis. He obtained tectonic subsidences in the basin and, combined with gravity modelling, have made
centre that suggested a total factor of about 1.8 maps of depth to Moho in the Viking Graben. They
consisting of TViassic ( 1.5) and late Jurassic ( 1.2) show that the reflection Moho and the gravity Moho
rifting phases. Vially (1988) made a regional study of are more or less coincident except locally under the
subsidence in the northern North Sea. He also distin- centre of the graben. Fichler and Hospers (in press)
guished the Permo-TYiassic and late Jurassic rifting calculate factors of more than 3.0 in the deepest
phases, but suggested that they were offset from one part of the graben between 60 and 61 where the
another with the axis of earlier rifting lying to the crustal thickness, between the base of the sediments
east of the Viking Graben. Vially shows a maximum and the Moho, is interpreted to be only 10 km.
factor for the Permo-THassic rifting beneath the They assumed a pre-stretching crustal thickness of
Horda Platform of 1.4 and for the late Jurassic rifting 31km.
a maximum factor of about 1.5 in the centre of the
Viking Graben.
Seismic interpretation
Badley et al. (1988) calculated factors for the
mid-Jurassic to early Cretaceous rifting phase in the In the present study a detailed interpretation has
northern Viking Graben. They noted that the fac- been made of a regional seismic line NNST-84-08,
tors obtained for the post-rift subsidences (1.14- made available by Nopec a.s. The line trends E-W
1.49) were greater than those for the syn-rift subsi- just to the north of lat. 61 and extends from the
dences (1.03-1.15) and proposed a depth-dependent 0ygarden Fault in the east across the northern part
extension model to account for this. They also sug- of the Horda Platform ("Sogn spur") and the Viking
gested that the symmetry of the syn- and post-rift Graben. On the west flank of the graben it crosses
sequences is incompatible with the Wernicke-type part of the Gullfaks structure on the lumpen Spur
model proposed by Beach (1986). In an earlier pa- before crossing the Norwegian/U.K sector boundary
per, Badley et al. (1984) published part of a seismic near to the Statfjord field (Fig. 1). The seismic line
line from the eastern part of the Horda Platform continues further east in the U.K sector, but for
showing deep-lying tilted fault blocks which they the purpose of this study we have used only the
related to the earlier Permo-THassic rifting phase Norwegian part.
postulated by Ziegler (1982). Badley et al. (1988) The interpretation was tied to well data, notably
suggested that both rift phases were formed by more wells 34/10-8 and 35/8-2 which lie on the line. It
or less orthogonal east-west extension and that there was depth converted using interval velocity data
is no evidence of major strike-slip or compressional corrected by a time/depth correlation from the wells.
phases of deformation. The depth converted cross-section is shown on Fig. 2.
Klemperer (1988) interpreted the NSDP-84 deep The Base Cretaceous unconformity is the easily
seismic profiles in the northern North Sea, one of recognised reflector truncating tilted fault blocks be-
which (NSDP-84-1) crosses the north Viking Graben low and with onlapping Cretaceous reflections above.
at about 61N approximately coincident with the line Over the highs this event represents a considerable
interpreted by Beach (1986). Klemperer (1988) in- time gap, perhaps as much as 100 m.y., but, where
terpreted the reflection Moho on most of the profile the sequence is more complete in the deepest part of
and provided confirmatory evidence for crustal thin- the troughs, it is assigned an age of 131 Ma (Ma =
ning beneath the graben. Crustal thicknesses were million years before present, m.y. = million years).
estimated to vary from about 20 km beneath the The time scale of Haq et al. (1987) was used for
graben to about 30 km beneath the Horda Plat- dating.
form. In contrast to Gibbs' (1987) interpretation The Top Brent reflector (Top Bathonian 156 Ma)
of the same data, Klemperer (1988) found no evi- has been tied to the well data and interpreted along
dence for the presence of continuous crustal shear the whole section. There are numerous faults at this
zones, but emphasised the continuity of flat-lying level and the interpretation is most problematic in
lower crustal events. Dipping events occur beneath the deepest part of the graben. Deeper events (Top
the Moho in the upper mantle, but these appear to Statfjord, Top Lomvi) can be interpreted, but these
show a symmetrical distribution about the graben. are parallel to the Top Brent and the interpretation
Klemperer (1988) and White et al. (1989) indeed in the graben centre is even less confident. The Top
Tectonic modelling of the northern North Sea using program RIFT 45

4E
62E

84-08

61

20 km

Fig. 1. Location of the area in the northern North Sea. Seismic line NNST-84-08, major structural features and important oil and gas
fields are shown.

Brent reflector is taken to be the top of the pre- Fault juxtaposes basement rocks against the Quater-
rift sequence with respect to the late Jurassic rifting nary sediments and was active during Cenozoic time
phase (Badley et al. 1988). (Rundberg, 1989).
The post-rift sequence, containing several intra-
Cretaceous reflectors Intra Albian (100 Ma), Top Subsidence analysis
Cenomanian (92 Ma) and Top Turonian (89 Ma)
and the Base Tertiary (65 Ma), Top Palaeocene An important method of estimating crustal and
(54 Ma) and Intra Eocene (49 Ma) events, is tied to lithospheric extension is an analysis of basin sub-
the wells and fairly well constrained by the interpre- sidence history (van Hinte, 1978; Bessis, 1986). In
tation. order to do this, the subsidence must be corrected
The base and top of the Miocene Utsira Forma- for compaction due to the progressive loading of the
tion are dated as 25 Ma and 5 Ma, respectively, sediments. The isostatic effects of sediment loading
following Rundberg and Smalley (1989). The Base are removed by backstripping to give the "tectonic"
Quaternary (2 Ma) forms an erosion surface which subsidence due to water loading alone (Steckler and
truncates older Cenozoic strata towards the east so Watts, 1978). The subsidence history at any num-
that Quaternary sediments rest on Palaeocene strata ber of selected points along the seismic line can be
at the eastern end of the section. The 0ygarden obtained by decompaction and backstripping using
46 S. Lippard and G. Liu

34/10-6 35/8-2
w E r 3E 4E F
1 A _ 1 (4) A 1 (6)

5
25
A J Z T - 49

65

L
89
92
100

-j A/l56\

0 10 km

Fig. 2. Depth-converted geoseismic line NNST-84-08 with top basement interpolated below. Ages of reflectors in million years (Ma)
see Table 1 for details. Points 1-6 (above) and 1-4 (below) refer to modelling locations.

TABLE 1 of the paleotopography of the basin and eustatic sea

Parameters used in backstripping level variations. Since mid-Jurassic times eustatic sea
levels have generally been above those of the present
Horizon Age (Ma) Lithology C /o day, with maximum levels of 100-300? m reached in
Top Basement ss/sh 0.39 0.56 the mid-late Cretaceous (Haq et al., 1987). Paleo-
Top Brent 156 sh 0.51 0.62 bathymetry is usually given by paleontological data
Base Cretaceous 131 sh (1st) 0.46 0.60
Intra Albian 100 sh 0.51 0.62 or can be estimated from seismic stratigraphy (Vail
Top Cenomanian 92 sh 0.51 0.62 et al., 1977). Later in this paper we will attempt to
Top Turonian 89 sh 0.51 0.62 derive some information about the paleobathymetry
Base Tertiary 65 sh (tuff) 0.43 0.59 from the subsidence curves.
Top Paleocene 54 sh 0.51 0.62
Intra Eocene 49 sh 0.51 0.62 The backstripping method we have assumed to be
Top Oligocene 25 ss 0.31 0.52 local, or Airy isostasy. Barton and Wood (1984) and
Top Miocene 5 ss/sh 0.39 0.56 Hellinger et al. (1989) have addressed the problem
Base Quaternary 2 ss/sh 0.39 0.56
of local versus regional isostatic compensation by
Lithology: ss = sand; sh = shale/mudstone; 1st = marl/limestone. modelling gravity data and have concluded that the
C = compaction factor (10~ 5 c m - 1 ) ; f0 = surface porosity.
elastic thickness of the lithosphere in the North Sea
Sediment grain density = 2.7 g/cm 3 .
is small and that, as a first approximation, flexural
effects can be ignored.
the methods first applied to the North Sea basin by
Sclater and Christie (1980). Compaction coefficients Tectonic modelling
and initial porosities were applied according to the
lithologies of each unit using the end member values Program RIFT (Odling, this volume) is based on
of Sclater and Christie (1980) (Tfoble 1). a McKenzie (1978) model of pure shear extension
Past water depths or paleobathymetry are a part of of the lithosphere. Finite rate extension (Jarvis and
the basin's subsidence history and can be the result McKenzie, 1980) and multiple rift phases are in-
Tectonic modelling of the northern North Sea using program RIFT 47

corporated in the model by calculating the effects TABLE 2

of extension on any pre-existing initial temperature One-phase (mid to late Jurassic) rifting results
profile. Non-uniform extension (Royden and Keen,
1 2 3 4 5 6
1980; Hellinger and Sclater, 1983; Rowley and Sa-
hagian, 1986) can be included by dividing the litho- c 1.4 1.75 1.8 1.45 1.35 1.10
m 1.4 1.75 1.8 2.0 1.7 1.6
sphere into a number of layers each of which can be 27.5 25 25 27.5 27.5 30
C\
stretched by a different amount. Between each rifting cp 25.1 21.5 21.2 24.3 24.6 27.5
phase the crustal thickness is calculated byfillingthe 0.94 1.18 1.25 0.68 0.61 0.56
basin with sediment to a specified water depth. In 0.92 1.16 1.29 0.61 0.60 0.51
T* 1.74 2.45 2.56 1.96 1.64 1.52
this way it is possible to use the crustal thickness data 1.82 2.47 2.53 1.99 1.68 1.46
TtP
as a control on the validity of the factors obtained
1 to 6 are the modelled points on the seismic line (Fig. 2).
by subsidence analysis. c crustal factor; m = mantle factor; C\ = initial crustal
In the modelling of the northern Viking graben thickness; C p = predicted present crustal thickness (including
profile we have initially modelled the mid-Jurassic sediments); Tia and Tip = actual and predicted syn-rift tectonic
to present-day subsidence patterns as a single rifting subsidences; 7\ a and 7\ p = actual and predicted total tectonic
subsidences.
phase which began at 156 Ma, lasted 25 m.y. and
was followed by 131 m.y. of post-rift subsidence.
The effect of earlier rifting phases has been taken dence patterns with thin or absent syn-rift sequences
into account by varying the initial crustal thicknesses and are difficult to model. Two subsidence curves;
across the graben until a best fit was obtained with the decompacted sediment-loaded subsidence ("total
the present crustal thickness. In the second stage we subsidence") and the backstripped water-loaded sub-
attempted to model the earlier rifting phase more sidence ("tectonic subsidence"), were obtained for
accurately by testing two two-phase rifting models in each point (Fig. 4).
which the timing of the first rift phase differs. Theoretical tectonic subsidence curves derived
from the RIFT program were fitted to each of the
One-phase rifting model actual tectonic subsidence curves (Fig. 4). The mid-
The mid Jurassic to present-day (156-0 Ma) sub- late Jurassic rifting event is assumed to have lasted
sidence history was modelled for six selected points 25 m.y. and the tectonic subsidence for this period is
along the seismic line. The points were chosen to lie fitted as closely as possible to the theoretical syn-rift
in the deepest parts of the troughs or in the centres subsidence. The initial crustal thicknesses were var-
of tilted fault blocks in order to avoid the effects ied from 27.5 to 25 km according to their position
of footwall uplift or hanging wall subsidence due to in the graben (Table 2). The initial lithosphere thick-
fault block tilting (Fig. 3) or flexural effects (Mars- ness was fixed at 125 km following McKenzie (1978).
den et al, 1989). The intra-basin highs; for example, The total subsidence was obtained by modelling to
the two-well locations (Fig. 2), give anomalous subsi- a point in the mid-Cenozoic (about 30 Ma). This is
the point at which there was a change from deep
to shallow water conditions in the area (Rundberg
and Smalley, 1989). For points 1-3, the data fit a
uniform stretching model (/?crust = /?mantie) with
factors of 1.4 on the west flank and 1.75 and 1.8
in the graben centre. For points 4-6 further east
on the Horda Platform, the curves cannot be fitted
to a uniform stretching model mainly because the
syn-rift subsidences are too small compared with the
total subsidences. However, a non-uniform stretch-
ing model (/?mantle > crust) C a n b e U S e d W i t h nantle
ranging from 1.7 to 2.0 and /?crust from 1.45 to 1.1
(Table 2).
It should be noted that in all cases the actual
subsidence curves deviate substantially from the the-
oretical curves, most notably in the Cretaceous and
Fig. 3. (a) Tilted fault blocks showing modelling locations A-I. early Tertiary (130-30 Ma), where the actual curves
(b) The irregular factor profile (solid line) is obtained using
all modelling locations. The dashed line shows a smoothed lie above the theoretical ones. The discrepancies are
factor profile obtained using only the mid-points of the tilted shown by the shaded areas on Fig. 4. An expla-
fault blocks (B, D, H). nation for this is the extent of the water depths
 Depth (km) Depth (km) Depth (km)

-"n" ^
o -
%
1
-if ^
'i /
/'V /
yi /
/ / ^^
1? vl ' ?
5 ~
^" -
tf 7 /
i\' / /
ex
o
;->/ /
< <rV7 /
i7 /


M
1
Depth (km) Depth (km) Depth (km
1
1 ' ^
-
1
j/
/ J
3C j 55 )
CO
. ca ft ca ca
6 / i J i /
/
] li
f 1 /
// 'J
^ / // 9
// J ^^
/ / si ^^^
/ / s / ^ ^
li / / / ^^^ -' / ^ ^
5
/'y / ' V
* ^^
H / ' / // / s ^
/ * /V/
ex ~"
< // / / /
/ ^ //
1 /
& / / / / s J
y /
f/
y ^ "^
Y^
o

M
00
Tectonic modelling of the northern North Sea using program RIFT 49

TABLE 3 et al. (1989) have recognised late THassic rifting

Paleobathymetries from subsidence analysis events in the area. Lervik et al. (1989) mention that
in well 30/6-1 on the Horda Platform, THassic rests
1 2 3 4 5 6 directly on the basement, suggesting that the rifting
100 Ma 350 650 580 650 320 400 commenced during the THassic. They conclude from
89 Ma 450 500 600 560 - - a regional study that faulting occurred throughout
65 Ma 300 210 250 150 210 -
the THassic, but that details of the timing are as
1 to 6 are the modelled points on the seismic line (Fig. 2). 100 yet unknown. Because of these uncertainties, we
Ma, 89 Ma and 65 Ma refer to Near Top Albian, Top luronian
and Base Tertiary. The predicted paleobathymetries are given in
tested two models; Permo-THassic (260-235 Ma)
metres. and late TYiassic (225-200 Ma) rifting, followed by
mid-late Jurassic rifting. The Tftassic sediments were
assumed to be an equal mixture of sand and shale
during the Cretaceous and early Tertiary. This is in for the decompaction calculations. The theoretical
agreement with paleobathymetric evidence (Nelson subsidence curves for four modelled points (shown
and Lamy, 1987) for deep water conditions (as much on the lower section on Fig. 2) are shown in Fig. 5.
as 1000 m?) in this period in the Viking Graben. In each case it can be seen that the factor for
The onlap of the post-rift Cretaceous sequences onto the late THassic rifting has to be slightly larger
the irregular fault blocks is evidence of a paleoto- (0.05) than for Permo-THassic rifting to obtain the
pography of several hundred metres at this time, same amount of subsidence at the onset of the mid-
although the effects of compaction and differential late Jurassic event. Tkble 4 shows that for Permo-
subsidence must also be taken into account. The THassic rifting the factors range from 1.3 on
small amount of sediment deposited in the early Cre- the flanks to 1.45 in the graben centre, for late
taceous suggests that the basin was sediment-starved THassic rifting the corresponding figures are 1.35 and
at this time (Bertram and Milton, 1988). From the 1.5. It is interesting to note that the factors for
differences between the theoretical and actual subsi- the earlier rifting events are more even across the
dence curves, it is possible to make some estimates Viking Graben than for the mid-late Jurassic event,
of water depths (Table 3). These results suggest that although the largest values still occur in the graben
water depths reached a maximum of at least 650 m centre. This supports the symmetrical distribution
in the mid-Cretaceous (100 Ma). Note that no eu- of the early and late rifting phases as proposed by
static sea level correction (+100 to 300 m?) has been Ziegler (1982) and Badley et al. (1988). In order to
made. obtain the best fit for the subsidence data, it was
necessary to vary the initial crustal thickness from
28 km in the centre of the graben to 30-32 km
Two-phase rifting models
on the flanks (Table 4). This implies that the crust
The seismic data do not allow the Top Basement to
beneath the graben was already thinned prior to
be identified with any confidence. A depth to base-
ment map was produced by Hospers and Ediriweera
(1988) by analysis of well data and modelling of the
gravity and magnetic data. This has been used to TABLE 4
fill in the top of the basement for modelling the ef- Two-phase rift rifting models
fects of earlier rifting phases on the mid-late Jurassic
1 2 3 4
rifting phase.
The timing of the earlier rifting event in the (a) Permo-Triassic followed by mid-late Jurassic rifting
/5p_T 1.3 1.45 1.45 1.35
northern Viking Graben is uncertain. Ziegler (1982) Cj 30 28 30 32
proposed a phase of late Permian-early Triassic C 156 28.0 25.3 27.2 29.6
rifting, whereas Morton et al. (1987) and Gabrielsen Cp 25.7 22.6 24.8 27.6
(b) Late Triassic followed by mid-late Jurassic rifting
1.35 1.5 1.5 1.4
Q 30 28 30 32
Fig. 4. Subsidence curves for points 1-6 (upper section on Fig. C 156 26.9 24.3 26.1 28.5
2). Solid line, total decompacted subsidence (sediment-loaded); Cp 25.6 22.4 24.6 27.4
short-dash line, tectonic subsidence (water-loaded); long-dash
line, theoretical modelled tectonic subsidence from program 1 to 4 are the modelled points on the seismic line (Fig. 2). p.j =
RIFT (see Table 2 for details), factors given in the top right- factor for Permo-Triassic (260-235 Ma) rifting; /?LT = fac-
hand corner (c, crust and m, mantle for two-layer stretching). tor for late Triassic (225-200 Ma) rifting; Q = initial crustal
Shaded areas represent misfits between actual and theoretical thickness; Q56 = predicted crustal thickness at 156 Ma (prior
tectonic subsidence curves (see text and Table 3 for further to mid-late Jurassic rifting); Cp = predicted present day crustal
explanation). thickness (including sediments).
50 S. Lippard and G. Liu

Age (Ma) Age (Ma)

250 200
J . L

, = 1.30 2= 1.40 , = 1.45 2= 1.85

, = 1.35 2= 1.40 , = 1.50 2= 1.85
Age (Ma) Age (Ma)
150 100

c , = 30

, = 1.45 2= 1-45 - , = 1.35 2= 1.1

, = 1.50 2= 1.45 - , = 1.40 2= 1 1

Fig. 5. Theoretical tectonic subsidence curves modelled by program RIFT for points 1-4 (lower section on Fig. 2). Solid line, scenario 1
(Permo-Triassic rifting followed by mid-late Jurassic rifting). Dashed line, scenario 2 (late Triassic rifting followed by mid-late Jurassic
rifting). Q , initial crustal thickness; lt factor for first rift phase; 2, factor for second rift phase (see Table 4 for further details).

the first rifting event. This suggests earlier Devonian the Caledonian orogeny extended and thinned by
or, perhaps, Carboniferous rifting although there is gravitational collapse, a model that has been used
no evidence from the seismic data to support this. for the Devonian basins of western Norway (Norton,
It is possible that the overthickened crust following 1986; Seranne and Seguret, 1987).
Tectonic modelling of the northern North Sea using program IUFT 51

A Be(Ma>
150 100

Fig. 6. Theoretical basement heat flow modelled by program RIFT for points 1-4. Scenario 1, solid line; scenario 2, dashed line.

Figure 6 shows the theoretical basement heat flow Discussion

evolution predicted by the two-phase rifting models.
The effect of the late Tfriassic event is to increase the Single and two-phase rifting models have been ap-
heat flow during the mid-late Jurassic rifting event plied to the subsidence history of the northern Viking
more than in the case of the Permo-Triassic event. Graben. The results are summarised and compared
Note that, in both cases, the heat flow anomaly with the result of others in T&ble 5. It can be seen
caused by the rifting events has decayed to back- that the present results differ substantially from those
ground value by the present day. of Giltner (1987) and Vially (1988). Giltner (1987)
modelled two-phase rifting in much the same way
as we have done. He modelled a first TYiassic rift
phase lasting for 35 Ma from 248 Ma to 213 Ma
52 S. Lippard and G. Liu

TABLE 5 the flanks and 2.5-3.0 in the graben centre (based on

Comparison of factors obtained in the present work and a uniform initial crustal thickness of 31 km).
published ones We have shown that the mid-late Jurassic rifting
Tampen Spur Viking Graben Horda Platform event can be accounted for by a uniform rifting
(a) Giltner (1987) model in the centre of the graben, but that a non-
- 1-2 1.5 1.2 uniform model, with more stretching in the mantle
ft-K 1.2 1.2 <1.2 part of the lithosphere than the crust, can be ap-
/?T0T 1-4 1.8 <1.4 plied to the Horda Platform. The deep seismic data
(b) Vially (1988) (Gibbs, 1987; Klemperer, 1988) provide evidence for
/3p_T 1.15 1.2 1.4 detachment zones in the upper mantle beneath the
ft-K 1.2 1.5 1.2
/?TOT 1.4 1.8 1.7
area, which may indicate that the crust and mantle
are partly decoupled. Badley et al. (1988) also pro-
(c) Badley et al. (1988)
/?j_K 1.05/1.38a 1.15/1.49* 1.03/1.14 a
posed that two-layer extension may account for the
discrepancies between initial and total subsidences
(d) This work
0- 1.35 1.5 1.4 in the area. White and McKenzie (1988) proposed
ft-K 1-4 1.8 1.4/1.85 b that a non-uniform stretching model with mantle
/3 T OT 1.9 2.7 2.0/2.6 b extension extending over a wider area than crustal
/?P_T = /? factor Permo-Triassic rifting; /?J_K = /? factor Juras- extension could account for the onlap of post-rift
sic-Cretaceous rifting; = ?- x /?K-J sequences onto the rift margins. The model assumes
a
First value from initial syn-rift subsidence; second from total overall equal amounts of stretching in the crust and
subsidence.
b
First value for crustal extension, second for mantle extension. mantle so that the crustal stretching is greater than
the mantle in the graben centre. Another mechanism
by which the syn-rift subsidence can be reduced is
and a second Jurassic-Cretaceous phase from 169 lateral transfer of heat from the graben centre to
Ma to 98 Ma (although he admits that most of the the flanks (Cochran, 1983). This is not taken into ac-
tectonic activity had ceased by 130 Ma). The initial count by the McKenzie model which assumes vertical
crustal thickness was taken as 35 km. The difference heat transfer. In large extensional basins such as the
with our results can be largely accounted for by this North Sea vertical heat transfer seems, however, to
large initial crustal thickness which results in the be a reasonable assumption. The present results are
need for lower factors to obtain the same tec- thus similar to the model of White and McKenzie
tonic subsidence. However, it is difficult to see how (1988), but suggest more irregular lateral and verti-
Giltner arrived at lower factors for the Jurassic- cal variations in factors and perhaps conform most
Cretaceous event than for Triassic in the centre of closely to a heterogeneous stretching model of the
the graben. Vially (1988) presented maps of total type proposed by Coward (1986).
subsidence, tectonic subsidence and subsidence rates
for various time periods and maps of factor dis- Conclusions
tribution for Permo-THassic and "Cimmerian" rifting
phases. As noted previously, he showed the axis of Analysis of backstripped subsidence curves using
Permo-Triassic rifting underlying the Horda Platform a pure-shear model of lithospheric extension has
and offset from the later rifting event in the Viking shown that the amounts of extension for the mid-
Graben proper. This is supported by Lervik et al. to late Jurassic rifting event in the northern Viking
(1989) who stated that, far from the TMassic rifting Graben at 6 vary from 1.75-1.8 in the graben
being a precursor to the later rifting, the two may centre to 1.35-1.45 on the flanking Tampen Spur and
have little in common. Vially's total factor of about Horda Platform areas. Whereas uniform stretching
1.8 for the centre of the Viking Graben at 61N is (/?crust=/?mantie) fits the data in the graben centre,
the same as Giltner's, but the values given to the non-uniform stretching (/?crust < /?mantie) provides a
early and late phases are reversed. Vially gives no better fit on the flanks.
information about the parameters (i.e., initial crustal The non-conformity of the subsidence curves with
thickness, etc.) used in the modelling. the theoretical ones for the Cretaceous and early
Fichler-Fettig (1989) and Fichler and Hospers part of the Tertiary can be accounted for by relatively
(1991) show that the present-day crustal thickness, large paleobathymetries, probably as much as 500-
between top basement and the Moho, varies from 20 600 m in the mid-Cretaceous. This agrees with pub-
km on the Horda Platform to less than 14 km and lished biostratigraphic data and seismic stratigraphy.
probably close to 12 km thick in the graben centre at The effects of an earlier rifting event are taken
61N. Their factor map shows values of 1.5-2.0 on into account in the one-phase modelling by varying
Tectonic modelling of the northern North Sea using program RIFT 53

the initial crustal thickness using present-day crustal Extension Tectonics. Geol. Soc. London, Spec .Publ., 28:
thickness as a control. Two-phase rifting models have pp. 467-476.
Bertram, G.T. and Milton, N.J., 1988. Reconstructing basin
been carried out by reconstructing the depth to top evolution from sedimentary thickness: the importance of
of the basement beneath the rift using published paleobathymetric control, with reference to the North Sea.
data. Two scenarios; late Permian-early THassic and Basin Res., 1: 247-259.
late Tftassic rifting, were tested. It is found that late Bessis, F., 1986. Some remarks on the study of subsidence of
Tfriassic rifting requires slightly larger factors to sedimentary basins. Mar. Pet. Geol., 3: 37-63.
Blair, D.G., 1975. Structural styles in North Sea oil and gas fields.
obtain the required amount of subsidence prior to In: A.W. Woodland (Editor), Petroleum and the Continental
the onset of the mid-late Jurassic event and has a Shelf of North West Europe, Vol. 1. Geology. Applied Science
greater influence on the heat flow during the latter Publishers, London, pp. 327-338.
event. The basement heat flow anomaly associated Bowen, W, 1975. The Brent oil field. In: A.W. Woodland
with all rifting events should have decayed by the (Editor), Petroleum and the Continental Shelf of North West
Europe, Vol. 1. Geology. Applied Science Publishers, London,
present day. 353-361.
The factors for THassic rifting are fairly uni- Christie, P.A.E and Sclater, J.G., 1980. An extensional origin
form across the section suggesting a broad zone of for the Buchan and Witchground graben in the North Sea.
fairly even extension contrasting with the later mid- Nature, 283: 730-733.
late Jurassic event the effects of which are strongly Cochran, J.R., 1983. Effect of finite rifting times on the devel-
opment of a sedimentary basin. Earth Planet. Sei. Lett., 66:
concentrated in the Viking Graben. 289-302.
The effect of both early rifting phases is an in- Coward, M.P., 1986. Heterogeneous stretching, simple shear and
crease in the predicted basement heat flow during basin development. Earth Planet. Sei. Lett., 80: 325-336.
the mid-late Jurassic event, but this does not signifi- Fichler, C. and Hospers, J., 1991. Gravity modelling in the
cantly affect the factors calculated by the one-phase Viking Graben area, North Sea. In: D J . Blundell and A.D.
Gibbs (Editors), Tectonic Evolution of the North Sea Rifts.
model. Clarendon Press, Oxford, pp. 71-80
The total factor obtained from both rifting Fichler-Fettig, C , 1989. The North Sea Viking Graben from 3-D
events is about 2.5 for the centre of the graben. This gravity modelling and deep seismic profiling. Terra Abstracts,
is in better agreement with the present-day crustal 1: 36.
thickness than previously published results. Gabrielsen, R., Fasrseth, R.B., Idil, S. and Steel, R.J., 1989.
Architectural styles of basin fill in the North Viking Graben.
Terra Abstracts, 1: 34.
Gibbs, A.D., 1987. Deep seismic profiles in the northern North
Acknowledgement Sea. In: J. Brooks and K.W Glennie (Editors), Petroleum
Geology of North-West Europe. Graham and Trotman, Lon-
mdon, pp. 1025-1028.
We thank IKU and Nopec a.s for granting permis- Giltner, J.P., 1987. Application of extensional models to the
sion to publish this paper. Northern Viking Graben. Nor. Geol. Tidsskr., 67: 339-352.
Halstead, P.H., 1975. Northern North Sea faulting. In: K.G.
Finstad and R.C. Selley (Editors), Jurassic Northern North
Sea Symposium. Norw. Pet. Soc, NP5, pp. 1-38.
References Haq, B.U., Hardenbol, J. and Vail, PR., 1987. Chronology of
fluctuating sea levels since the Triassic (250 million years ago
to present). Science, 235: 1156-1166.
Badley, M.E., Egeberg, T. and Nipen, O., 1984. Development Hellinger, S.J. and Sclater, J.G., 1983. Some comments on two-
of rift basins illustrated by the structural evolution of the layer extensional models for the evolution of sedimentary
Oseberg feature, Block 30/6, offshore Norway. J. Geol. Soc.
basins. J. Geophys. Res., 88: 8251-8269.
London, 141 639-649.
Hellinger, S.J., Sclater, J.G. and Giltner, J., 1989. Mid-Jurassic
Badley, M.E., Price, J.D., Rambech Dahl, C. and Agdestein, T., through mid-Cretaceous extension in the Central Graben of
1988. The structural evolution of the northern Viking Graben the North Sea Part 1. Estimates from subsidence. Basin
and its bearing upon extensional models of basin formation. Res., 1: 191-200.
J. Geol. Soc. London, 145: 455-472. Hollinger, K. and Klemperer, S.L., 1988. A comparison of the
Barton, P. and Wood, R., 1984. Tectonic evolution of the North Moho interpreted from gravity data and from deep seismic
Sea basin; crustal stretching and subsidence. Geophys. J. R. reflection data in the northern North Sea. Geophys. J., 97:
Astron. Soc, 79: 987-1022. 247-250.
Beach, A., 1985. Some comments on sedimentary basin de- Hospers, J. and Ediriweera, K.K., 1988. Mapping of the top of
velopment in the northern North Sea. Scott. J. Geol., 21: the crystalline continental crust in the Viking Graben area,
493-512. North Sea. Nor. Geol. Unders., Spec. Publ., 3: 21-28.
Beach, A., 1986. A deep seismic reflection profile across the Jarvis, G.T. and McKenzie, D.P., 1980. Sedimentary basin forma-
northern North Sea. Nature, 323: 53-55. tion with finite extension rates. Earth Planet. Sei. Lett., 48:
Beach, A., Bird, T. and Gibbs, A.D., 1987. Extensional tec- 42-52.
tonics and crustal structure: deep seismic reflection data Klemperer, S.L., 1988. Crustal thinning and the nature of
from the northern North Sea Viking Graben. In: M.P. Cow- extension in the northern North Sea from deep seismic
ard, J.F. Dewey and P.L. Hancock (Editors), Continental reflection profiling. Tectonics, 7: 803-821.
54 S. Lippard and G. Liu

Lervik, K.S., Spencer, A.M. and Warrington, G., 1989. Outline an integrated approach. Dr.-Ing. Thesis, University of
of Triassic stratigraphy and structure in the central and Trondheim, NTH, 292 pp.
northern North Sea. In: J.D. Collinson (Editor), Correlation Rundberg, Y. and Smalley, P.C., 1989. High resolution dating of
in Hydrocarbon Exploration, Norwegian Petroleum Society, Cenozoic sediments from the Northern North Sea using
87
Graham and Trotman, London, pp. 173-189. Sr/ 86 Sr stratigraphy. Am. Assoc. Pet. Geol., Bull., 73:
Marsden, G., Kusznir, N., Roberts, A. and Yielding, G., 1989. 298-308.
Numerical modelling of basin formation in the northern Sclater, J.G. and Christie, P.A.F., 1980. Continental stretching:
North Sea. Terra Abstracts, 1: 34. an explanation of the post-mid-Cretaceous subsidence of the
McKenzie, D.P., 1978. Some remarks on the development of North Sea basin. J. Geophys. Res., 85: 3711-3739.
sedimentary basins. Earth Planet. Sei. Lett. 40: 25-32. Seranne, M. and Seguret, M., 1987. The Devonian basins of
Morton, N., Smith, R.M., Golden, M. and James, A.V., western Norway, tectonics and kinematics of an extending
1987. Comparative stratigraphic study of the Triassic-Jurassic crust. In: M.P. Coward, J.F. Dewey and P. Hancock (Editors),
sedimentation and basin evolution in the northern North Continental Extensional Tectonics. Geol. Soc. London, Spec.
Sea and northwest of the British Isles. In: J. Brooks Publ., 28: 537-548.
and K.W. Glennie (Editors), Petroleum Geology of North- Steckler, M.S. and Watts, A.B., 1978. Subsidence of the Atlantic-
West Europe. Graham and Trotman, London, pp. 697- type continental margin off New York. Earth Planet. Sei.
710. Lett., 41: 1-13.
Nelson, P.H.H. and Lamy, J.M., 1987. The M0re-West Shetlands Vail, PR., Mitchum, R.M. and Thompson, S., 1977. Seismic
area: a review. In: J. Brooks and K.W. Glennie (Editors), stratigraphy and global changes of sea level, Part 3. Relative
Petroleum Geology of North-West Europe. Graham and changes of sea level from coastal onlap. In: C.E. Payton
Trotman, London, pp. 775-784. (Editor), Seismic Stratigraphy Application to Hydrocarbon
Norton, M.G., 1986. Late Caledonide extension in Western Exploration. Am. Assoc. Pet. Geol., Mem. 26: 63-81.
Norway; a response to extreme crustal thickening. Tectonics, van Hinte, J.E., 1978. Geohistory analysis application of
5: 195-204. micropaleontology to exploration geology. Am. Assoc. Pet.
Odling, N.E., 1992. RIFT, a model of sedimentary basin evo- Geol., Bull., 62: 201-222.
lution by finite rate, non-uniform, pure shear extension of Vially, R., 1988. La subsidence dans le Viking Graben (mer du
the lithosphere. In: R.M. Larsen, H. Brekke, B.T. Larsen and nord septentrionale). Rev. Inst. Fr. Petrole, 43: 629-646.
E. Talleraas (Editors), Structural and Tectonic Modelling and Wernicke, B., 1985. Uniform sense simple shear of the continen-
its Application to Petroleum Geology. Norwegian Petroleum tal lithosphere. Can. J. Earth Sei., 22: 108-125.
Society (NPF), Special Publication 1. Elsevier, Amsterdam, White, N. and McKenzie, D., 1988. Formation of the "steer's
pp. 457-467 (this volume). head" geometry of sedimentary basins by differential stretch-
Rowley, D.B. and Sahagian, D., 1986. Depth-dependent stretch- ing of the crust and mantle. Geology, 16: 250-253.
ing: a different approach. Geology, 14: 32-35. White, N. and the BIRPS Group, 1989. Nature and magnitude
Royden, L. and Keen, C.E., 1980. Rifting processes and thermal of extension in the North Sea from regional seismic reflection
evolution of the continental margin of Eastern Canada deter- profiles. Terra Abstracts, 1: 33.
mined from subsidence curves. Earth Planet. Sei. Lett., 51: Wood, R. and Barton, P., 1983. Crustal thinning and subsidence
343-361. in the North Sea. Nature, 302: 134-136.
Rundberg, Y., 1989. Tertiary sedimentary history and basin Ziegler, P.A., 1982. Geological Atlas of Western and Central
evolution of the Norwegian North Sea between 60-62N Europe. Elsevier, Amsterdam, 130 pp.

S. LIPPARD Continental Shelf and Petroleum Technology Research (IKU), N-7034 Trondheim, Norway
Present address: Department of Geology and Mineral Resources Engineering, Norwegian Institute of Technology (NTH), N-
7034 Trondheim, Norway
G. LIU Continental Shelf and Petroleum Technology Research (IKU), N-7034 Trondheim, Norway
 55

Voring Basin: subsidence and tectonic evolution

J. Skogseid, T. Pedersen and V.B. Larsen

The post-Caledonian history of the V0ring Basin has been examined by tectonic subsidence and a uniform lithosphere
extension model. The study is based on depth converted seismic reflection lines integrated with seismic velocities and well data.
Tectonic subsidence calculated for the pre-Cretaceous, Cretaceous and Cenozoic time intervals shows that the main episode of
late Jurassic-early Cretaceous extension was centred in the eastern V0ring Basin, but affected the entire basin province, whereas
the early Cenozoic rifting, associated with high magmatic activity and continental separation, was restricted to the central and
western parts of the basin. NE-SW trending zones of weakness, probably inherited from the Caledonian crustal configuration,
are inferred throughout the evolution of the basin. In some regions high extension rates during the early Cenozoic caused
only minor faulting. This is attributed to ductile deformation/mobilization of Cretaceous shales and to intrusions. Although the
uniform extension modelling approach has relevance in a regional setting, we suggest that differential crustal thinning, lateral
heatflow and magmatic underplating have to be considered when analysing selected local basin features. The importance of
previous tectonic episodes is demonstrated in the Bod0 High region. Here the pre-Mesozoic structural configuration has a large
rotated basement block and a deep detachment plane which controls the later pattern of crustal deformation and subsidence.
Source rocks for petroleum hydrocarbons include the Kimmeridgian Clay formation and possibly early Cretaceous and Paleocene
sediments believed to have been deposited under conditions of restricted water circulation in the western V0ring Basin. A
potential reservoir unit may occur in redeposited sediments derived from a 250 km wide Paleocene-earliest Eocene landmass
along the line of continental separation.

Introduction mic lines have established a stratigraphic framework.

Finally, the first-order configuration of the deeper
Scientific and commercial investigations of the parts of the basin and the crust has been obtained
V0ring continental margin off mid-Norway (Fig. 1) by digital sonobuoy and expanded spread profiling
document a post-Caledonian history of progres- (ESP) experiments (Mutter et al., 1984; Eldholm and
sive, predominantly extensional tectonics leading Mutter, 1986; Planke et al., 1991).
towards continental breakup and formation of the Based on these data this study investigates the
Norwegian-Greenland Sea at about 57 Ma. The re- history of vertical movements at the margin focusing
gional sedimentary V0ring Basin, which occupies on the V0ring Basin. In particular, we attempt to
the central part of the margin, provides important resolve and quantify the different tectonic episodes
information about the pre-Cenozoic intracontinen- by means of their tectonic subsidence. Furthermore,
tal basin evolution between Norway and Greenland, we discuss the basin evolution and aspects related to
the sequence of events during breakup and the the temperature history of the sediments and their
post-opening development of the passive continental hydrocarbon potential.
margin.
Although the margin has been explored for two Geological setting
decades, it has only recently been possible to per-
form quantitative modelling of the margin in space The V0ring Margin comprises three main geolog-
and time. This is largely due to the acquisition of a ical provinces: the TY0ndelag Platform, the V0ring
database of multichannel seismic (MCS) lines cov- Basin and the V0ring Marginal High (Fig. 2). The
ering the entire margin (B0en et al., 1984; Skogseid area's main structures and structural units, as defined
and Eldholm, 1989). In addition, borehole data from at the base Cretaceous level, are described and given
commercial drilling on the continental shelf (Spencer formal names by Gabrielsen et al. (1984) and Blystad
et al., 1984; 1986; Dalland and Worsley, 1988) and et al. (in press).
scientific drilling at the V0ring Plateau (TMwani et The Tr0ndelag Platform is a 150 km wide area
al., 1976; Eldholm et al., 1987; 1989) tied to the seis- between the Norwegian mainland and the V0ring

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 55-82. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
56 J. Skogseid, T. Pedersen and VB. Larsen

V>u

Fig. 1. Regional physiography and structural features of the ocean basins and continental margins in the Norwegian-Greenland Sea
(Eidholm et al., 1990). The box shows the study area.

Basin. Although representing a relatively stable area northernmost Fulla Ridge (Planke et al., 1991).
during late Mesozoic times, it was previously divided The V0ring Basin is terminated to the west by
into several structural units. the V0ring Escarpment: the eastern boundary of the
The V0ring Basin is a large sedimentary basin V0ring Marginal High. The formation of this high,
province with grabens, sub-basins and structural during early Cenozoic breakup, was associated with
highs. The eastern basin province includes the deep massive extrusive and intrusive magmatic activity. A
Ttena Basin and Ras Subbasin, flanked by the Hal- sill/flow complex in a 10-40 km wide area east of
ten and D0nna terraces adjacent to the elevated the V0ring Escarpment inhibits seismic resolution to
Nordland Ridge in the east, and the Fulla Ridge in deeper levels (Fig. 2;flowlimit basalts).
the west. The Fulla Ridge forms a prominent ridge The seismic resolution in the V0ring Basin deteri-
across the entire V0ring Basin separating the eastern orates below the base Cretaceous horizon, although
basin province from the Fenris and Hel Grabens a series of deeper reflectors are observed in most
west of the ridge. The northern part of the Fulla of the study area. A minimum of 3 km of pre-
Ridge includes a synform, the Nyk Syncline. In ad- Cretaceous sediments is inferred regionally (B0en
dition to the base Cretaceous structures, separate et al., 1984; Skogseid and Eldholm, 1989), while
structures also exist at deeper stratigraphical levels, as much as 10 km may exist locally (Eldholm and
exemplified by the Bod0 High, recognized by a major Mutter, 1986; Planke et al., 1991). Thus, the pre-
gravity anomaly between the Traena Basin and the Cretaceous sediment thicknesses in the V0ring Basin
V0ring Basin: subsidence and tectonic evolution 57

MAIN STRUCTURAL FEATURE llllllllll CONTINENT-OCEAN BOUNDARY DSDP/ODP DRILL SITE

BATHYMETRY (M) FLOW LIMIT (BASALTS) GRAVITY HIGH ( > 50 mGal)

Fig. 2. Main structural features in the V0ring Basin and adjacent areas based on Skogseid and Eldholm (1989) and Blystad et al. (in
press). F-SE, Faeroe-Shetland Escarpment; COB, Continent-ocean boundary; NBFZ: Norway Basin fracture zone; CJMFZ, central Jan
Mayen fracture zone; EJMFZ, East Jan Mayen fracture zone; LFZ, Lofoten fracture zone; VFZ: V0ring fracture zone.

are comparable to those underlying the Tr0ndelag tablished during a late Jurassic-early Cretaceous
Platform as well as the East Greenland sedimentary extensional episode and subsequent Cretaceous sub-
sequences. These regions document several Paleo- sidence. This episode represented a late stage in a
zoic and Mesozoic tectonic episodes prior to the late tectonic evolution dominated by extension follow-
Jurassic (Bukovics et al., 1984; Surlyk et al., 1984; ing the end of the Caledonian Orogeny. Seismic
Larsen, 1987; Ziegler, 1988; Graham and Robin- information recording the pre-Cretaceous events ex-
son, 1989; Larsen et al., 1989; Manby and Hambrey, ists mainly from the TY0ndelag Platform, whereas
1989). onshore geology documents large-scale extensional
The present structural division between the deformation dating back about 400 Ma. Based on
TY0ndelag Platform and the V0ring Basin was es- these data, information from gravity anomalies
58 J. Skogseid, T. Pedersen and VB. Larsen

LAURENTIAN MARGIN BALTOSCANDIAN MARGIN

CARBONIFEROUS - MIDDLE JURASSIC

General extension
AREA UNDER EXTENSION

LATE JURASSIC - EARLY CRETACEOUS

General extension

\c( Voring Bodo High Trondelag Platform

'Escarpment
coastline

EARLY TERTIARY
Extension - Continental separation

ORDOVICIAN SEDIMENTS ISLAND ARC VOLCANICS

DEVONIAN SEDIMENTS
M OCEANIC CRUST - OPHIOLITE NOT TO SCALE

Fig. 3. Schematic diagrams illustrating an evolutionary model of the margin off mid-Norway from late Ordovician to early Tertiary
times.
V0ring Basin: subsidence and tectonic evolution 59

(Tklwani and Eldholm, 1972; Planke et al., 1991) Cretaceous times when the entire TY0ndelag Plat-
and the subsidence modelling of this study, we sum- form and Halten Terrace were tilted westward caus-
marize the tectonic evolution of this region since late ing a contemporaneous transgression. The base Cre-
Ordovician times (Fig. 3, stages 1-6): taceous reflector is faulted along the Fulla Ridge and
(1) Continental convergence between the Lauren- in the vicinity of the Fenris and Hei Grabens, reveal-
tian and Baltoscandian margins, referred to as the ing early Cretaceous subsidence apparently without
Caledonian Orogeny, has been suggested to have erosion. Consequently the base Cretaceous reflector
occurred in two main phases; the Finnmarkian, may define a diachron, implying that the central and
lasting until late Ordovician (stage 1), and the western parts of the V0ring Basin developed in a
Scandian that caused the final closure of the Ia- later stage of the late Jurassic-early Cretaceous tec-
petus Ocean during the late Silurian-early Devonian tonic episode or, more likely, experienced a more
(stage 2) (Dallmeyer, 1988). Obducted island arcs rapid subsidence.
along the orogen suggest a phase of ocean-ocean (5) In the early Cenozoic, renewed extension cen-
collision which may argue for the existence of an tred in the western V0ring Basin led to breakup and
early subduction zone to the west. The Scandian continental separation between Norway and Green-
phase caused reactivation of former thrust zones, land in the earliest Eocene (stage 6). Associated
obduction of oceanic crust and mantle materials in with continental breakup, the region experienced
front of the Laurentian continental crust, and finally syn-tectonic listric faulting, uplift and erosion, and
continent-continent collision with large-scale litho- large-scale magmatic activity and volcanism (Eld-
spheric thickening (Austrheim, 1987; Jamtveit, 1987; holm et al., 1989; Skogseid and Eldholm, 1989).
Bucher-Nurminen, 1991). Post-tectonic Cenozoic subsidence was restricted to
(2) The thick isostatically unbalanced lithosphere the 150-200 km wide region of the V0ring Basin east
responded by gravitational collapse and extensional of the V0ring Escarpment.
crustal thinning (Dewey, 1988; England and House- (6) During Neogene time the Norwegian mainland
man, 1988) during the Devonian, mainly by reactiva- and adjacent regions were uplifted and eroded. This
tion of the original thrusts by normal slip (stage 3). process produced huge amounts of Plio-Pleistocene
This process is documented onshore where Devonian sediments which were transported westward and de-
half-grabens formed and filled with deposits of Old posited as the shelf edge prograded to its present
Red sandstone (Surlyk et al., 1984; Hossack, 1986; position.
McClay et al., 1986; Norton, 1986; Ziegler, 1988).
Similar offshore structures are described by Gowers Tectonic subsidence
and Lunde (1984), Duindam and Van Hoorn (1987),
Nelson and Lamy (1987) and Graham and Robinson Mode/ling procedure
(1989). Vertical movements, corrected for sediment com-
(3) Between Devonian and late Jurassic times sev- paction and loading as well as paleo water depth,
eral extensional episodes occurred (stage 4): late provide important information for understanding the
Carboniferous-early Permian regional rifting and tectonic evolution of sedimentary basins. To quantify
basement-involved block faulting, mid-to-late THas- this tectonic subsidence (negative values represent-
sic basement-involved block faulting and, finally, late ing uplift), we use the "backstripping" procedure
THassic-early Jurassic growth faulting with detach- (Steckler and Watts, 1978; Sclater and Christie,
ment in TYiassic evaporites (Surlyk et al., 1984; B.T 1980). Here, the one-dimensional tectonic subsi-
Larsen, pers. comm.). dence, Y, is:
(4) Late Jurassic-early Cretaceous extension led
to major fault activity including reactivation of older = Hspm ~ Ps + Wd (1)
fault zones which generally created slightly rotated
fault blocks and caused subsequent subsidence along where Hs is the decompacted sediment thickness,
major rift systems (stage 5) (R0nnevik et al., 1983; W& is the paleo water depth, and p m , ps and pw
B0en et al., 1984; Hinz et al., 1984; Mutter, 1984; are the densities of mantle, sediments and water,
Skogseid and Eldholm, 1989). Along the Nordland respectively. Equation 1 assumes an Airy compensa-
Ridge, the Halten Terrace and the eastern flank tion mechanism and eustatic sea level changes are
of the V0ring Basin, the base Cretaceous reflec- ignored.
tor is an erosional unconformity which truncates From eqn. 1 and observational data the "observed"
late Jurassic fault blocks (Skogseid and Eldholm, tectonic subsidence can be calculated as a function
1989). Brekke and Riis (1987) found that these of time for a given point within the basin. Further-
regions remained relatively elevated until middle more, it is possible to derive first-order estimates of
60 J. Skogseid, T. Pedersen and VB. Larsen

lithosphere evolution as a function of time and space TABLE 1

by comparing the "observed" tectonic subsidence at Lithospheric parameters used in this study
different locations with synthetic subsidence curves
C\ 3.6 km Hot mantle geoid
from geodynamic models. We use the McKenzie
Ci = 7.8 km Cold mantle geoid
(1978) model with instantaneous and uniform litho- L = 125 km Lithosphere thickness
sphere extension, as formulated by Le Pichon and T = 65 Ma Thermal time constant
Sibuet (1981). pm = 3.3 g c m - 3 Mantle density
The initial fault controlled subsidence, Sj, is: pw = 1.03 g c m - 3 Water density

Si = ( C ! - W ) ( l - i ) (2) calculated at about 750 locations forming a 15 x 15

km grid in the V0ring Basin. The grid allows us
where C\ is the "hot" mantle geoid depth, W is the to present both subsidence profiles and maps. Five
pre-rift water depth (after sediment unloading) and stratigraphic horizons, with ages of 0, 65, 144, 200
is the lithosphere stretching factor. The lithosphere and 380 Ma, were used in the calculations. Other
returns to thermal equilibrium an infinitely long parameters are shown in Table 1. There are two
time after extension with a total tectonic subsidence, reasons for incorporating the 200 and 380 Ma hori-
Stot : zons in the modelling in spite of their relatively
poor definition. Neglecting sediments below base
SXol = (C2-W)(l-^ (3) Cretaceous may lead to: (1) overestimation of the
tectonic subsidence, or "pseudosubsidence" (Roll,
1974), due to compaction of the pre-Cretaceous sed-
where C2 is the "cold" mantle geoid depth. The iments (Fig. 4), and (2) underestimation of tectonic
thermal subsidence, S ^ , approaches its final value subsidence, due to an erroneous pre-rift water depth
with a time constant, r: (sediment unloaded basement depth) in the expres-
sion for both initial and total tectonic subsidence
Sthe = (Stot-Si)(l-e-' / T ) (4) (eqns. 2 and 3) (Fig. 4). These effects will generally
not cancel one another out despite their opposite
where t is time. signs.
Note that the "hot" mantle geoid depth is equiv- The modelling recognizes the two well-docu-
alent to the depth of a mid-ocean ridge without mented late Jurassic-early Cretaceous and early
an oceanic crust, and the "cold" value is the corre- Cenozoic rift episodes. We incorporated the
sponding depth after cooling of the asthenospheric Jurassic-Cretaceous rifting in the entire V0ring
anomaly (Turcotte et al., 1977; Celerier, 1986). Stot is Basin, whereas the early Cenozoic episode only af-
the theoretical equivalent of the observational "total fected the western and central V0ring Basin (Ped-
tectonic subsidence", or TTS, of Sawyer (1985). ersen and Skogseid, 1989; Skogseid and Eldholm,
The rationale for using this approach is threefold. 1989).
First, extension is documented to play an important
role in the evolution of the V0ring Basin. Second, Data
the above extension model contains a minimum of Our study is based mainly on the seismic inter-
parameters. These are: the "hot" and "cold" man- pretation of Skogseid and Eldholm (1987; 1989)
tle geoids, the lithospheric thermal time constant (Fig. 5) which outlines the lateral distribution of
and stretching factor. This is important as it mini- sedimentary units and their structural setting. How-
mizes the number of necessary a priori known crust ever, these data resolve only the "shallow" structure
and mantle lithosphere properties. Third, the tem- of the crust, generally down to the base Creta-
poral resolution of subsidence in the V0ring Basin ceous level. To evaluate better the deeper basin
at the present time does not support the use of and the crustal configuration we have integrated
more complex extension models, in particular since velocity-depth curves interpreted from seismic re-
it is difficult to separate syn- and post-tectonic subsi- fraction data (Mutter et al., 1984; Eldholm and
dence components. We do believe, however, that this Mutter, 1986; Planke et al., 1991). Assuming a 6.0-
approach allows us to make first-order estimates of 6.5 km/s velocity for the crystalline basement and
vertical movements and lithosphere extension during using the SEASAT free-air gravity field as a guide-
the V0ring Basin history. line, we have inferred a 380 Ma basement level in
To quantify the spatial (two-dimensional) basin the basin. The velocity-depth information has also
evolution, the tectonic subsidence and factors were been used to determine whether seismic reflection
V0ring Basin: subsidence and tectonic evolution 61

TIME (Ma)
400 350 300 250 200 150 100 50

13 POST CRET. TEC. SUB.

h-
CL
LU APPARENT SUBSIDENCE: 2.1 KM
ACTUAL SUBSIDENCE: 1.0 KM

) I I I I I

400 350 300 250 200 150 100 50

TIME(Ma)
0 100 200
I
_) | | -o

kx/ /
1 BETA =1.5
- 1

E 2 -2

, 7 BETA =: 1.5
0- 3
LU
Q
W "3

J -4

APPARENT BETA SECOND PHASE: 1.26

1 -5
200 400

Fig. 4. Top: An example of pseudosubsidence, the difference between apparent and actual subsidence due to compaction of underlying
sediments. The apparent tectonic subsidence of 2.1 km (upper curve) is reduced here to 1.0 km when compaction of pre-Cretaceous
sediments is accounted for (lower curve). Bottom: A two-phase uniform extension model with values of 1.5 is erroneously interpreted
as a two phase model with values of 1.5 and 1.26 if the pre-rift sediments unloaded water depth of 2500 m at 200 Ma is ignored and
a zero depth is assumed.

interpretations are consistent within each subregion Lithology and paleo water depth data are nec-
(Tkble 2) and to depth convert the seismic reflection essary in order to calculate tectonic subsidence.
profiles. To demonstrate regional trends and anoma- Lithologies were obtained from commercial wells
lies in the crust we have constructed crustal profiles west of the Nordland Ridge and on the Halten
across the margin based on seismic reflection, refrac- Terrace, and from the DSDP and ODP wells on
tion and gravity data. The profiles provide a tool to the V0ring Plateau (Fig. 2). Extrapolation into the
compare deep velocity anomalies to shallow struc- V0ring Basin was guided by general relationships be-
tures, and to correlate these with the gravity field tween lithology, seismic reflectivity and velocity. The
(Fig. 6). sediments were separated into five main sequences,
62 J. Skogseid, T. Pedersen and VB. Larsen

MAIN STRUCTURAL FEATURE llllllllll CONTINENT-OCEAN BOUNDARY O DSDP/ODP DRILL SITE

J
BATHYMETRY (M) 'L FLOW LIMIT (BASALTS)

Fig. 5. MCS profiles in the V0ring Basin and V0ring Marginal High. The location of profiles presented in this paper are showindicated
by bold lines with figure number annotated. For abbreviations see Fig. 2.

to which we ascribed three different lithologies (Ta- Basin (Fig. 7). The total tectonic subsidence post-
ble 3). Paleo water depth is difficult to determine, dating the base Cretaceous level increases westward.
both from well and seismic data. However, because There is also a general increase in Cenozoic tectonic
the Jurassic-Cretaceous and early Cenozoic tectonic subsidence from east to west, reflecting increased
episodes are bounded by erosional unconformities, tectonic activity as the zone of final continental
we have assumed an averaged shallow water depth at separation between Greenland and Norway is ap-
these times. proached (Skogseid and Eldholm, 1987). Moreover,
initial uplift is observed both in the Fenris Graben
Results and at the western Fulla Ridge. The best-fitting fac-
The subsidence modeling is exemplified by a pro- tors, estimated from the tectonic subsidence curves
file across the main structural elements in the V0ring after correction for remaining thermal anomaly from
 I 3"

R
O

a.

R-

Fig. 6. V0ring Margin crustal profiles including isovelocity contours, the faulted base Cretaceous horizon (bold lines), and other prominent deep reflectors (thin lines). Zone A = normal oceanic
crust; zone B = sub-aerially formed oceanic and transitional crust; zone C = V0ring Basin; and zone D = Tr0ndelag Platform region. CL = coast line; BH = Bod0 High; FR = Fulla Ridge; CO
HG = Hel Graben; NR = Nordland Ridge; NS = Nyk Syncline; TB = Traena Basin; TP = Tr0ndelag Platform; VMH = V0ring Marginal High; VE = V0ring Escarpment. From Planke et al.
(1991). Location in Fig. 5 (upper profile is northern profile).
64 J. Skogseid, T. Pedersen and VB. Larsen

TABLE 2

Mean seismic refraction velocities (km/s) at identified seismic horizon levels

Horizon VB Sub-province Wells

1 2 3 4 5 6
27 6 5 8 3 4
O' 2.3 *2.3 *2.3 2.3 2.2 2.0 2.0 2.5
44 3 7 n 13 4 4
A 2.3 2.4 2.4 2.4 2.1 2.2 2.2
28 2 7 13 4
A' 2.6 2.5 2.6 2.5 4.9 2.2
32 4 5 5 3
TP 2.8 !2.5 2.5 "2.7 3.4 2.5 2.2
c 44
7
3.0 !2.4 3
2
2.5 7
5
2.4 n
3.1 10
3.4 7
2.6 4
4
2.4
D' 4.1 4.0 3.9 3.0
31 x 3 4 n 5 3 3
D 4.3 2.8 4.0 4.5 4.4 4.3 3.7 3.2
12 4 6 3
E 5.1 *5.5 4.9 5.2 *4.7 4.1
4 2
BM 5.9 *5.3 6.5 !5.1
4 x 2
LCR 8.0 8.0 8.1 ^.O

Abbreviations: VB: V0ring Basin, 1: Nordland Ridge and Tr0ndelag Platform, 2: Halten and D0nna terraces, 3: Traena Basin and Ras
Subbasin, 4: Fulla Ridge, 5: Fenris Graben, 6: Hel Graben. ': base Pliocene, A: lower Miocene, A7: middle Oligocene, TP: earliest
Eocene tuff marker, C: base Tertiary, D': middle Cretaceous, D: base Cretaceous, E: approximate base Jurassic, BM: Basement, LCR:
Lower Crustal Reflector. Indexes show number of velocity values used to compute averages.

TABLE 3
Stratigraphical division and parameters of the exponential relation between porosity and depth for various normal pressure North Sea
lithologies

Sequence Lithology /o C Psg CondSg

(km" 1 ) (g cm" 3 ) (Wm-1 K-1)
Pliocene, Pleistocene
and Paleocene sandy shale 0.58 0.41 2.70 3.3
Miocene, Oligocene,
Eocene and Cretaceous shale 0.63 0.51 2.72 2.0
Pre-Cretaceous shaley sand 0.53 0.34 2.68 3.0
Sediment grain densities (p sg ) and conductivities (Cond sg ) are for the same lithologies (from Sclater and Christie, 1980, and Zoth and
Haenel, 1988). Fractional porosity / = fQ e~Cz where z is depth in kilometer, C is a decay constant, and fQ is surface porosity.

earlier rift phases, are summarized in Table 4. Note (1) The western V0ring Basin experienced re-
that the pre-Cretaceous sediments were used only to gional uplift, listric faulting and the formation of a
estimate the water-loaded basement depth (Wd) in well-defined rift unconformity of early Tertiary age
eqns. 2 and 3 for the Jurassic-Cretaceous and early (Fig. 10).
Cenozoic stretching. (2) The pre-Cenozoic Fenris Graben (Fig. 10), and
possibly also the Hel Graben underwent additional
Cenozoic. Cenozoic tectonic subsidence indicates local inversion. In the seismic data the inverted areas
that a broad region of the western V0ring Basin was are generally observed in apparent association with a
affected by extension prior to continental separation well-defined lower crustal reflector at typical depths
(Fig. 8). The crustal cross-section in Fig. 9 illustrates between 10 and 14 km.
several important observations related to the early (3) Areas affected by large-scale erosion associ-
Cenozoic rifting. ated with both the regional uplift and the local inver-
sion (Fig. Ha) are probable provenance areas for the
TABLE 4
deposition of the thick Paleocene sequence towards
Stretching factors for individual tectonic episodes assuming
the east (Figs. 9 and lib). Assuming a conformable
uniform extension (McKenzie, 1978) pre-erosional Cretaceous sediment sequence, the
amount of erosion was estimated by adding layer
Area Pre -Cretaceous Cretaceous Cenozoic thicknesses truncated by the base Tertiary rift uncon-
l l ft formity.
Nordland Ridge - 1.2 _
D0nna Terrace - 1.3 -
(4) Extensive magmatic activity is represented by
Ras Subbasin - 1.7 1.4a the inferred large magmatic body underplating the
Fulla Ridge - 1.1 1.9 continent-ocean transitional region, sills in the Cre-
Fenris Graben - 1.3 2.2 taceous sedimentary sequence, and massive emplace-
a
Western basin. ment of lavas during breakup (Fig. 9).
 CENTRAL V0RING BASIN

Voring Nordland Trondelag

Escarpment Fenris Graben Fles Fault Zone Ras Subbasin Donna Terrace Ridge Platform

NW

TECTONIC SUBSIDENCE

200 BC 100 BT 200 BC 100 BT 200 BC 100 BT 200 BC 100 BT 200 BC 100 BT 200 BC 100 BT 200 BC 100 BT
I I I _i _1
west
Fenris Graben Fulla Ridge Ras Subbasin Donna Terrace

N.

Fig. 7. Tectonic subsidence of selected stratigraphic levels computed for different sub-provinces in the V0ring Basin (black dots). Between the dots subsidence is inferred according to a uniform
lithospheric extension model (solid line) with adjustments made to fit observation points (dashed lines). BC = base Cretaceous; BT = base Tertiary. Location in Figs. 8 and 12.
66 /. Skogseid, T. Pedersen and VB. Larsen

MAIN STRUCTURAL FEATURE TTTTTTTTTT CONTINENT-OCEAN BOUNDARY DSDP/ODP DRILL SITE

BATHYMETRY (M) -IIL FLOW LIMIT (BASALTS)

Fig. 8. Cenozoic tectonic subsidence. Contour interval 100 m. Location of profiles in Figs. 7 and 9 shown with figure numbers
annotated. For abbreviations see Fig. 2.

Cretaceous. Late Jurassic-early Cretaceous rift- A comparison of the Cretaceous and Cenozoic
ing created the main structural subdivision of the subsidence patterns (Figs. 8 and 12) demonstrates a
V0ring Margin and is represented by Cretaceous major westward shift in the location of the rift zones.
tectonic subsidence (Fig. 12). The main area of This trend has also been observed by Bukovics and
extension and subsequent subsidence occurred in Ziegler (1985) and Graham and Robinson (1989).
the Ras Subbasin and TYaena Basin, whereas the Specifically, the Cretaceous subsidence patterns
Tr0ndelag Platforrn/Nordland Ridge experienced rift reveal (Figs. 12 and 13):
flank uplift (Figs. 7 and 12). The Fulla Ridge and (1) The zone of maximum subsidence narrows
Fenris Graben regions experienced relatively minor towards the northeast, and north of 67N it shifts to
Cretaceous subsidence (Fig. 7). the west.
 V0RING VOLCANIC MARGIN Voring Voring
Marginal High Escarpment Fenris Graben Fulla Ridge Fles Fault Zone
Km
100 200 300
NW

Regional Paleocene Uplift

Fig. 9. Schematic crustal profile across the western V0ring Basin and transition between continental and oceanic crust. The profile shows relative distribution of different magmatic/volcanic and
structural features forming the V0ring Volcanic Margin. Approximate location in Fig. 8.
68 J. Skogseid, T. Pedersen and V.B. Larsen

C-164
Fenris Graben
Km

10

B-9-81

h10

" 15
Fig. 10. Top: line drawing of profile C-164 illustrating listric faults truncated by the base Tertiary rift unconformity, C. Bottom: profile
B-9-81 documenting Paleocene local inversion in the Fenris Graben, with earliest Eocene flood basalts onlapping the inverted region.
The uplift is apparently associated with a prominent lower crustal reflector, LCR. O' = base Pliocene; A = lower Miocene; A' =
middle Oligocene; TP = earliest Eocene tuff marker; C = base Tertiary; D = base Cretaceous. Location in Fig. 5.

(2) The Fenris and Hei Grabens subsided less than troughs and ridges. The fact that the map represents
the eastern basin province. a time span of about 240 Ma, encompassing several
(3) There is an apparent overall relationship be- tectonic episodes, may explain the much wider region
tween late Jurassic-early Cretaceous faulting and the of subsidence compared to those for the Cretaceous
underlying basement topography. In particular, the and Cenozoic. This implies that pre-Cretaceous ex-
Nyk Syncline may have been formed passively due tension took place regionally, or shifted between
to differential extension and subsidence between the separate rift zones.
deeper basement structures (Fig. 13). The pre-Cretaceous subsidence map is considered
(4) An apparent asymmetry exists between max- poorly constrained, except for those regions where
imum basin subsidence and Moho shallowing. This there is reasonable basement control, for example
is particularly evident in the central V0ring Basin in the vicinity of the Bod0 High (Fig. 6). Here
profile (Fig 13), but also apparent in the northern ESP velocities and the positive gravity anomaly indi-
V0ring Basin profile. We point out, however, that the cate a high-density lower crustal body (TMwani and
Moho is still poorly defined. Eldholm, 1972; Planke et al., 1991), whereas the re-
flection profiles identify a horizon which correlates
Pre-Cretaceous. We are not able to resolve the with the 6.0-6.5 km/s velocity contours and dips deep
various pre-Cretaceous tectonic episodes within the into the Ttena Basin (Fig. 15). We believe that the
V0ring Basin. Nevertheless, the incorporation of a Bod0 High represents a large rotated fault block,
basement surface, albeit poorly constrained, permits once adjacent to the Nordland Ridge, that formed
construction of a pre-Cretaceous subsidence map above a "shallow detachment" during the Devonian
(Fig. 14). This shows a pattern of fairly uniform sub- collapse of the Caledonides (Fig. 3). The following
sidence with the exception of local NE-SW trending observations are supporting evidence:
V0ring Basin: subsidence and tectonic evolution 69

(1) The block contains a deep core with high interpretation of reflector D is supported by analysis
seismic velocity and density. of velocity-depth curves from a great number of seis-
(2) The base of the rotated block appears to "rest" mic refraction profiles in the different V0ring Basin
on the Moho discontinuity. sub-provinces (Table 2; Fig. 16). These show typical
(3) The regions on each side of the Bod0 High velocity gradient changes associated with the base
have experienced differential subsidence with respect Tertiary and the base Cretaceous horizons. Eldholm
to each other and to the rotated block. and Mutter (1986) have suggested that the strati-
(4) The Bod0 High gravity anomaly is part of a graphically lower gradient change takes place just
regional elongated system of gravity belts between below reflector D. This is in agreement with wells at
Lofoten and Scotland that follow the trend of the the Halten Terrace which document that the velocity
Caledonian orogenic belt (Planke et al., 1991). increase occurs within the upper Jurassic sequence
rather than at the Jurassic-Cretaceous boundary.
Evaluation of observational data Thus, the velocities appear to favour the interpre-
tation of Skogseid and Eldholm (1989) that the
We have used seismic refraction velocities aver- Fulla Ridge experienced non-deposition and/or ero-
aged for each subregion within the basin to evaluate sion during the early Cretaceous. The early Tertiary
the overall and local velocity-depth relationships (Ta- sills may have been preferentially emplaced between
ble 2). The large differences in velocities along the Jurassic sands and Cretaceous marine shales which
stratigraphic boundaries are related to the amount are of different competences. It may be argued, how-
of overburden, whereas velocities as a function of ever, that sills or composite sills may increase the
depth below seafloor are fairly consistent throughout velocity although this effect is not observed else-
the V0ring Basin (Table 5). Average velocities for where along the margin (Eldholm and Mutter, 1986).
each sub-region have been used in the depth conver- The parameters in the backstripping calculations
sion assuming constant velocity gradients. We have may introduce errors in the tectonic subsidence.
observed that the refraction velocities generally are Both the decompacted sediment thicknesses and the
slightly higher than corresponding well and interval sediment densities depend on the lithology. In addi-
velocities derived during processing of MCS data. tion, the use of proper compaction formulae, present
This may cause an overestimation of depths in the porosities and grain densities contain uncertainties.
deepest basins. Because we have used parameters that essentially are
Only Neogene well control exists in the V0ring "North Sea" averages (Tables 1 and 3), the uncer-
Basin proper. Basin subsidence and lithospheric ex- tainties for each lithology may be large. However,
tension are therefore obtained from an interpreted if the lithologic parameters are fairly uniform across
seismic data set. However, interpretation of seismic the basin, the relative subsidence pattern may be
data in areas with poor well control is often ambigu- representative, even though the absolute values are
ous. poorly constrained.
The reliability of the history of vertical movements Water depth changes affect subsidence estimates
is critically dependent on the correct definition of linearly, and lack of precise paleo water depth indi-
stratigraphic boundaries. In particular, identification cators will cause errors. Present water depths across
of the base Cretaceous horizon over the north- the V0ring Basin range from 360 to 1400 m, showing
ern Fulla Ridge may need reinterpretation in view that sedimentation has not kept pace with subsidence
of a denser grid of new better-quality data. The during Cenozoic time. Nevertheless, the erosional
base Cretaceous horizon (reflector D) of Skogseid unconformities that formed during rifting indicate
and Eldholm (1989) could be interpreted as early fairly shallow waters. We cannot exclude local wa-
Tertiary sills, placing the base Cretaceous unconfor- ter depths in the range of 0-500 m, or even more,
mity at a deeper level. Although we agree that sills within this large basin region. In this context we
exists at these stratigraphic levels, the question is would like to draw attention to the Cenozoic his-
whether the sills lie within the thick Cretaceous se- tory of the V0ring Basin and the adjacent Nordland
quence or were emplaced near the base Cretaceous Ridge region. During this period the maximum sub-
level. We chose the latter interpretation for two rea- sidence was in the western V0ring Basin. In the basin
sons. First, the Cretaceous sediments generally ex- modelling we have assumed that the early Cenozoic
hibit poorly developed stratification compared with extension did not affect areas east of the central
the pre-Cretaceous and Cenozoic sediments. At the Ras Subbasin (Figs. 7 and 8). Consequently, the Ras
Fulla Ridge, there are thick, generally well-stratified Subbasin has not undergone major extension since
sediments below reflector D, in contrast to the Creta- Jurassic-Cretaceous rifting. The model curve in Fig.
ceous sediments in the adjacent basins. Second, our 7 does not pass through the base Tertiary "obser-
70 J. Skogseid, T. Pedersen and VB. Larsen

MAIN STRUCTURAL CONTINENT-OCEAN DSDP/ODP DRILL

FEATURE BOUNDARY SITE
BATHYMETRY (M) FLOW LIMIT (BASALTS)

Fig. 11. (a) Contours of estimated amounts of Paleocene erosion. Contour interval 200 m. For abbreviations see Fig. 2.

TABLE 5
1
Mean seismic refraction velocities (km s ) at depth (s) below seafloor

Depth VB Sub-province:
1 2 3 4 5 6
58 13 6 15 12 14 n
1 2.4 2.6 2.5 2.4 2.4 2.1 2.5
51 13 4 14 12 14 7
2 3.1 3.2 2.8 2.7 3.0 3.3 3.5
42 10 4 14 n 10 3
3 4.1 5.2 4.1 3.8 4.1 4.3 4.4
4 264? 4
4.6 10
4.4 7
4.7 3
4.6 2
5.0
13 5 4 2
5 5.2 *5.5 4.8 5.4 5.3 *5.2
Abbreviations: VB: V0ring Basin, 1: Nordland Ridge and Tr0ndelag Platform, 2: Halten and D0nna terraces, 3: Traena Basin and Ras
Subbasin, 4: Fulla Ridge, 5: Fenris Graben, 7: Hel Graben. Indexes show number of velocity measurements used to compute averages.
Velocities higher than 6.0 km s _ 1 are assumed to represent crystalline basement and are not shown.
V0ring Basin: subsidence and tectonic evolution 71

MAIN STRUCTURAL CONTINENT-OCEAN DSDP/ODP DRILL

FEATURE BOUNDARY SITE
BATHYMETRY (M) FLOW LIMIT (BASALTS)

Fig. 11 (continued), (b) Isopach map of Paleocene sediments. Contour interval 200 m. For abbreviations see Fig. 2.

vation" points for either the Ras Subbasin or the 1500 m. This is not considered in the simple uniform
D0nna Terrace and Nordland Ridge. If the differ- extension model applying Airy isostasy, but has been
ence between observation and model is due entirely approximated by the dotted line in Fig. 7.
to paleo water depth, a depth of about 500 m is In view of the gravity field, Airy isostasy seems
inferred during the early Tertiary. reasonable (Tklwani and Eldholm, 1972), although
The Nordland Ridge model curve is incorrect the condition of Airy isostasy may not have prevailed
because there is evidence of syn-rift uplift during the in the past. Considering an elastic plate model, struc-
late Jurassic-early Cretaceous extension. Yielding tures with short wave-lengths will be most affected
and Roberts (this volume) have ascribed this to by deviations from Airy isostasy. Thus, features such
"footwall" uplifting along the V0ring Basin boundary as the Nordland Ridge, the Fenris Graben and the
fault and have calculated effective uplift of maximally Bod0 High region may have experienced a flexural-
72 J. Skogseid, T. Pedersen and VB. Larsen

MAIN STRUCTURAL FEATURE inlllllll CONTINENT-OCEAN BOUNDARY DSDP/ODP DRILL SITE

BATHYMETRY (M) -IIL FLOW LIMIT (BASALTS)

Fig. 12. Cretaceous tectonic subsidence. Contour interval 100 m. Location of N, northern, C, central and S, southern V0ring Basin
profiles in Fig. 13. The profile in Fig. 7 corresponds to profile C. For abbreviations see Fig. 2.

type basement response. A finite flexural rigidity tiary level where subsidence may have been overesti-
model of the lithosphere could increase the observed mated by 180-300 m. This estimate is obtained by mul-
tectonic subsidence in the Fenris Graben by approx- tiplying an assumed sea level rise of 120-200 m (Watts
imately 10%, maximally 150-200 m, and contribute and Thorne, 1984; Haq et al., 1987) with an isostatic
to the assumed Jurassic-Cretaceous uplift at the loading factor of 1.5 (Steckler and Watts, 1978).
Nordland Ridge by 100-300 m.
Ignoring eustatic sea level changes may affect the Aspects of basin development
absolute values of tectonic subsidence, but, by defini-
tion, will not influence its geographical pattern. The By applying relatively simple models of litho-
maximum error here probably relates to the base Ter- spheric extension and basin subsidence, we have
V0ring Basin: subsidence and tectonic evolution 73

NORTHERN V0RING BASIN

Voring Fulla Ridge/ Bodo Traena Donna Nordland

Escarpment Hei Graben Nyk Syncline High Basin Terrace Ridge

S10-

CENTRAL VORING BASIN

SOUTHERN VORING BASIN

Voring Halten
Escarpment Fenris Graben Fulla Ridge Fles Fault Zone Ras Subbasin Terrace
C21-329_
Km s/

Fig. 13. Crustal profiles across the V0ring Basin. The southern profile includes the velocity-depth curve from sonobuoy C21-329
(Eldholm and Mutter, 1986) showing "shallow" basement velocities. Location in Figs. 5 and 12.

described the regional aspects of the basin evolution plain some local features in the V0ring Basin which
based on a two-phase rift history: the late Jurassic- we address below.
early Cretaceous and early Cenozoic rift episodes.
Several pre-Cretaceous rift events are probable, but Pre-Cenozoic basin
cannot be resolved. Furthermore, we have indicated The Bod0 High region is the only "window" into
that the regional modelling approach does not ex- the pre-Cretaceous part of the basin. The large Bod0
74 J. Skogseid, T. Pedersen and VB. Larsen

_ MAIN STRUCTURAL FEATURE nli'Miii CONTINENT-OCEAN BOUNDARY O DSDP/ODP DRILL SITE

BATHYMETRY (M) FLOW LIMIT (BASALTS)

Fig. 14. Pre-Cretaceous tectonic subsidence. Contour interval 200 m. For abbreviations see Fig. 2.

High block of crystalline basement (Fig. 15) was deformation and thinning, consequently bringing the
formed during Scandian compression and the fol- mantle up towards the base of the observable base-
lowing Devonian collapse of the orogen (Fig. 3). ment block. This is schematically illustrated in Fig. 3,
It appears, however, that the block has been later- whereas Figs. 6, 13 and 15 show the present configu-
ally displaced without experiencing large-scale inter- ration where the Bod0 High basement block appears
nal deformation during the subsequent extensional to be bounded at depth by the hanging wall fault
episodes. Speculatively we relate this to successive plane, directly above Moho. During the respective
reactivation of the bounding listric faults and as- episodes of extension this configuration caused dif-
sociated formation of large half-grabens on either ferential syn-tectonic vertical movements, which may
side of the main basement blocks. The lower crust, have been expressed in actual uplift of the block
on the other hand, reacted to extension by ductile proper.
V0ring Basin: subsidence and tectonic evolution 75

BOD0 HIGH
SCALE: H=V
C-167
SP5000 SP6000 SEAFLOOR
I
BTERTIARY

30 30

NORDLAND
FULLA RIDGE TR/ENA BASIN RIDGE
C-165
SP2000 SEAFLOOR

km/s

30

C-164
SP 5000 SP6000 SEAFLOOR SP7000
I
RTFRTIARY

Fig. 15. Crustal sections and isovelocities across the Bod0 High region. The faulted base Cretaceous horizon (bold lines) and
other prominent deep reflectors (thin lines) are modified from Skogseid and Eldholm (1989). Shading indicates crystalline basement
underneath the Nordland Ridge and in the rotated fault blocks discussed in the text. There is no vertical exaggeration. Locations in
Fig. 5.

The Bod0 High lies at the southern extension of entire mid-Norwegian margin, although offset by ap-
a positive gravity anomaly belt which trends along parent transfer zones, overlie a series of crystalline
the Lofoten shelf subparallel to a more prominent basement blocks similar to the Bod0 High. However,
anomaly over the Lofoten Islands (TMwani and Eld- the most prominent gravity anomalies are produced
holm, 1972). These gravity anomalies are also charac- only where the basement blocks are associated with
terized by shallow Moho (Goldschmidt-Rokita et al., high-density rocks, probably mantle material and/or
1988) and shallow basement (Chroston and Brooks, eclogite zones (TMwani and Eldholm, 1972).
1989), most likely composed of rotated basement The basement configuration has implications for
blocks (Mokhtari et al, 1989). We speculate that the lateral extent of Cretaceous basin subsidence.
similar gravity anomaly belts which exists along the If the Bod0 High experienced uplift during Cre-
76 J. Skogseid, T. Pedersen and V.B. Larsen

Velocity (km/s) gional sinistral shear movements along a Great Glen

0 2 4 6 8 10 Fault trend. Drawn from our observations, we put
1 1 1 1 1 1
forward an alternative explanation where the uplift,
Fulla Ridge or relative lack of subsidence, is caused by differen-
1
tial lithospheric extension accompanied by footwall
2 J * V* uplift along the major normal faults (Yielding and
/ j *
Roberts, this volume), and that shear movements
3 (Gabrielsen and Robinson, 1984; Gowers and Lunde,
1984; CaselU, 1987; Gr0nlie and Roberts, 1987) rep-
l Base Tertiary
4 resent a second-order effect.
' \' Base
5 Jv. Cretaceous L
S" Cenozoic basin
Q

J .1\ The magmatic surge during breakup at 57.5 Ma

* 1
v Base I
Jurassic
is manifested by a large volume of extrusive rocks
J \ l Basementr
including the wedges of seaward dipping reflectors
that partly compose the V0ring Marginal High. The
high is underlain by a thickened, high-velocity, lower
crustal body (Fig. 6).
9 T 1 i 1 1 "T This "underplated" body is believed to be a result
Fig. 16. Relationship of stratigraphy and refraction velocities of decompressional melting as the asthenosphere
along the Fulla Ridge. Vertical bars indicate average velocities rose passively beneath the stretched and thinned
(Table 1) and range in depth of the respective horizons. lithosphere (White et al., 1987). Many, often con-
flicting, evolutionary models have been advanced to
account for the V0ring volcanic margin, and the
taceous rifting and, subsequently, less and delayed drilling of ODP Site 642 (Fig. 2) recently has pro-
subsidence compared with the adjacent basins, the vided key boundary conditions. The volcanic mar-
tectonic subsidence map (Fig. 12) does not accu- gin evolution and its relationship to the North At-
rately represent the area of lithospheric deformation. lantic Volcanic Province (Morton and Parson, 1988)
In this case Cretaceous rifting becomes more uni- have been discussed in detail by Eidholm et al.
form along strike, and the apparent westward shift (1989). Here, we restrict ourselves to the thermo-
of the area of maximum subsidence north of 67N mechanical modelling of Pedersen and Skogseid
becomes less pronounced. (1989). Assuming instantaneous rifting, they esti-
We observe that the late Jurassic-early Cretaceous mated the thermal anomaly needed to produce the
extension generally formed collapse grabens within observed amount of melt and regional uplift across
older half-graben basins (Figs. 13 and 15). This the V0ring Margin. By calculating the amount of
relation between Jurassic-Cretaceous faults and the partial melting, they concluded that an astheno-
underlying basement topography shows that the deep spheric thermal anomaly of about +50C is neces-
structural setting influenced the location of later sary.
faulting. The old basement-involved fault zones were Seismic data in the V0ring Basin (Skogseid and
locally reactivated, but Jurassic-Cretaceous faults are Eldholm, 1989) and the M0re Basin (Myhre et al,
generally basement-detached. 1989) document that the construction of the marginal
The asymmetry observed between the location of highs was preceded by intrusive activity which pro-
maximum basin subsidence and Moho shallowing gressively moved up-section and was focused towards
along the central V0ring Basin profile (Fig. 13) may the line of continental breakup. The time between
be explained in light of the early Cretaceous erosion initial rift-related faulting and final breakup may
and rift flank uplift described above. If the Moho have been as long as 17 Ma if the observed faults
relief also reflects the region of maximum mantle- are dated as old as Campanian/Maastrichtian as pro-
lithosphere thinning, the main thermal anomaly was posed by Skogseid et al. (in press) In the models we
probably located under the east flank of the basin have placed the initiation of Cenozoic rifting at 65
in the Halten and D0nna terraces and the Nord- Ma, the Cretaceous-Tertiary boundary.
land Ridge area. In this case this area would have This episode is characterized by high amounts of
experience syn-tectonic uplift and associated erosion extension, with values ranging from 1.4 in the west-
rather than subsidence. Brekke and Riis (1987) have ern Ras Subbasin to 2.2 in the Fenris Graben (Ta-
suggested that these areas were peneplaned during ble 4). Extension of this magnitude is not reflected
the early Cretaceous and related the uplift to re- in the observed faults. However, rift-related faulting
V0ring Basin: subsidence and tectonic evolution 77

2-.

100 200 Km
Continent-Ocean Voring Fenris Fulla Ras
Boundary Escarpment Graben Ridge Subbasin

Fig. 17. Diagram showing the general relationship between early Tertiary uplift/erosion and Paleocene redeposition.

is normally associated with syn-tectonic subsidence. pw are the densities of magma, mantle and water,
But Fig. 7 shows that this subsidence is almost at a respectively. Using the values in Table 1 for mantle
minimum during the early Cenozoic rifting. Because and water, 3-4 km for i7 mag and 2.8-2.9 g cm - 3
extension took place as gradual stretching associated for /9mag (McKenzie, 1984), AS" is 500-900 m with a
with intrusions and heating over a period of 17 Ma, corresponding increase in .
we question whether relatively unconsolidated Cre- In addition to magmatic effects, lateral heat flow
taceous shales would have deformed in a manner will dominate the heat budget for narrow basins
resolvable as faults in seismic data, instead under- such as the Fenris and Hei Grabens (Alvarez et
going ductile deformation. The Cretaceous sequence al., 1984). This has two important consequences:
generally shows poor internal stratification. the total subsidence will be approached much faster
The values derived from subsidence only (Ta- than indicated in eqn. 4, and the lateral heat flow
ble 4) may be reduced by taking into account the will cause syn-rift thermal uplift of the flanks. A
intrusive activity expressed as dikes in the lower crust first-order estimate of the rapid cooling effect was
and sills in the sediments (Royden et al., 1980). How- obtained by approximating 5tot in eqn. 3 based on the
ever, a reduction in the stretching estimates is not present total subsidence. Adding this correction to
supported by the 4-6 km crustal thickness following AS calculated above, we obtain an early Cenozoic
early Cenozoic rifting. This implies relatively high factor of 1.9-2.4. A flank uplift at the adjacent Fulla
rates of extension and thinning. Ridge of about 300-400 m is inferred (Alvarez et al.,
With reference to eqn. 2, which describes the 1984) and in good agreement with observations.
initial "fault-controlled" subsidence due to extension, We believe these calculations made at selected
the modelling in the Fenris Graben yields an initial locations (Fig. 7) provide first-order estimates of ef-
uplift component consistent with the local inversion fects due to rapid cooling and magmatic underplating
within the graben as observed on seismic data (Fig. along the entire length of these grabens because the
10). This local inversion, which is superposed on the Cenozoic subsidence, graben width and underplated
regional uplift (Figs. 11a and 17), is caused in part "magma thickness" vary little along strike. The local
by the large pre-Cenozoic depth to basement in the inversion of the Fenris Graben, due to the under-
graben (cf. eqn. 2), equivalent to a much reduced plating and "infill" of the Moho relief, probably will
pre-rift crustal thickness (McKenzie, 1978). not be reversed during the later subsidence as the
In addition the high magmatic production would underplated magma represents an actual growth of
have affected both the initial and total tectonic the crustal thickness.
subsidence/uplift (Foucher et al., 1982). To examine Apparent differential vertical movements in the
the influence of magmatism on the early Cenozoic early Tertiary are also observed along the western
estimates in the Fenris Graben we isostatically cor- flank of the Fulla Ridge/Nyk Syncline (Fig. 13).
rected the total tectonic subsidence (Fig. 7) for the These movements are related to similar processes of
presence of the magmatic underplating. This gives an differential crustal thinning as discussed at the Bod0
increase, AS: High. The basement block beneath the western Fulla
Ridge has probably experienced syn-tectonic uplift
Pmag
AS=HmgPm (5) or differential subsidence in the early Tertiary, and
Pm - Pw formation of the Nyk Syncline may be due, in part,
where Hmag is the magma thickness, and p mag , pm and to these relative vertical movements. Similarly, we
78 J. Skogseid, X Pedersen and VB. Larsen

observe a correlation between an inverted Cenozoic sands within the thick (3000 m) Paleocene sequence
structure along the Fles Fault Zone (Fig. 13) and in the Flett Basin (Hitchen and Ritchie, 1987) as
deeper basement structures. This inverted area, as- analogs, we infer sands within the V0ring Basin Pa-
sociated with reverse faults and termed the Heiland- leocene sequence. During the late Paleogene and
Hansen Arch (Blystad et al., in press.), however, Neogene the western V0ring Basin subsided rapidly
is marginally within the area that experienced early and marine shales provide a suitable cap rock. Sand
Cenozoic extension and it may be difficult to docu- bodies may also exist along both sides of the Fulla
ment syn-tectonic uplift. Ridge and locally along the Bod0 High. These highs
seem to have received little to no sediments through-
Petroleum potential out the early Cretaceous, because they experienced
local rift-related uplift and, thus, were prone to ero-
Hydrocarbon exploration in parts of the eastern sion. These sequences may be genetically correlated
V0ring Basin will provide valuable information when with the Cenomanian and TUronian sand-prone de-
assessing the hydrocarbon potential in the western posits drilled west of the Nordland Ridge (Hastings,
V0ring Basin. However, the impact of the early 1986). Cretaceous marine shales interfingering with
Cenozoic rifting and associated magmatic activity the sand lobes could form suitable cap rocks for
must also be considered. these possible reservoirs.
The main source rock for oil in the North Sea The main uncertainty in evaluating the area's hy-
and offshore mid-Norway is the Kimmeridgian hot- drocarbon potential are related to quantifying the
shale (Faber and Stahl, 1984; Baird, 1986). These amount of extension and the associated thermal in-
organic-rich sediments are also found onshore East fluence. Tbo many parameters are unknown making
Greenland and have been correlated with the North it impossible to determine accurate temperatures and
Sea sequences both stratigraphically and geochem- degrees of maturity. In particular we have poor con-
ically. This indicates deposition under similar envi- trol on the actual distribution of different lithologies.
ronmental conditions throughout the North Atlantic Thus, conductivity may exceed acceptable ranges
realm (Requejo et al., 1989). In addition to this for one or more units, which could alter tempera-
source, we speculate that potential source rocks may tures dramatically. However, we tentatively infer that
have been deposited within basins with restricted wa- the potential reservoirs are promising, as shown in
ter circulation conditions during two other periods. Fig. 18.
During the early Cretaceous, favourable conditions
for source rock deposition may have existed in the Summary and conclusions
Fenris and Hei Grabens, which were separated from
the eastern basins by the Fulla Ridge-Bod0 High Although uniform extension modelling is a sim-
complex. During the Paleocene and earliest Eocene plification of the actual geodynamic processes, it
a land area along the central axis of the developing provides a method for quantifying lithospheric ex-
Norwegian-Greenland Sea may have created condi- tension and amounts of subsidence. Because many
tions for restricted basins along the present margin critical parameters for basin modelling, particularly
(Eidholm et al., 1989). This was a period domi- thermal modelling, remain unknown in the V0ring
nated by a hot, humid climate and high rates of Basin, our results are primarily presented as tectonic
biogenic production (Froget et al., 1989). In partic- subsidence maps.
ular, the Paleocene-early Eocene V0ring Basin may The tectonic evolution of the V0ring Basin com-
have been closed towards the north, as the margin prises a series of predominantly extensional episodes
approached the uplifted Lofoten-Vesterlen area. since Devonian time, which governed the initial
From seismic data we infer the existence of poten- crustal and lithospheric configuration in the basin
tial reservoir rocks deposited in two different strati- modelling. The two latest main episodes, the late
graphic intervals and tectonic settings. One is the Jurassic-early Cretaceous and early Cenozoic rift-
product of the major erosional event caused by early ing, have had a dominant influence on the regional
Cenozoic extension and uplift. These sediments, de- structural and subsidence pattern in the study area,
rived from the area west of the Fulla Ridge, were whereas older structures are only exposed locally.
transported eastward and deposited in the V0ring However, the regional configuration of the crys-
Basin where they reach a maximum thickness of talline basement surface can be derived from seismic
about 1500 m (Fig. 17). Using the Paleocene and refraction data.
Eocene sands derived from the East Shetland Plat- values for Jurassic-Cretaceous extension range
form and deposited in the Southern Viking Graben from 1.1 to 1.7 along a basin transect. Maximum
(Lovell, 1984; Conort, 1986), or even more likely the extension occurred in the Ras Subbasin and the
V0ring Basin: subsidence and tectonic evolution 79

Subsidence Temperature History Burial H i s t o r y

300 200
a se Cretaces
ou

150

a se T e r t i ayr

a
E

200 100 0
Ma.

Fig. 18. Correlation between tectonic subsidence, temperature and burial history for an assumed well at the western Fulla Ridge.
Lithology and conductivity parameters are given in Table 3.

Tfrasna Basin, whereas the Fenris and Hei Grabens should be considered.
experienced minimal extension on the order of (3) Early Cenozoic magmatic activity caused both
1.1-1.3. subsidence and thermal anomalies not inherent in
The locus of maximum early Cenozoic stretching uniform extension modelling. In particular the high
shifted westward with respect to that of the Jurassic- amounts of extension in the western V0ring Basin
Cretaceous episode. Estimates of values range do not correlate with observed extensional faults.
from 1.4 in the western Ras Subbasin to 1.9-2.5 in This is accounted for in part by assuming ductile
the Fenris Graben. This Cenozoic extension lasted deformation of the Cretaceous shales and by intro-
for at least 10 Ma before itfinallycaused continental ducing extension by magmatic intrusion documented
separation in the earliest Eocene. During the syn- by the dikes and sills in the western V0ring Basin
rift phase the western V0ring Basin was uplifted, region. Moreover, the effects of underplating will in-
and an area about 250 km wide (Fig. 17) may have crease the amount of initial uplift and decrease total
been exposed to erosion. Also associated with rifting subsidence.
was high magmatic activity which caused accretion (4) Finally, lateral heatflow might be important
of thick magmatic bodies at the base of the crust, in the development of the "narrow" Fenris and Hel
emplacement of sills in Cretaceous sediments and Grabens. The high amount of extension, /?=1.9-2.5,
regional fissure eruptions. During breakup a surge in the Fenris Graben resulted in a crystalline crust
of extrusive volcanism formed large lava wedges only 4-6 km thick. This suggests that these regions
represented as seaward dipping reflectors. were close to the stage of seafloor spreading in
Local features in the V0ring Basin show subsi- the late Paleocene. The narrow rift configuration
dence patterns that deviate from a uniform extension caused considerable lateral heatflow and rapid cool-
model. Among these are: ing, which may have strengthened the lithosphere
(1) The Bod0 High region which represents and subsequently lead to continental separation fur-
a "window" into the pre-Cretaceous part of the ther west. Lateral heatflow is also indicated by uplift
basin. Here the pre-Jurassic structural setting caused of the western Fulla Ridge and the relatively rapid
apparent differential crustal thinning during the thermal subsidence during Cenozoic time.
main extensional episodes. The lithospheric thin- These results, which provide an initial quantitative
ning caused probable syn-tectonic uplift of the Bod0 description of the Mesozoic and Cenozoic evolu-
High block followed by post-tectonic differential sub- tion of the V0ring Basin, impact assessment of the
sidence. petroleum potential of the basin. Although lithology
(2) Offset between Moho shallowing and maxi- and conductivity data are required in order to resolve
mum basin subsidence may explain rift-flank uplift the temperature history and allow reliable maturity
during Jurassic-Cretaceous rifting, particularly in estimates, the most prospective formations appear to
the eastern V0ring Basin, the Halten and D0nna be possible Cretaceous sands along the Fulla Ridge
terraces and Nordland Ridge. This development sug- and Bod0 High and the thick Paleocene sequence in
gests that simple shear models of basin subsidence the western V0ring Basin.
80 J. Skogseid, T. Pedersen and VB. Larsen

Acknowledgments Dalland, A. and Worsley, D., 1988. A lithostratigraphic scheme

for the Mesozoic and Cenozoic succession offshore Norway
This study has benefited from the drilling re- north of 62N. Norw. Pet. Dir., Bull., 4 (ISBN 82-7275-241-9),
sults and, in particular, subsequent seismic studies 67 pp.
presented in the Leg 104 Scientific Result Volume Dallmeyer, R.D., 1988. Polyphase tectonothermal evolution of
the Scandinavian Caledonides. In: A.L. Harris and D J . Fettes
(Eldholm, Thiede, Taylor et al., 1989). We also ac- (Editors), The Caledonian-Appalachian Orogen. Geol. Soc.
knowledge data made available by Statoil A.S. We London, Spec. Publ., 38: 365-379.
thank Steinar Thor Gudlaugsson and Sverre Planke Dewey, J.F., 1988. Extensional collapse of orogens. Tectonics, 7:
for stimulating discussions and are particularly grate- 1123-1139.
ful to Olav Eldholm for initiating the project and Duindam, P. and Van Hoorn, B., 1987. Structural evolution of
the western Shetland continental margin. In: J. Brooks and
for providing advice and critical review. We also
K.W. Glennie (Editors), Petroleum Geology of North-West
thank two anonymous reviewers for their critical and Europe. Graham and Trotman, London, pp. 765-773.
constructive reviews and for providing editorial ex- Eldholm, O. and Mutter, J.C. 1986. Basin structure of the Nor-
pertise. This study was supported by grants from the wegian Margin from analysis of digitally recorded sonobuoys.
Norwegian Research Council for Science and the J. Geophys. Res., 91: 3763-3783.
Humanities, Statoil A.S. and VISTA. Eldholm, O., Thiede, J., Taylor, E. et al. 1987. Proceedings of
the Ocean Drilling Program, Initial Rep. 104, (Pt. A), 783 pp.
Eldholm, O., Thiede, J., Taylor, E., 1989. Evolution of the
References V0ring volcanic margin. In: O. Eldholm, J. Thiede, E. Taylor
et al. (Editors), Proc. ODP, Sei. Results, 104, College Station,
Alvarez, E, Virieux, J. and Le Pichon, X., 1984. Thermal con- Texas, (Ocean Drilling Program), pp. 1033-1065.
sequences of lithospheric extension over continental margins: Eldholm, O., Thiede, J., Taylor, E. et al. 1989. Proc. ODP,
the initial stretching phase. Geophys. J. Astron. Soc, 78: Sei. Results, 104, College Station, Texas, (Ocean Drilling
389-411.
Program), 1141 pp.
Austrheim, H., 1987. Eclogitization of lower crustal granulites by
Eldholm, O., Skogseid, J., Sundvor, E. and Myhre, A.M., 1990.
fluid migration through shear zones. Earth Planet. Sei. Lett.,
The Norwegian-Greenland Sea. In: A. Grantz, G.L. Johnson
81: 221-232.
and J. Sweeney (Editors), The Arctic Ocean Region, Vol.
Baird, R.A., 1986. Maturation and source rock evaluation of
L, The Geology of North America. Geological Society of
Kimmeridge Clay, Norwegian North Sea. Am. Assoc. Pet.
America, Boulder, Colo., pp. 351-364.
Geol., Bull., 70: 1-11.
England, P.C. and Houseman, G.A., 1988. The mechanics of the
Blystad, P., Faerseth, R.B., Larsen, B.T., Skogseid, J. and
Tibetan Plateau. Philos. Trans. R. Soc. London, Ser. A, 326:
T0rudbakken, B., in press. Structural elements of the Norwe-
301-320.
gian continental shelf, Part 2. The Norwegian Sea. Norw. Pet.
Faber, E. and Stahl, W, 1984. Geochemical surface exploration
Dir., Bull., 6
for hydrocarbons in the North Sea. Am. Assoc. Pet. Geol.,
Brekke, H. and Riis, F., 1987. Tectonics and basin evolution
of the Norwegian Shelf between 62N and 72N. Nor. Geol. Bull., 68: 363-386.
Tidsskr., 67: 295-322. Foucher, J.-P, Le Pichon, X. and Sibuet, J.-C, 1982. The ocean-
Bucher-Nurminen, K., 1991. Mantle fragments in the Scandina- continent transition in the uniform lithospheric stretching
vian Caledonides. Tectonophysics, 190: 173-192 model: role of partial melting in the mantle. Philos. Trans. R.
Bukovics, C. and Ziegler P.A. 1985. Tectonic development of the Soc. London, Ser. A, 305: 27-43.
Mid-Norway continental margin. Mar. Pet. Geol., 2: 2-22. Froget, C , Desprairies, A., Latouch, C. and Miallet, N., 1989.
Bukovics, C., Shaw, N.D., Cartier, E.G. and Ziegler, P.A., 1984. Paleoenvironmental significance of Cenozoic clay deposits
Structure and development of the Mid-Norway continental from the Norwegian Sea: ODP Leg 104. In: O. Eldholm, J.
margin. In: A.M. Spencer et al. (Editors), Petroleum Geol- Thiede, E. Tayloret al. (Editors), Proc. ODP, Sei. Results, 104,
ogy of the North European Margin. Graham and Trotman, College Station, Texas, (Ocean Drilling Program), pp. 41-60.
London, pp. 407-423. Gabrielsen, R.H. and Robinson C , 1984. Tectonic inhomo-
B0en, F., Eggen, S. and Vollset, J., 1984. Structures and basins geneities of the Kristiansund-Bod0 Fault complex, off shore
of the margin from 62-69N and their development. In: A.M. mid-Norway. In: A.M. Spencer et al. (Editors), Petroleum Ge-
Spencer et al. (Editors), Petroleum Geology of the North ology of the North European Margin. Graham and Trotman,
European Margin. Graham and Trotman, London, pp. 3-28. London, pp. 397-406.
Caselli, F., 1987. Oblique-slip tectonics, mid-Norway shelf. In: Gabrielsen, R.H., Faerseth, R., Hamar, G. and R0nnevik, H.C.,
J. Brooks and K.W. Glennie (Editors), Petroleum geology 1984. Nomenclature of the main structural features on the
of North-West Europe. Graham and Trotman, London, pp. Norwegian Continental Shelf north of the 62nd parallel. In:
1049-1063. A.M. Spencer et al. (Editors), Petroleum Geology of the
Celerier, B., 1986. Models for the evolution of the Carolina North European Margin. Graham and Trotman, London, pp.
Trough and their limitations. Ph.D. Thesis, Massachusetts 41-60.
Institute of Technology, Cambridge, Mass., 206 pp. Goldschmidt-Rokita, A., Sellevoll, M.A., Hirschleber, H.B. and
Chroston, P.N. and Brooks, S.G., 1989. Lower crustal seismic Avedik, F , 1988. Results of two seismic refraction profiles off
velocities from Lofoten-Vesterlen, North Norway. Tectono- Lofoten, Northern Norway. Nor. Geol. Under., Spec. Publ.,
physics, 157: 251-269. 3: 49-57.
Conort, A., 1986. Habitat of Tertiary hydrocarbons, South Viking Gowers, M.B. and Lunde, G., 1984. The geological history of
Graben. In: A.M. Spencer et al. (Editors), Habitat of Hy- Traenabanken. In: A.M. Spencer et al. (Editors), Petroleum
drocarbons on the Norwegian Continental Shelf. Graham and Geology of the North European Margin. Graham and Trot-
Trotman, London, pp. 159-170. man, London, pp. 237-252.
V0ring Basin: subsidence and tectonic evolution 81

Graham, R.H. and Robinson, N.D., 1989. A regional structural Myhre, A.M., Skogseid, J., Eldholm, O., Faleide, J.I., Gud-
study of the Norwegian Continental Shelf between 6250' and laugsson, S.T., Planke, S. and Vgnes, E., 1989. Norwegian
67. Paper presented at the Norwegian Petroleum Society Continental margin: structural and stratigraphical styles. 28th
Workshop, 18-20 October, 1989, Stavanger. Int. Geol. Congr., Washington, D.C. (abstract).
Gr0nlie, A. and Roberts, D., 1987. Dextral strike-slip duplexes Nelson, P.H.H. and Lamy, J.M., 1987. The M0re/West Shetland
of Mesozoic age along the Hitra-Snsa and Verran faults, area: a review. In: J. Brooks and K.W Glennie (Editors),
M0re-Tr0ndelag Fault Zone, Central Norway. NGU-Rep., 87, Petroleum Geology of North-West Europe. Graham and
139 pp. Trotman, London, pp. 775-784.
Hastings, D.S., 1986. Cretaceous stratigraphy and reservoir Norton, M.G., 1986. Late Caledonian extension in western
potential, mid Norway continental shelf. In: A.M. Spencer Norway: A response to extreme crustal thickening. Tectonics,
et al. (Editors), Habitat of Hydrocarbons on the Norwegian 5: 195-204.
Continental Shelf. Graham and Trotman, London, pp. 287- Pedersen, T. and Skogseid, J., 1989. V0ring Plateau volcanic
298. margin: extension, melting and uplift. In: O. Eldholm, J.
Haq, B.U., Hardenbol, J. and Vail, PR., 1987. Chronology of Thiede, E. Taylor et al. (Editors), Proc. ODP, Sei. Results,
fluctuating sea levels since the Triassic. Science, Reprint Ser., 104, College Station, Texas, (Ocean Drilling Program), pp.
235: 1156-1167. 985-991.
Hinz, K., Dostman, H.J. and Hanisch, J. 1984. Structural ele- Planke, S., Skogseid, J. and Eldholm, O., 1991. Crustal structure
ments of the Norwegian Sea continental margin, Geol. Jahrb., off Norway, 62-70N. Tectonophysics, 189: 91-107.
A75: 193-211. Requejo, A.G., Hollywood, J. and Halpern, H.I., 1989. Recogni-
Hitchen, K. and Ritchie, J.D., 1987. Geological review of the tion and source correlation of migrated hydrocarbons in upper
West Shetland area. In: J. Brooks and K.W Glennie (Editors), Jurassic Hareelv Formation, Jameson Land, East Greenland.
Petroleum Geology of North-West Europe. Graham and Am. Assoc. Pet. Geol., Bull., 73: 1065-1088.
Trotman, London, pp. 737-749. Roll, A., 1974. Langfristige Reduktion der Mchtigkeit von
Hossack, J.R., 1986. The geometry of listric faults in the Sedimentgesteinen und ihre Auswirkung eine bersicht.
Devonian basins of Sunnfjord, W. Norway. J. Geol. Soc. Geol. Jahrb., Ser. A, 14, 76 pp.
London, 141: 629-637. Royden, L., Sclater, J.G. and von Herzen, R.P., 1980. Continental
Jamtveit, B., 1987. Metamorphic evolution of the Eiksunddal margin subsidence and heatflow: important parameters in
eclogite complex, Western Norway, and some tectonic impli- formation of petroleum hydrocarbons. Am. Assoc. Pet. Geol.,
cations. Contrib. Mineral. Petrol., 95: 82-99. Bull., 64: 173-187.
R0nnevik, H.C., Eggen, S. and Vollset, J., 1983. Exploration
Larsen, H.C., Armstrong, G., Marcussen, C , Moore, S. and
of the Norwegian Shelf. In: J. Brooks (Editor), Petroleum
Stemmerik, L., 1989. Deep seismic data from the Jame-
Geochemistry and Exploration of Europe. Blackwell, Oxford,
son Land Basin, East Greenland. Abstracts, IFP/ILP /IUGS
pp. 71-94.
Exploration Research Conference, Aries. Terra Abstracts, 1.
Sawyer, D.S., 1985. Total tectonic subsidence: a parameter for
Larsen, V.B., 1987. A synthesis of tectonically-related stratig-
distinguishing crust type at the U.S. Atlantic continental
raphy in the North Atlantic-Arctic region from Aalenian to
margin. J. Geophys. Res., 90: 7751-7769.
Cenomanian time. Nor. Geol. Tidsskr., 67: 281-193.
Sclater, J.G. and Christie, P.A.F., 1980. Continental stretching:
Le Pichon, X. and Sibuet, J.-C, 1981. Passive margins: a model
an explanation of the post-mid-Cretaceous subsidence in
of formation. J. Geophys. Res., 86: 3708-3720.
the central North Sea basin. J. Geophys. Res., 85: 3711-
Lovell, J.P.B., 1984. Cenozoic. In: K.W. Glennie (Editor), In-
3739.
troduction to the Petroleum Geology of the North Sea.
Skogseid, J. and Eldholm, O., 1987. Early Cenozoic crust at the
Blackwell, Oxford, pp. 151-169.
Norwegian continental margin and the conjugate Jan Mayen
Manby, G.M. and Hambrey, M.J., 1989. The structural setting Ridge. J. Geophys. Res., 92: 11471-11491.
of the Late Proterozoic tillites of East Greenland. In: R.A. Skogseid, J. and Eldholm, O. 1989. V0ring Continental Margin:
Gayer (Editor), The Caledonide Geology of Scandinavia. seismic interpretation, stratigraphy and vertical movements.
Graham and Trotman, London, pp. 299-312. In: O. Eldholm, J. Thiede, E. Taylor et al. (Editors), Proc.
McClay, K.R., Norton, M.G., Coney, P. and Davis, G.H., 1986. ODP, Sei. Results, 104, College Station, Texas, (Ocean Drilling
Collapse of the Caledonian orogen and the old red sandstone. Program), pp. 993-1030.
Nature, 323: 147-149. Skogseid, J., Pedersen, T, Eldholm, O. and Larsen, B.T., in
McKenzie, D., 1978. Some remarks on the development of press. Tectonism and magnatism during during NE Atlantic
sedimentary basins. Earth Planet. Sei. Lett., 74: 25-32. continental break-up: The V0ring Margin. In: B.C. Storey, T.
McKenzie, D. 1984. A possible mechanism for epiorogenic uplift. Alablaster and R J . Plankhurst (Editors), Magmatism and the
Nature, 307: 616-618. Causes of Continental Break-up. Geol. Soc. London, Spec.
Mokhtari, M., Pegrum, R.M. and Seilevoll, M.A., 1989. A geo- Publ.
physical study of the Norwegian continental margin between Spencer, A.M. et al. (Editors), 1984. Petroleum Geology of
67 and 69N. Seismological Observatory, Univ. of Bergen, the North European Margin. Graham and Trotman, London,
Seismo-Series. 436 pp.
Morton, A.C. and Parson, L.M., 1988. Early Tertiary volcanism Spencer A.M. et al. (Editors), 1986. Habitat of Hydrocarbons
and the opening of the NE Atlantic. Geol. Soc. London, on the Norwegian Continental Shelf. Graham and Trotman,
Spec. Publ., 39, 477 pp. London, 354 pp.
Mutter, J.C., 1984. Cenozoic and late Mesozoic stratigraphy and Steckler, M.S. and Watts, A.B., 1978. The Gulf of Lion: sub-
subsidence history of the Norwegian margin. Bull. Geol. Soc. sidence of a young continental margin. Nature, 287: 425-
Am., 95: 1135-1149. 429.
Mutter, J.C., Talwani, M. and Stoffa, PL., 1984. Evidence for Surlyk, F., Piasecki, S., Rolle, F., Stemmerik, L., Thomsen, E.
a thick oceanic crust adjacent to the Norwegian margin. J. and Wrang, P., 1984. The Permian basin of East Greenland.
Geophys. Res., 89: 483-502. In: A.M. Spencer et al. (Editors), Petroleum Geology of
82 J. Skogseid, T. Pedersen and VB. Larsen

the North European Margin. Graham and Trotman, London, White, R.S., Spence, G.D., Fowler, S.R., McKenzie, D.P., West-
pp. 303-315. brook, G.K. and Bowen, A.N., 1987. Magmatism at rifted
Talwani, M. and Eldholm, O., 1972. The continental margin continental margins. Nature, 330: 439-444.
off Norway. A geophysical study. Geol. Soc. Am. Bull., 83: Yielding, G. and Roberts, A., 1992. Footwall uplift during
3375-3608. normal faulting implications for structural geometries in
Talwani, M., Udintsev, G. et al. 1976. Initial Reports of the the North Sea. In: R.M. Larsen, H. Brekke, B.T Larsen and
DSDP, Vol. 38. U.S. Government Printing Office, Washington, E. Talleraas (Editors), Structural and Tectonic Modelling and
O.C., 1256 pp. its Application to Petroleum Geology. Norwegian Petroleum
Tiircotte, D.L., Haxby, W.F. and Ockendon, J.R., 1977. Litho- Society (NPF), Special Publication 1. Elsevier, Amsterdam,
spheric instabilities. In: M. Talwani and Pitman W.C. Ill pp. 289-304 (this volume).
(Editors), Island Arcs, Deep Sea Trenches and Back-arc Ziegler, P.A., 1988. Evolution of the Arctic-North Atlantic and
Basins. Am. Geophys. Union, Maurice Ewing Ser., 1: 63-69. the western Tethys. Am. Assoc. Pet. Geol., Mem., 43: 198 pp.
Watts, A.B. and Thorne, J.A., 1984. Tectonics, global changes Zoth, G. and Haenel, R., 1988. Thermal conductivity. In: R.
in sea-level and their relationship to stratigraphic sequences Haenel, L. Rybach and L. Stegena (Editors), Handbook of
at the U.S. Atlantic continental margin. Mar. Pet. Geol., 1: Terrestrial Heat-flow Density Determination. Kluwer Aca-
319-339. demic Publisher, Dordrecht, pp. 449-466.

J. SKOGSEID Department of Geology, University of Oslo, P.O. Box 1047, Blindem, 0316 Oslo 3, Norway
T. PEDERSEN Department of Geology, University of Oslo, Oslo, Norway
VB. LARSEN StatoilA/S, Stavanger, Norway
 83

The orientation of minor fault plane striae and the

associated deviatoric stress tensor as a key to the fault
geometry in part of the More-Trondelag Fault Zone,
on-shore central Norway
D. Bering

The Verran Fault and the associated Hitra Fault are prominent regional features within the northern part of the M0re-
Tr0ndelag Fault Zone. On the northern side of the Beitstadfjord the main Verran Fault Zone consists of at least three main
parallel faults. These structures, together with smaller interconnecting faults, are considered to form part of an intricately
linked system. The faults are clearly post-Caledonian in age with a poly-phasal brittle deformation history which can be seen
from the pervasive brecciation on all scales together with hydrothermal alteration of the rock products. Measurements of the
orientation of slickensides and other kinematic indicators on minor polished fault planes establish a system of slip-vectors with
an oblique-slip normal to pure dip-slip normal main component. The slip-vectors together with the orientation of the minor
slip planes can be used to compute the deviatoric stress tensor responsible for the deformation. It is concluded that the fault
pattern along the main Verran Fault can best be interpreted as a set of dip-slip normal faults which are a response to pure
crustal extension. The minor interconnecting faults between the major zones are considered to represent transfer faults linking
the motions on the main faults.

Introduction TM image data, Rindstad and Gr0nlie (1986) and

Gr0nlie and Roberts (1989) established an interest-
The purpose of this paper is to demonstrate the ing pattern of anastomosing major and minor faults
importance and usefulness of establishing the pale- in the Verran area north of the Beitstadfjord in
ostress responsible for the deformations when dis- the northern part of Tr0ndelag (Fig. I). This pat-
cussing the geometry and movements along major tern resembles the "strike-slip duplex" structures of
fault systems which lack mappable marker horizons. Woodcock and Fischer, 1986, and, together with the
The technique is applied to the Verran Fault Zone field observations of strong cataclasis and the de-
which is part of the M0re-TY0ndelag Fault Zone velopment of thick breccias along the majority of
(Fig. 1). The study of this major fault zone is the sub- the faults, led the authors to postulate that major
ject of ongoing work and a more detailed description dextral lateral movements are responsible for this
of the field relationship, an assessment of various structure. This would imply that at least part of the
methods of paleostress estimation and a discussion M0re-TY0ndelag Fault Zone, with its off-shore ex-
of the different stress systems active along the zone, tension, had a strong component of strike-slip move-
will be the subject of future papers. ments.
The M0re-TY0ndelag Fault Zone (Gabrielsen and However, the bedrock consists of ductile poly-
Ramberg, 1979; Gabrielsen et al., 1984) was defined deformed ortho- and paragneisses lacking typical
by satellite imagery analysis and consists of strong marker horizons, and the relative motions along the
lineaments following the Caledonian structural grain different faults are thus difficult to establish. This
from lesund on the coast of Norway through M0re paper is a part of a major work in which orientation
and Tr0ndelag (Fig. I). Based on the similarity of the data for minor fault planes with striae are used to
geometry and orientation, these authors linked the classify the individual slip and to establish the stress
lineaments with established faults on land and with system(s) responsible for the deformation and, thus,
the development of fault structures off-shore. possibly provide arguments for the fault geometry
On the basis of detailed analyses of Landsat and kinematics operating on a larger scale.

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 83-90. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
84 D. Bering

Fig. 1. The M0re-Tr0ndelag Fault Zone in relation to the main structural elements oif-shore Norway. The main map shows
the structural features north of the Trondheimsfjord discussed in the text. Note the three almost parallel faults with the minor
interconnecting faults resembling strike-slip duplexes (see text for discussion).

The method in dm and metre scales. They are characterized

by polished hard greenish surfaces developed in
The fault products along the major and minor coherent fine-grained roch powder. The planes show
faults in the Verran System are characterized by poly- striae developed as rather coarse grooves and locally
phasal brecciation, hydrothermal alteration and local occurrence of kinematic indicators, such as comb
prehnite and stilbite veining (Gr0nlie and Roberts, structures, tool marks etc. (Hancock and Barke,
1989). It is evident from the field that the bedrock 1987), are observed. The similarity of the slip planes
has undergone more than one phase of brittle defor- and their structural style are believed to reflect the
mation. One earlier phase is marked by quartz-filled same deformational event.
extensional veining and rather coherent fault prod- The stilbite veining serves as a further time con-
ucts with quartz-epidote matrix. straint as it is sometimes cut and polished by the slip
The structures considered in this paper cut the planes: in other situations the slip planes are cut by
earlier ones, and are considered to be of the second small extensional veins filled with stilbite.
(or third) phase in the brittle deformation history. After development of the slip planes, the major
They consist of numerous small discrete fault planes fault breccias show evidence of one (or possibly two)
The fault geometry in part of the M0re-Tr0ndelag Fault Zone, onshore central Norway 85

BIAXIAL STRAIN ing the movement on a larger scale. However, taken

2 together at different localities, the minor fault planes
a

/TV \
must in some way reflect the overall stress-strain re-
lationship for that particular deformation on a larger

L
scale. Classification of the individual planes into nor-

( (-}
U* W3 mal, reverse, dextral and sinistral slip planes, and
computation of the deviatoric stress tensor responsi-
ble for their development, will therefore contribute
to the discussion of the sense of movements and the
\J iU ^^\L^ kinematics of the faults on a regional scale.
From the work of Anderson (1942) it is known that
the orientation of faults is geometrically linked to the
orientation of the principal axes of the stress ellipsoid
TRIAXIAL STRAIN
responsible for the deformation. In a volume of rock
deformed under plane strain condition (Fig. 2a),
one conjugate fault system will develop. These faults
will be oriented in such a way that the intersection
line between them will define 2, \ will bisect the
acute angle, and 3 will bisect the obtuse angle
( > 2 > 3 being the three principal directions in
the stress ellipsoid responsible for the deformation).
Under triaxial strain conditions, the situation is
more complicated (Fig. 2b). Two systems of conju-
gate faults must develop to accommodate the strain.
Fig. 2. The three principal compressive stress axes in rela- 2 will in this situation bisect the acute angle be-
tion to the orientation of conjugate faults under (a) biaxial,
and (b) triaxial strain condition. Modified from Hancock, 1985. tween the lines of intersection of the two systems.
Likewise, ?> will bisect the larger angle and \ will
be oriented normal to the 2 - 3 plane (Reches and
later phase of brittle deformation. This deformation Dieterich, 1983).
is marked by development of irregular kink bands in Most natural faults are oblique-slip and do not
the dm and metre scales and produced a rather loose frequently occur in conjugate systems, and the sym-
and incoherent late brecciae. metry criterion above is seldom met. Based on the
The latest deformation in the area is not just Bott (1959) analysis, many methods of stress esti-
confined to the fault breccias, but is also observed mation assume that the slip along a fault occurs in
in the relatively undisturbed bedrock. The structures the direction of the resolved shear stress (e.g., Carey
developed are mostly small fractures and fissures, but and Bruinier, 1974; Angelier, 1979; 1984; Gephart
some fault planes with slickensides show evidence of and Forsyth, 1984; Reches, 1987; and others), and
movements. Both in their orientation and the asso- several inversion techniques have been developed.
ciated mineralization, these structures are evidently The one used here is the direct inversion method of
separated, both in time and association, from the Angelier et al. (1984), which minimizes a function
brittle deformation structures related to the devel- that depends on both the slip-shear angle and the
opment of the thick fault breccias, and will not be shear.
analysed further in this contest. The deformational
events subject to analysis in this paper are, thus, con- Application of paleostress estimates to the
sidered to be somewhere in the middle of the brittle Verran fault zone
deformation history of the major fault zones; not
the earliest and not the latest, but they do reflect an In the Fosen area, northwest of the TYondheims-
event which had a great influence in the development fjord, two main fault systems are observed the
of the thick breccias along the major faults as seen Hitra Fault and the Verran Fault (see Figs. 1 and 3).
on a regional scale. The Hitra Fault is considered to be an older system
The movement along the regional faults is thus than the Verran Fault, mainly on the basis of de-
accommodated (at least in part) by development of formation products (see also Gr0nlie and Roberts,
the minor slip planes. The orientation and type of 1989). The fault rock along this zone displays older
movement along one particular small fault may be elements of ductile-semibrittle deformation of the
just a local accommodation, not necessarily reflect- minerals quartz and feldspar and almost always have
86 D. Bering

Fig. 3. Map of the Verran area showing the location of the Hitra and Verran Fault. The Hitra Fault displays Riedel-like shears (upper
central) and the Verran Fault System consists of three almost nearly parallel faults with minor interconnecting faults. Eaqal-area (lower
hemisphere) projection of minor faults with their associated striae from the main localities a, d, g, h, 1, q and r are shown together with
the different populations after separation. See text for discussion.

a fine-grained cataclastic matrix. Very little veining axis could have had a near vertical orientation at
is observed, except for localities where the fault has some point in the history of the fault, thus indicating
been reactivated. Analyses of the stress system re- a strike-slip movement.
sponsible for the Hitra Fault are not presented here. The Verran Fault system consists of three almost
It should be mentioned, however, that the TM image parallel major faults (Fig. 3), with at least three
displays some Riedel-like shears in the central part minor interconnecting faults between the two larger
of the fault, indicating that the intermediate stress ones. In the Beitstadfjord, seismic profiling by NGU
The fault geometry in part of the M0re-Tr0ndelag Fault Zone, on-shore central Norway 87

DEX R SIN

DEX H SIM

DEX \ SIM 'DEX R SIM DEX I SIN

'^
/

Y / .

* #
/ r\\ * * * /*
\ \+* ** 7
\;
* * / : \ /'
DEX N SIN DEX . It SIM DEX K SIN

Fig. 4. Classification of the minor fault planes into dip-slip normal or reverse, sinistral or dextral strike-slip faults (from Angelier,
1984). In (a) faults with a high degree of certainty regarding the relative movements from the whole area are plotted, and diagrams b-f
show the areal distribution of the different types. The majority of the faults show oblique-slip normal movements although some pure
strike-slip faults are recognized, particulary in the central part of the main zone and in the interconnecting faults.

(B0e, 1989) shows a series of faults developed in movements is given by computing the distribution
what is considered to be rocks of Middle Jurassic of transverse and lateral movements, expressed as
age. These faults follow the trend to both the major ratios relative to the net slip. The components are
and minor ones on land (Fig. 3). computed as simple trigometric functions of the dip
All the faults are marked in the field by strong of the fault and the pitch of the striation. As the sum
brecciation and development of numerous minor slip of the three vectors that represent vertical, horizontal
planes. Although the slip planes show a wide variety transverse and horizontal lateral components of slip
in orientation, the majority of the smaller faults equals the net slip vector, all the points in the
are oriented almost parallel to the larger structures. diagram fall within a circle (Fig. 4)
(Fig. 3, stereograms a, d, g, h, 1, q and r). Where fault In Fig. 4a faults from the whole area are plot-
populations show conjugate systems, the intersection ted. Only those faults with a very high degree of
lines of these faults are parallel with the larger certainty regarding the relative movements are used.
structure (i.e., stereograms j , k, m and p in Fig. 3). Although faults with all types of relative motion are
There is then a strong indication of a link between found, oblique-slip normal faults are in the majority.
the development of the major and minor struc- Of the 206 faults plotted, 82% show a normal-slip
tures. component. The oblique-slip and pure strike-slip
Based on the measurements of the attitude of faults show a near equal distribution of sinistral and
the minor faults and the direction of the slip vector dextral components.
observed from the striations, all the faults have been Looking at the areal distribution of the various
classified as normal, reverse, dextral and sinistral faults (Fig. 4b-h), some general pattern can be seen.
relative movements. By using the principles drawn The central part of the southern main fault (the
up by Angelier et al, 1984, a good picture of fault Verran Fault proper), Fig. 4h, contributes most to
88 D. Bering

the population of oblique-slip faults with a sinistral Looking at the minor interconnecting faults, one
component. In two of the interconnecting faults, i.e., gets a more complicated picture. Although some so-
Fig. 4g and d, faults with a larger degree of "oblique- lutions show pure extension (a and e, and to some
ness" are found, and some classify as pure strike-slip extent f and g), one solution has the minimum prin-
faults with both sinistral and dextral relative move- cipal axis, 3, near the vertical (Fig. 5b), showing
ments. that a pure compressional event existed locally. Solu-
From the stereograms in Fig. 3, it is evident that tion h gives an intermediate position of the principal
more than one population exists in most of the main axes with both extension and compression in the
localities chosen for this analysis (i.e., Fig. 3a, d, horizontal plane, i.e., strike-slip faulting.
h, 1). Inversion of these unmodified data sets will The horizontal extension directions from locali-
obviously give a solution for the deviatoric stress ties in the major NE-SW striking faults (Fig. 51-n)
tensor which has no geological meaning. One must have the same NW-SE direction. Solution c gives a
first try to separate the data sets into homogeneous uniaxial stress ellipsoid and an extension direction
sub-sets. The separation technique adopted here is cannot be inferred. The central part of the southern
firstly based on the small amount of field evidence fault, together with this general extension direction,
of cross-cutting relationships that exists in a few also shows an almost E-W to ENE-WSW direction,
localities. Subsequently, the whole population from indicating that the region as a whole departs from
one locality is inverted and the mean stress tensor a plane strain condition (see Fig. 2 and discussion
is computed. This tensor is then used, for each fault above).
plane, to compute the theoretical slip vector. The The paleostress pattern from the whole region
angular difference between this and the actual slip then gives a picture of a general stress regime with
direction (from the striations), called the misfit angle, the maximum principal compressive stress, , ori-
is then a measure of how good the tensor solution is ented almost vertical. This stress regime will then
for that particular fault. Faults with high misfit angles generally favour development of normal, more or
are then eliminated from the population and a new less dip-slip faults. The departure from this general
stress tensor is computed. By repeatedly eliminating stress state found locally, is believed to represent lo-
faults, and accepting a certain amount of misfit (10- cal stress, variable both in time and space, induced by
15 degrees in this analysis), one obtains a population movements along the major faults. This conclusion is
of faults with one homogenous solution. The faults also supported by the fact that the minor intercon-
which were eliminated are then used for computing a necting faults show the largest degree of departure
second tensor, with the same process of eliminating from the regional stress state. It is believed that the
faults with high misfit angles as in the first case and overall fault geometry in the context of this stress
the process is repeated. After computing a series pattern is that of a large system with parallel major
of tensors in this way, one must check whether faults with dip-slip movements. The interconnecting
the faults could belong to more than one solution, faults in this picture will represent transfer faults
indicating that one or more of the tensors are of little linking the relative motions on the bigger structures.
geological meaning. This process is very important, In a large fault system like the M0re-TY0ndelag
particularly in cases where field evidence is scarce or Fault Zone, with its long history, it is likely that
lacking. different stress systems have been active at different
The results of this separation procedure are shown times and at different places, and it is not believed
in Fig. 3, where the sub-sets of fault slip data are that the analysis presented above represent other
plotted. The resulting deviatoric stress tensors are than a part of the total deformation history. It
represented in Fig. 5a-n by plotting the orientations cannot, then, be ruled out that lateral relative move-
of the principal axes in the different stress ellipsoids. ments have been active at some time in the region,
The stereograms also show the directions of the although no field evidence to support this has been
horizontal extension and compression. found in the Verran Fault System. On the basis of
Most of the data sets from the major NE-SW the data and analysis presented above, it is believed
striking faults give solutions with a near vertical that the mappable fault pattern now met within the
orientation of the maximum principal axis, , (i.e., Verran area is the response to pure crustal extension.
Fig. 5a, c, i, j , k, 1, m, n). The three solutions i,
j and k from the same locality have almost the Conclusions
same orientations and probably represent the same
tectonic event. One solution from one locality (d) The orientation of minor fault planes along with
has the intermediate principal axis, ^ in the vertical their associated striae in fault zones can be a valu-
position, indicating strike-slip movements. able tool in the discussion of the kinematics of large
The fault geometry in part of the M0re-Tr0ndelag Fault Zone, on-shore central Norway 89

x> ^
#
4^

Fig. 5. The results after computation of the deviatoric stress tensors (a-n) from the fault populations in Fig. 3, shown here as an
orientation of the principal stress axes (symbols: five-, four- and three-legged stars mark the position of , 2 and ^ axes, respectively,
in the stress ellipsoids. Also shown (arrows) are the direction of the horizontal extension (compression) directions.

faults. This is particularly important in areas lack- References

ing mappable marker horizons and where the fault
pattern can be interpreted on a large scale both as Anderson, E.M., 1942. The Dynamics of Faulting. Olivier and
Boyd, Edinburgh, 1st ed.
strike-slip or dip-slip movement.
Angelier, J., 1979. Determination of the mean principal direc-
In the case of the Verran area, the analysis above tions of stresses for a given fault population. Tectonophysics,
points to the conclusion that the major and minor 56: T17-T26.
faults in the area are of dip-slip normal type, and Angelier, J., 1984. Tectonic analysis of fault slip data sets. J.
the fault pattern on the map scale is not evidence Geophys. Res., 98(B7): 5835-5848.
Bott, M., 1959. The mechanisms of oblique faulting. Geol. Mag.,
for major strike-slip movements along the M0re-
96.
Tr0ndelag Fault Zone. B0e, R., 1989. NGF Vinterm0te, Geolognytt Nr. 22 (abstract).
Carey, E. and Brnier, B., 1974. Analyse thorique et numerique
d'un modele mecanique elementaire applique l'etude d'une
Acknowlegdement population de failles. C.R. Acad. Sei. Paris, 279D: 891-894.
Gabrielsen, R.H. and Ramberg, I.B., 1979. Fracture patterns
in Norway from Landsat imagery: results and potential use.
The author wishes to thank the Norwegian oil Proc. Norwegian Sea Symp. Troms0, 23: 1-28.
companies Norsk Hydro A/S and Statoil for financing Gabrielsen, R.H., Faerseth, R., Hamar, G. and R0nnevik, H.,
the work. The cooperation with B.T. Larsen and R. 1984. Nomenclature of the main structural features of the
Faerseth is especially acknowledged. Norwegian Continental Shelf north of the 62nd parallel. Pet.
90 D. Bering

Geol. N. Eur. Margin, pp. 41-60. Hancock, P.L. and Barka, A.A., 1987. Kinematic indicators on
Gephart, J.W. and Forsyth, D.W., 1984. An improved method for active normal faults in Western Turkey. J. Struct. Geol.,
determining the regional stress tensor using earthquake focal 9: 573-584.
mechanism data: application to the San Fernando earthquake Reches, Z. and Dieterich, J.H., 1983. Faulting of rocks in three-
sequence. J. Geophys. Res., 89: 9305-9320. dimensional strain fields. I. Failure of rocks in polyaxial,
Gr0nlie, A. and Roberts, D., 1989. Resurgent strike-slip duplex servo-control experiments. Tectonophysics, 95: 111-132.
development along the Hitra-Snsa and Verran Faults, M0re- Reches, Z., 1987. Determination of the tectonic stress tensor
Tr0ndelag Fault Zone, Central Norway. J. Struct. Geol., 11: from slip along faults that obey the Coloumb yield condition.
295-305. Tectonics, 6: 849-861.
Hancock, P.L., 1985. Brittle microtectonics: principles and prac- Woodcock, N.H. and Fischer, M., 1986. Strike-slip duplexes. J.
tices. J. Struct. Geol., 7: 437-457. Struct. Geol., 8: 725-735.

D. BERING Geology Institute, University of Bergen, Bergen, Norway

Present address: Norwegian Petroleum Directorate, P.O. Box 600, N-4001 Stavanger
 Extensional tectonic history of the western Barents Sea
C.A. Dengo and K.G. R0ssland

Extensional tectonics was the dominant mode of deformation in the western Barents Sea beginning in Late Devonian-Early
Carboniferous time and culminating with the break-up between Greenland and Norway in Paleocene time. This paper describes
extensional tectonic elements that have been recently mapped, places them within a regional tectonic framework and, discusses
the significance that this extensional deformation may have to petroleum exploration in the Barents Sea. The oldest tectonic
event that can be mapped regionally in the western Barents Sea occurred in Late Devonian-Early Carboniferous time in
response to the initial rifting between Greenland and Norway. This event established a fundamental basement architecture of
half-grabens and intervening highs that influenced the location of younger basins and hydrocarbon traps, the deposition of
source, reservoir and seal facies and the timing of hydrocarbon maturation and migration. Major basins that formed at that time
are the Troms0, Bj0rn0ya, Hammerfest and Nordkapp basins, as well as a newly identified basin named the Dia Graben. The
Nordkapp and Bj0rn0ya basins are each sub-divided into two major half-grabens.
Salt deposition in these basins marks the end of rifting and the beginning of thermal subsidence, in Late Carboniferous
time. During Permian through Late Triassic, a time of relative tectonic quiescence, some basement-involved normal faults were
reactivated, possibly in response to the Uralian Orogeny and loading by westward prograding clastic sediments. Fault reactivation
is apparent along the eastern margin of the Dia Graben which provided a structural control for the position of the Early-Mid
Triassic shelf edges. Basement-detached normal faults developed in association with salt withdrawal, mostly in the Dia Graben
and Nordkapp basin.
In Late Jurassic-Early Cretaceous times there was renewed crustal extension in the Barents Sea. However, only those basins
west of the Loppa High underwent rapid tectonic subsidence. The Cretaceous Troms0, Bj0rn0ya and south-western Hammerfest
basins, and perhaps the Kval0y and Harstad basins are interpreted as pull-apart basins that formed in response to oblique rifting
between the Barents Sea and Greenland. Basement faults bounding the Troms-Finnmark Platform and southern Nordkapp
Basin also were reactivated. Basement-detached normal faulting developed east of the Loppa High and west of the southern
Nordkapp Basin. There was additional salt withdrawal in the Nordkapp basins, Dia Graben and possibly the Troms0 Basin.
Northward-propagating rifting led to sea floor spreading in Late Paleocene time and the development of a passive margin
along the western Barents Sea. Most of the observed deformation is west of the Senja Ridge; however, there was continued
movement along basement-involved normal faults bounding the Harstad, Troms0, Bj0rn0ya and southern Nordkapp basins,
accompanied by additional salt withdrawal. Neogene contractional deformation, manifested primarily as basin inversion, is
apparent along most of the passive margin, west of Loppa High.

Introduction ceous time and culminated with sea floor spreading

in Late Paleocene time.
The western Barents Sea is located on a passive This paper discusses the extensional deformation
continental margin that developed as the North At- that occurred in the western Barents Sea, focusing
lantic and Arctic oceans opened in response to the primarily on the Late Devonian-Early Carbonifer-
break-up between Greenland and Norway in Late Pa- ous deformation. Mapping at this structural level
leocene time. Multiple pre-break-up tectonic events has permitted identification of tectonic elements in
are recorded in the Barents Sea. The earliest is the addition to those already described (R0nnevik and
Caledonian Orogeny, related to the collision and Jacobsen, 1984; Faleide et al., 1984; Berglund et al.,
suturing of Laurentia and Baltica (and intervening 1986). During the Late Devonian-Early Carbonif-
microplates) in Late Silurian-Early Devonian time erous, asymmetric crustal extension imparted to the
(see review by Smith, 1988). Subsequent tectonic basement of the western Barents Sea a fundamental
events were characterized by crustal extension that architecture characterized by half-grabens and inter-
began with the initial rifting between Greenland and vening highs. Recognizing the importance of this rift-
Norway in Late Devonian-Early Carboniferous time, ing event is critical to understanding the petroleum
continued with rifting in Late Jurassic-Early Creta- geology of the western Barents Sea because it es-

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 91-107. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
92 CA. Dengo and K. G. R0ssland

LEGEN
D WESTERN BARENTS SEA
REGIONA
L SEISMI
C GRID

0 25 50 75 100 125k
m

Fig. 1. Location map of the western Barents Sea. The study area is located between the northern coast of Norway and 75N and
between 15 and 35. The regional seismic grid used for this study is indicated and numbered lines represent geoseismic lines
referenced in the text.

tablished the tectonic framework that influenced (1) The area of study is located between longitudes
the location of younger basins and structures, (2) the B W and 3300/ E and between latitudes 7430/ and
distribution of source, reservoir and seal facies and 7030' N (Fig. 1). This area covers only a portion
(3) hydrocarbon maturation and migration. Some of the entire Barents Shelf; however, it is the area
of these aspects are discussed at the end of this that has been opened for hydrocarbon exploration
paper. and, consequently, it is where the geophysical data
Extensional tectonic history of the western Barents Sea 93

TECTONIC TIMING CHART

INTERIOR RIFT B A S I N ,
BASEMENT DETACHED N O R M A L F A U L T I N G
BASEMENT I N V O L V E D EXTENSION

Fig. 2. Summary chart showing timing of major tectonic events and resultant styles of deformation in the western Barents Sea.

coverage is best. Regional syntheses of the entire tures were mapped using the regional seismic grid
Barents Shelf and its tectonic relationship to adjacent shown in Fig. 1. An infill seismic grid and grav-
Svalbard, Franz Josef Land, the Soviet Arctic and ity data were used to compliment the regional grid.
Greenland have recently been published in a volume Key stratigraphic horizons were dated using available
on the geological evolution of the Barents Shelf well data and the seismic horizons were correlated
(Harland and Dowdeswell, 1988). regionally using concepts of sequence stratigraphy.
Three regional paleo-tectonic maps were made A summary of the tectonic events and associated
to represent the major periods of crustal exten- deformational styles for the major basins described
sion, the Late Devonian-Early Carboniferous, Late in this paper is presented in Fig. 2.
Jurassic-Early Cretaceous and Tertiary. Tectonic fea-
94 CA. Dengo and K. G. R0ssland

Pre-rift deformation half-grabens that formed during the initial phase

of crustal extension between Greenland and Nor-
The Caledonian Orogeny, which occurred in Early way (R0nnevik, 1981). Within the study area, these
Paleozoic time with the closure of the Iapetus Ocean half-grabens constitute the TYoms0, Bj0rn0ya, Ham-
and the suturing of Greenland to Norway and Spits- merfest and Nordkapp basins and the Dia Graben.*
bergen, formed the metamorphic basement of the Basement-involved normal faulting was the domi-
Barents Sea. Caledonian structures have a north- nant style of deformation (Fig. 2).
east trend in northern Norway (Sturt et al., 1978; R0nnevik and Jacobsen (1984) and Faleide et al.
Townsend, 1987), a north-west trend in Spitsbergen (1984) used the tectonic reconstructions performed
(see Harland, 1985) and most likely extend across the by Harland (1969) and Ziegler (1978) to propose that
Barents Sea (Faleide et al., 1984). The trend of Cale- the Late Devonian-Early Carboniferous crustal ex-
donian structures is remarkably coincident with the tension in the western Barents Sea resulted from the
trend of many of the younger extensional basins, sug- combined effects of sinistral strike-slip faulting in the
gesting that Caledonian structures had fundamental western part of the Barents Sea and a conjugate dex-
control in localizing the position of major north- tral strike-slip fault in the central Barents Sea. The
east and north-west trending basin-bounding normal interpretations presented here support and build on
faults. The influence that pre-existing thrust faults the concept that the Late Devonian-Early Carbonif-
have in the localization of younger extensional faults erous deformation is important in the Barents Sea.
has been documented in many rift basins world-wide,
However, they differ from those of previous workers
particularly where deep seismic reflection data are
in that we propose that the deformational kinemat-
available (e.g., Nelson et al., 1985; Cheadle et al.,
ics is dominantly dip-slip normal faulting with little
1987; Ratcliffe and Burton, 1985; Swanson, 1986).
demonstrable strike-slip deformation.
Chadwick (1985) and Lake and Karner (1987) pro-
posed that reactivation of Caledonian and Variscan- Tectonic elements
age thrust faults as normal faults controlled the
The major tectonic elements that were active In
development of several basins in southern Britain
Late Devonian-Early Carboniferous time are shown
and adjacent offshore.
on the paleo-tectonic map (Fig. 3) and on six parallel,
Following the closure of the Iapetus Ocean, in east-trending, regional geoseismic lines (Figs. 4-9).
Late Silurian-Early Devonian time, erosion of the The structures represented on the paleo-tectonic
Caledonian Orogen led to the deposition of thick map are at deeper structural levels than the Upper
continental clastic sediments, collectively termed the Carboniferous horizon interpreted on the geoseismic
Old Red Sandstone (see Roberts and Sturt, 1980). lines (Figs. 4-9). Three regional structural trends
In Spitsbergen, these rocks are found largely in a are observed on the paleo-tectonic map (Fig. 3):
half-graben that was interpreted to have formed (1) north-west trending structures in Finnmark East
in response to divergent strike-slip movement on (south-east of Nordkapp), (2) north-east trending
the paleo-Hornsund and Billefjorden faults (Harland structures that dominate the fabric and which extend
et al., 1974; Harland, 1985). Arguments in favour across most of the study area from the Norwe-
and against major strike-slip movements in Late gian coast to about 74N and (3) north-north-west
Devonian time have been proposed (Lamar et al., trending structures that extend from 74N to Spits-
1986; Harland et al., 1988). Not withstanding, in the bergen.
western Barents Sea there are indications of deep The principal features of the Late Devonian-
extensional structures that may be of the same age Early Carboniferous deformation are discussed be-
as those in Spitsbergen. These structures are poorly
low. Although the Kval0y Basin is shown on the
imaged on seismic data and to date have not been
paleo-tectonic map it is not possible at this time to
mapped.
determine whether or not it is sited over a Late
Devonian-Early Carboniferous predecessor.
Early Devonian-Early Carboniferous interior
rift phase The Troms-Finnmark and Berlevg Faults. Two
major faults separate the less deformed TYoms-
The oldest sequence of rocks that can be mapped
regionally in the western Barents Sea is composed *At the time of this writing this article, there is not yet an
of Upper Devonian to Lower Carboniferous rift-fill officially approved nomenclature for some tectonic elements
alluvial fan and floodplain clastic sediments mixed described here. The Norwegian Petroleum Directorate has pro-
posed that the names Maud Basin and Nordsel High be used
with carbonates and evaporites (Faleide et al., 1984; for elements described here, i.e., the Dia Graben and the Saiika
Steel and Worsley, 1984). These were deposited in High, respectively.
Extensional tectonic history of the western Barents Sea 95

LEGEND
BASEMENT-INVOLVED NORMAL FAULTS
WESTERN BARENTS SEA
BASIN AXIS, SHOWING DIRECTION OF PLUNGE LATE DEVONIAN - EARLY CARBONIFEROUS
PALEOTECTONIC MAP
+
EXPLORATION WELLS

0 25 50 75 100 125km

Fig. 3. Late Devonian-Early Carboniferous paleo-tectonic map. Areas highlighted in grey represent basement highs relative to adjacent
basins.

Finnmark Platform from the much more de- here the Berlevg Fault. The Berlevg Fault is
formed Barents Shelf. West of 24, the TYoms- the southern boundary of the Nordkapp Basin.
Finnmark Fault zone separates the Platform from It has a consistent down-to-the-north offset, yet
the Hammerfest and TYoms0 basins. East of 24, it appears to have a more complicated history
the structural boundary is a fault located to the than just one of dip-slip motion. Displacement
north of the TYollfjord-Komagelv fault, named along the basin-bounding normal faults at the
 - LINE 722500-
8 - LINE 7225A-8
5
WEST BJ0RN0Y
A BASI
N LOPP
A HIGH
U- FENRI
S GRABE
N * ' ^^

182230-8
4

Fig. 4. Geoseismic lines 722500-84 and 7225A-85. Horizons shown represent base Tertiary (T), base Cretaceous unconformity (K), Mid Triassic (Anisian, A), top Permian (P), and top
Carboniferous (C). Dashed lines are form lines shown only to represent configuration of stratal patterns. S represents salt. Lines have been squeezed and are vertically exaggerated five times. For
the locations of lines see Fig. 1.

LINE 7 1 5 7 - 2
8 -
I I OAI ivm
iI i - HAMMERFES
T BASI
N
KVAL0
Y BASI
N SENJ
A RIDG
E - - TROMS
0 BASI
N - - LOPP
A HIGH - ^ F E NSRGRABE
I N* * FENRI
S HIGH * , *
^ - SAIIK
A HIGH -

201
4 2115-8
2

a.
ft
S*3

Fig. 5. Geoseismic lines 7157-82 and 7155-86. Lines have been squeezed and are vertically exaggerated five times. A legend for the horizons shown can be found in Fig. 4. For the location of lines
see Fig. 1.
 INE 7 3 0 5 - 8-
WEST BJORN0Y
A BASI
N DIA GRABE
N

Fig. 6. Geoseismic lines 7305-84 and 7305-85. Lines have been squeezed and are vertically exaggerated five times. A legend for the horizons shown can be found in Fig. 4. For the location of lines
see Fig. 1.

Fig. 7. Geoseismic lines 7245-84 and 7245-85. Lines have been squeezed and are vertically exaggerated five times. A legend for the horizons shown can be found in Fig. 4. For the location of lines
see Fig. 1.
 - LINE 7 3 2 5 - 8 4 LINE 7 3 2 5 - 8 5 CD
EAST oo
STAPPEN HIGH WEST BJ0RN0YA BASIN BJ0RN0YA DIA HIGH
WEST BASIN
2015-84

Fig. 8. Geoseismic lines 7325-84 and 7325-85. A legend for the horizons shown can be found in Fig. 4. Lines have been squeezed and are vertically exaggerated five times. For the location of lines
see Fig. 1.

LINE 7 3 4 5 - 8 5

- U DIA HIGH

-2.0 f

11

a
Fig. 9. Geoseismic lines 7345-84 and 7345-85. Lines have been squeezed and are vertically exaggerated five times. A legend for the horizons shown can be found in Fig. 4. For the location of lines
see Fig. 1.
Extensional tectonic history of the western Barents Sea 99

southern end of Nordkapp must be transferred Loppa-Nordkapp Platform area. A significant

to the Berlevg Fault as strike-slip motion. Be- change has been made to previously published maps
tween 28-31E there is a small elongated basin in the area traditionally called the Loppa-Nordkapp
that is bounded to the south by the Berlevg Platform. This area has been divided into a series
Fault and to the north by a basement-high. This of subparallel, north-east trending half-grabens and
basin may be a pull-apart basin that opened in re- intervening highs. The Late Devonian-Early Car-
sponse to strike-slip motion along the Berlevg Fault. boniferous elements now recognized and displayed
No evidence was found to indicate a north-westward on Figs. 3-9 are: (1) Dia High, (2) Dia Graben,
extension of either the Berlevg or TYollfjord- (3) Fenris High, (4) Fenris Graben and (5) Loppa
Komagelv fault zones into the Hammerfest Basin, High.
westof24E. The Loppa High is mapped to extend north of
7305/ N. At this latitude the Loppa High is faulted,
Finnmark East area. The Finnmark East area, with the northern segment of the block being struc-
located between the Nordkapp and Varanger basins turally lower than the southern segment. The adja-
and the TYoms-Finnmark Platform (Fig. 3), is unique cent Fenris High continues northward, unfaulted. At
because it is the only area where north-west trending 7335' the Fenris and Loppa Highs merge and
faults are dominant. become a single basement block extending north as
North-west and north-east trending faults intersect the eastern flank of the S0rkapp Basin. The north-
to form the broad Tkna low (Fig. 3). Flanking this east trending Senja Ridge merges with the Loppa
low are two basement highs (Alta and Varanger) High forming a single basement block at 7240' ,
that are the basement flanks of the Nordkapp and an interpretation that is also supported by the gravity
Varanger basins, respectively. data (Fig. 10). The Ringvass0y-Loppa Fault Com-
plex, which was active in Late Permian time, may
Nordkapp Basins. The Nordkapp Basin is divided have been active during Early Carboniferous time.
into two adjacent, opposing half-grabens: Nordkapp The Fenris and Dia Highs form a single basement
Basin South and Nordkapp Basin North. The two block south of latitude 72N, north of which they
half-grabens are resolved on the free-air gravity map bifurcate into two distinct basement highs (Figs.
as distinct lows (Fig. 10). Nordkapp North is the 3-7). The Dia High trends north-east as one of the
deeper of the two basins and is separated form most prominent features of the study area. The Dia
Nordkapp South by a prominent interbasinal ridge Graben, flanked by the Dia, Fenris and Loppa Highs
(Nordkapp High) that trends north-west, oblique to is a narrow graben at about 72N and widens to
the basinal axes. Both basins are flanked by basement the north (Figs. 3-9). A small interbasinal ridge is
highs and have high-relief margins and an opposing located within the Dia Graben towards its northward
low-relief margin. The Saiika High forms the high- extension (Fig. 3).
relief, north-west margin of Nordkapp South and the
Alta High forms the low-relief, south-east margin of Bjomoya and Tromso Basins and Stappen
this basin. The Saiika and Nordkapp highs flank the High. The Bj0rn0ya Basin has been subdivided into
Nordkapp Basin North. A complex pattern of multi- two half-grabens, East Bj0rn0ya and West Bj0rn0ya,
directional fault trends marks the intersection of the latter being the larger and deeper of the two
the Nordkapp High with the graben-flanking highs. basins (Figs. 3, 8 and 9). The eastern margins of
The asymmetry of Nordkapp has previously been both basins is formed by the faults flanking the Senja
recognized, particularly for Nordkapp Basin South Ridge and its continuation as the Loppa High. The
(R0nnevik et al., 1982; Faleide et al., 1984; R0nnevik north-western margin of the West Bj0rn0ya Basin
and Jacobsen, 1984). is the Stappen High. This margin is characterized
by a series of down-to-the-south-east normal faults.
Hammerfest Basin. In Late Devonian-Early A narrow intrabasinal ridge, located at about 21,
Carboniferous time the Hammerfest Basin was a separates East from West Bj0rn0ya basins (Figs. 3, 8
north-east trending half-graben with the high-relief and 9).
margin on the north-west flank. The basin is flanked The TYoms0 Basin has a very thick Cretaceous
by the Dia High to the north-west (Figs. 4 and 5), fill that obscures the deeper structure (Fig. 5). The
the Vads0 High to the north (Figs. 6 and 7) and basin-bounding faults show Early Carboniferous off-
the Saiika High to the south-east. The south-west set; however, the presence of salt in the TYoms0
boundary of Hammerfest is a complex fault zone that Basin, which is of Late Carboniferous age elsewhere
displaces the TVoms0 Basin down-to-the-west relative in the western Barents Sea, is indirect evidence that
to the Hammerfest Basin. Troms0 was an actively subsiding basin at that time.
iootr !6W o
o

2
FREE AIR GRAVITY MAP ro*oo'H
120Km

20*00* 30W

Fig. 10. GECO free-air gravity map. a

Extensional tectonic history of the western Barents Sea 101

Late Carboniferous-Jurassic sag phase to the West Bj0rn0ya Basin. This is the position of
the Cretaceous hinge line, which for the most part
In the Late Carboniferous active crustal extension is controlled by reactivation of Late Devonian-Early
ceased, leading to a transition into a slowly subsiding Carboniferous basin-bounding normal faults (Figs. 2,
interior sag (Fig. 2). Sedimentation changed from 4-9 and 11; compare Figs. 3 and 11). East of the
continental to shallow-marine siliciclastic deposition Loppa High, prominent basement faults that were
and ultimately to carbonate and evaporite deposition reactivated are the Ttoms-Finnmark fault, the north-
(R0nnevik and Jacobsen, 1984). Salt deposition is west portion of the Berlevg Fault and the southern
interpreted to have occurred in the Dia Graben, bounding faults of Nordkapp Basin South.
r
Oms0, Nordkapp, West Bj0rn0ya and Hammerfest The north-east structural trends inherited from
basins (Fig. 2). Slow subsidence continued through the Late Devonian-Early Carboniferous deforma-
most of the Permian. In Late Permian times the tion predominate across the western Barents Sea.
western portion of the Loppa High was faulted Two exceptions to this are noted in the Hammerfest
down to the west along the Ringvass0y Loppa Fault Basin. In the central Hammerfest, at about 2140'E,
Complex. The Loppa High was subsequently rotated north-west trending faults are superimposed on the
down to the east. pre-existing north-east trending fabric (Fig. 11).
Closing of the Uralian Sea during the Late Per- These faults are found along strike of the Berlevg
mian through Early THassic placed the Barents Shelf Fault. West-trending faults are also superimposed on
in a distal foreland position to the Ural Mountains. earlier structural trends along the north-west bound-
Great volumes of sediment were deposited in Finn- ary of the Hammerfest Basin, forming the boundary
mark East, the Nordkapp Basins and across the Dia between it and the Loppa High (Fig. 11). This fault
Graben and Loppa High. Sediment loading had a system continues to the north-east, but the displace-
two-fold effect. First, there was reactivation of some ment is transferred to a drape flexure located above
basement-involved normal faults, most notably the the Late Devonian-Early Carboniferous margin of
eastern margin of the Dia Graben, which provided a the basin.
structural control for the position of the Early-Mid- Detached normal faults and low-amplitude folds,
Triassic shelf edges (Figs. 2, 3 and 8). Secondly, there east of Loppa High, were partially localized over
was increased withdrawal of salt, primarily in the Dia deeper, pre-existing Late Devonian-Early Carbonif-
Graben and the Nordkapp basins, accompanied by erous faults (Figs. 4-9). The folds are either flexures
the local development of basement-detached normal over normal fault blocks, or low-side roll-over struc-
faults. tures (extensional fault-bend folds; Groshong, 1989).
Salt-cored folds are also present, but no effort was
Late Jurassic-Early Cretaceous rift phase made to differentiate these on the map.
Two possible zones of strike-slip faulting have
Renewed crustal extension between Greenland been identified, each having minor displacements
and Norway initiated a continued and long-term (Fig. 11). A north-east trending divergent(?) strike-
uplift of the Loppa High. This episode of rifting was slip fault is located along the eastern margin of the
most intense during Late Jurassic-Early Cretaceous Fenris Graben. Displacement along this fault appears
(Barremian) time (Fig. 2). Major structures that were to be minimal, yet it was sufficient to form a small
active at this time are shown on the paleo-tectonic depression along its strike. The southern boundary
map in Fig. 11. of the Vads0 High was also reactivated, possibly as a
Regionally, most of the deformation occurred west convergent(?) strike-slip fault.
of the Loppa High. A thick, Hauterivian-Albian
section, sourced primarily from the west, was de- Tertiary passive margin phase
posited in the rapidly subsiding TYoms0, Bj0rn0ya
and Kval0y Basins. These basins can be interpreted Tectonic activity during the Tertiary is mostly re-
to have formed as pull-apart basins in response to lated to the progressive northward opening of the
oblique rifting between the Barents Sea and Green- North Atlantic and Arctic Oceans in Late Pale-
land (Fig. 11). ocene time (magnetic anomaly 24; Figs. 2 and 12).
The Tunheim Terrace developed on the Loppa Opening began near the V0ring Plateau escarpment
High, characterized by a relatively wide zone of tilted and propagated northwards along the Senja fracture
fault blocks underlain by detached normal faults. A zone. By Early to mid-Miocene times the Knipovich
prominent set of faults with multi-directional trends and Mohns Ridges were established as spreading
extends along the western boundary of the Loppa centres (Fig. 12). An excellent review of the Ceno-
High from the TYoms0 Basin, across the Senja Ridge, zoic evolution of this margin has been presented by
102 CA. Dengo andK.G. R0ssland

4"

>

a \\

\
*i$^7
ECONs

BASEMENT-INVOLVE
D NORMA
L FAULTS (INCLUDE
S LATE
DEVONIAN-EARL
Y CARBONIFEROU
S REACTIVATE
D
FAULTS
)
WESTERN BARENTS SEA
ANTICLINAL AXIS, SHOWING DIRECTIO
N OF PLUNG
E LATE JURASSIC - EARLY CRETACEOUS
SYNCLINAL AXIS, SHOWIN G DIRECTIO
N OF PLUNG
E PALEOTECTONIC MAP
MONOCLINE , SHOWIN G DIRECTIO
N OF PLUNGE

<5> LIMIT
S OF SALT
25 50 75 100 125k
m

Fig. 11. Late Jurassic-Early Cretaceous paleo-tectonic map.

Eldholm et al. (1984) and Dowling (1988). Here, the Berlevg and Ttoms-Finnmark faults. East of
we note only that most of the deformation occurred the Loppa High the region was relatively stable.
west of the Loppa High and Senja Ridge, along reac- There was deformation related to salt movement and
tivated, pre-existing faults. Active basement-involved basement-detached normal faulting overlying older
normal faults are the basin-bounding faults of the basement-involved faults.
Harstad, Kval0y, TVoms0, Bj0rn0ya and Nordkapp Subsequent to the initiation of sea floor spreading
Basin South. Some movement is also recorded along (Early Eocene), the western margin of the Barents
Extensional tectonic history of the western Barents Sea 103

*<*>>

*rfr-

**$?

/ o U G O C E H E-. 3 6 M / -

/--~-~. /
-
/

^ . -9, MIDDL
E EOCENE
. 42 Ma
_-j
-" 1
/^
^r^J-^^--^-~\
^
7 /
E A R L V j o a H L - ;

LEGEN D
BASEMENT-INVOLVE
DEVONIAN-EARL
D NORMA L FAULT
Y CARBONIFEROU
S (INCLUDE
S LATE
S REACTIVATE
D WESTERN BARENTS SEA
FAULTS)
BASEMEN T - DETACHE D NORMAL FAULTS
TERTIARY PALEOTECTONIC MAP
EARLY TERTIARY EROSIONA L LIMIT
OCEANIC CRUS T
MAGNETI C ANOMALIE S 0 25 50 75 100 125k
m
LIMIT
S OF SALT

Fig. 12. Tertiary paleo-tectonic map.

Sea developed as a sheared margin (Myhre et al., in Bj0rn0ya Island and the Bj0rn0ya West region.
1982) with as much as 550 km of dextral strike-slip To the east there is little evidence of compressive
displacement on the Hornsund Fault. Compressive deformation in the Barents Sea. From mid-Miocene
deformation associated with this strike-slip faulting time to the present the western Barents Sea has been
led to the development of the Spitsbergen orogenic regionally uplifted.
belt, and to the south led to compressive structures
 s

NORDKAPP BASIN SOUTH SE

NW

Fig. 13. Portion of the deep seismic reflection line, IKU-B, across Nordkapp Basin South. A low-angle normal fault is interpreted to extend beneath the entire basin to the base of the crust. The
line has been squeezed and is vertically exaggerated five times. For the location of the seismic line see Fig. 1.

Extensional tectonic history of the western Barents Sea 105

Discussion Basins across the Senja Ridge). The arrangement of

these half-grabens and highs is similar to that ob-
The principal conclusions derived from this study served in the East African Rift (Rosendahl et al.,
are that: (1) extensional tectonics has been the dom- 1986). Although the types of structural trap devel-
inant mode of deformation in the western Barents oped in these basins should be typical of graben
Sea since Late Devonian-Early Carboniferous time settings (Harding, 1984), the mode of linkage of the
with little demonstrable large-scale strike-slip defor- basement-involved normal faults can modify and or
mation, (2) a fundamental basement architecture of enhance trap closure capacity, as illustrated recently
half-grabens and intervening highs formed in Late for the North Viking Graben (Scott and Rosendahl,
Devonian-Early Carboniferous time and these con- 1989). This is particularly evident where the inter-
trolled the location of younger basins, and (3) defor- basinal highs intersect the graben-flanking highs, for
mation progressed westwards from the time of initial example the intersection of the Nordkapp and Saiika
extension until the time of break-up in the Late Highs. Cross-cutting (multi-directional) trends can
Paleocene. The recognition in the last several years form traps that otherwise would not be expected
that lithospheric extension leads to the development within a full-graben.
of asymmetrical basins (half-grabens) characterized Basin asymmetry will also have an important in-
by tilted fault-blocks (e.g., Wernicke and Burchfiel, fluence on the distribution of the rift-fill clastic
1982; Jackson et al., 1982; Gibbs, 1984; Rosendahl et sediments. Different types of facies relationship
al., 1986) provides a framework with which to view will develop as a function of margin relief and
the extensional deformation of the western Barents slope, fault linkage patterns, drainage directions,
Sea. Equally important, the concepts of asymmetrical etc. (Leeder and Gawthrope, 1987). Thus, predictive
extensional structuring have a direct application in models for reservoir and seal distribution can be
hydrocarbon exploration because they can be used to achieved with knowledge of the basin architecture.
improve models of heat flow and maturation and to This was demonstrated by Gerdes et al. (this volume)
predict trap geometry and reservoir distribution. who describe a case for the Barents Sea where the
The western Barents Sea extensional fabric shares structural analysis of the margin of a half-graben, in
similarities with extensional areas elsewhere. Region- conjunction with other data, was used to construct
ally, the Late Devonian-Early Carboniferous paleo- depositional models for different seismic sequences
tectonic map shows a fabric of subparallel ridges and and from these to determine areas favoured for
basins that are similar in scale and arrangement to reservoir deposition.
the Basin and Range Province of the western United Additional studies are required in order to further
States (see Myers and Hamilton, 1964; Wernicke and the understanding of the deep crustal structure of
Burchfiel, 1982). Low-angle detachments may have the Barents Sea and the mechanism(s) by which
a significant control on the extensional geometries lithospheric extension has occurred whether this
observed in the western Barents Sea, as indicated has been dominantly by lithospheric-scale pure shear
by the deep seismic reflection profiles available. For (McKenzie, 1978), simple shear (Wernicke, 1985) or
example, in the case of the Nordkapp Basin South, some other mechanism. To understand the entire rift
it is possible to interpret a low-angle normal fault evolution fully it will also be necessary to take into
extending beneath the entire basin to lower crustal account the structure of the complimentary margin
levels (Fig. 13). Additional studies may reveal that of north-east Greenland.
regionally extensive basement highs (for example Identification of the mechanism(s) of lithospheric
the Loppa High) may be analogous to the uplifted extension during the various rift phases may help
footwall block associated with low-angle crustal de- in understanding problems such as the cause of the
tachment systems, as described for parts of the Basin Tertiary uplift and erosion in the western Barents
and Range Province (Wernicke and Axen, 1988). It is Sea (Fig. 12). Some possible causes for this uplift
important to recognize, however, that the asymmetry are discussed by Nyland et al., Kjemperud and Fjeld-
observed in the Barents Sea does not have to repre- skaar and Riis and Fjeldskaar (all in this volume).
sent a process that is unique to lithospheric exten- Although glacial rebound accounts for some of the
sional deformation everywhere, but may represent uplift, other processes, operative at the lithospheric
the control imposed by the pre-existing Caledonian scale and probably directly related to the mecha-
structural fabric. nism of lithospheric extension, must account for the
Locally in the western Barents Sea, half-grabens remaining uplift. Thermal subsidence and the time-
are arranged with alternating polarity across inter- integrated anomalous heat flow will differ depending
basinal ridges (e.g., Nordkapp Basins across the on the mode of crustal extension; for example, cal-
Nordkapp High and the Ttoms0 and West Bj0rn0ya culations for a low-angle detachment model indicate
106 CA. Dengo and K. G. R0ssland

that they will be about one-half the values calculated E. Talleraas (Editors), Structural and Tectonic Modelling and
for a model of pure shear extension (Voorhoeve and its Application to Petroleum Geology. Norwegian Petroleum
Society (NPF), Special Publication 1. Elsevier, Amsterdam,
Houseman, 1988). Consequently, accurate matura- pp. 133-152 (this volume).
tion modelling of source rock intervals, hydrocarbon Gibbs, A.D., 1984. Structural evolution of extensional basin
type and yield predictions, and construction of ther- margins. J. Geol. Soc. London, 141: 609-620.
mal subsidence curves for geohistory analysis depend Groshong, R.H., 1989. Half-graben structures: balanced models
directly on the extensional model assumed. It is im- of extensional fault bend folds. Geol. Soc. Am. Bull., 101:
96-195.
portant to continue our studies of the deep crustal
Harding, T.P., 1984. Graben hydrocarbon occurrences and struc-
structure of the western Barents Sea. tural style: Am. Assoc. Pet. Geol. Bull., 68: 333-362.
Harland, W.B., 1969. Contribution of Spitsbergen to under-
Acknowledgments standing of tectonic evolution of North Atlantic Region. Am.
Assoc. Pet. Geol. Mem., 12: 817-851.
We thank Esso Norge a.s., Exxon Company Inter- Harland, W.B., 1985. Caledonide Svalbard. In: D.G. Gee and
B.A. Stuat (Editors), The Caledonide Orogen Scandinavia
national and Exxon Production Research Company and Related Areas. John Wiley and Sons, Chichester, pp.
for permission to publish this work. We also acknowl- 999-1016.
edge the Norwegian Petroleum Directorate, GECO Harland, W.B. and Dowdeswell, E.K. (Editors), 1988. Geological
and IKU for permitting us to publish the geoseismic Evolution of the Barents Shelf Region. Graham and Trotman,
sections, the gravity map and the seismic line IKU-B, London, 176 pp.
respectively. This study benefited from the contribu- Harland, W.B., Cutbill, J.L., Friend, P.F., Gobbett, D.J., Holliday,
D.W., Maton, P.I., Parker, J.R. and Wallis, R.H., 1974. Bille-
tions made by many of our colleagues at Exxon, in fjorden Fault Zone, Spitsbergen. Nor. Polarinst. Skr., 161, 72
particular T. Nardin, B. Wiggins, D. Bergslien, O. pp.
Hostad, S. Ballestad, J. Demarest and S. Greenlee. Harland, W.B., Perkins, P.J. and Smith, M.P., 1988. Cam-
The manuscript was improved by reviews made by brian through Devonian stratigraphy and tectonic develop-
T Hauge and L. Fairchild from Exxon Production ment of the western Barents Shelf. In: W.B. Harland and
E.K. Dowdeswell (Editors), Geological Evolution of the Bar-
Research Co., I. Stewart from BP Norway and an
ents Shelf Region. Graham and Trotman, London, pp. 7 3 -
anonymous reviewer. 88.
Jackson, J.A., King, G. and Vita-Finzi, C , 1982. The neotectonics
of the Aegean: an alternative view. Earth Planet. Sei. Lett.,
References 61: 303-318.
Kjemperud, A. and Fjeldskaar, W., 1992. Pleistocene glacial
Berglund, T., Auguston, J., Faerseth, R., Gjelberg, J. and
isostacy implications for petroleum geology. In: R.M.
Ramberg-Moe, H., 1986. The evolution of the Hammer-
fest Basin. In: Habitat of Hydrocarbons on the Norwegian Larsen, H. Brekke, B.T. Larsen and E. Talleraas (Editors),
Continental Shelf. Norw. Pet. Soc., Graham and Trotman, Structural and Tectonic Modelling and its Application to
London, pp. 319-338. Petroleum Geology. Norwegian Petroleum Society (NPF),
Chadwick, R.A., 1985, Upper Jurassic to Cretaceous sedimen- Special Publication 1. Elsevier, Amsterdam, pp. 187-195 (this
tation and subsidence. In: A. Whittaker (Editor), Atlas of volume).
Onshore Sedimentary Basins in England and Wales: Post- Lamar, D.L., Reed, W.E. and Douglass, D.N., 1986. Billefjorden
Carboniferous Tectonics and Stratigraphy. Blackie, Glasgow, fault zone, Spitsbergen. Is it part of a major Late Devonian
pp. 49-52. transform? Geol. Soc. Am. Bull., 97: 1083-1088.
Cheadle, M.J., McGeary, S., Warner, M.R. and Matthews, D.H., Lake, S.D. and Karner, G.D., 1987. The structure and evolution
1987. Extensional structures on the western U.K. continental of the Wessex basin, southern England: an example of
shelf: a review of evidence from deep seismic profiling. inversion tectonics. Tectonophysics, 137: 347-378.
In: M.P. Coward, J.F. Dewey and PL. Hancock (Editors), Leeder, M.R. and Gawthrope, R.L., 1987. Sedimentary models
Continental Extensional Tectonics. Geol. Soc. London Spec. for extensional tilt-block/half-graben basins. In: M.P. Coward,
Publ., 28: pp. 445-465. J.F. Dewey and P.L. Hancock (Editors), Continental Exten-
Dowling, L.M., 1988. Cenozoic evolution of the western margin sional Tectonics. Geol. Soc. Londen Spec. Publ., 28: pp. 139-
of the Barents Shelf. In: W.B. Harland and E.K. Dowdeswell 152.
(Editors), Geological Evolution of the Barents Shelf Region. McKenzie, D.P., 1978. Some remarks on the development of
Graham and Trotman, London, pp. 157-169. sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32.
Eldholm, O., Sundvor, E., Myhre, A.M. and Faleide, J.I., 1984. Myers, W.B. and Hamilton, W, 1964. Deformation associated
Cenozoic evolution of the margin off Norway and Svalbard: with the Hebgen Lake earthquake of August 17, 1959. U.S.
In: A M . Spencer et al. (Editors), Petroleum Geology of the Geol. Surv. Prof. Pap., 435: 55-98.
North European Margin. Graham and Trotman, London, pp. Myhre, A.M., Eldholm, O. and Sundvor, E., 1982. The margin
3-18. between Senja and Spitsbergen fracture zones: implications
Faleide, J.I., Gudlaugsson, St.T. and Jacquart, G., 1984. Evolu- from plate tectonics. Tectonophysics, 89: 1-32.
tion of the western Barents Sea. Mar. Pet. Geol., 1: 123- Nelson, K.D., Arnow, J.A., McBride, J.H., Willemin, J.H.,
149. Huang, J., Zheng, L., Oliver, J.E., Brown, L.D. and Kaufman,
Gerdes, K., Hurst, J. and Jeans, R., 1992. Seismostratigraphic S., 1985. New COCORP profiling in the southeastern United
study of a synrift megasequence, from the Barents Sea, north- States, Part I: Late Paleozoic suture and Mesozoic rift basin.
ern Norway. In: R.M. Larsen, H. Brekke, B.T Larsen and Geology, 13: 714-717.
Extensional tectonic history of the western Barents Sea 107

Nyland, B., Jensen, L.N., Skagen, J., Skarpness, O. and Vorren, the North European Margin, Norwegian Petroleum Society,
T., 1992. Tertiary uplift and erosion in the Barents Sea; Graham and Trotman, London, pp. 19-32.
magnitude, timing and consequences. In: R.M. Larsen, H. Scott, D.L. and Rosendahl, B.R., 1989. North Viking Graben:
Brekke, B.T. Larsen and E. Talleraas (Editors), Structural and an East African perspective. Am. Assoc. Pet. Geol. Bull., 73:
Tectonic Modelling and its Application to Petroleum Geology. 155-165.
Norwegian Petroleum Society (NPF), Special Publication 1. Smith, M.P., 1988. Cambrian through Devonian Panarctic tec-
Elsevier, Amsterdam, pp. 153-162 (this volume). tonic events. In: W.B. Harland and E.K. Dowdeswell (Ed-
Ratcliffe, N.M. and Burton, W.C., 1985. Fault reactivation mod- itors), Geological Evolution of the Barents Shelf Region.
els for the origin of the Newark basin and studies related to Graham and Trotman, London, pp. 7-17.
eastern U.S. seismicity. In: G.P. Robinson and A.J. Froelich Steel, R.J. and Worsley, D., 1984. Svalbard's post-Caledonian
(Editors), Proc. Sec. U.S.G.S. Workshop on the Early Meso- strata an atlas of sedimentational patterns and pale-
zoic Basins of the Eastern United States. U.S. Geol. Surv. ogeographic evolution. In: A.M. Spencer et al. (Editors),
Circ, 946: pp. 36-45. Petroleum Geology of the North European Margin. Nor. Pet.
Riis, F. and Fjeldskaar, W, 1992. On the magnitude of the Late Soc, Graham and Trotman, London, pp. 109-135.
Tertiary and Quaternary erosion and its significance for the Sturt, B.A., Pringle, I.R. and Ramsay, D.M., 1978. The Finn-
uplift of Scandinavia and the Barents Sea. In: R.M. Larsen, H. markian phase of the Caledonian Orogeny. J. Geol. Soc
Brekke, B.T Larsen and E. Talleraas (Editors), Structural and London, 135: 597-610.
Tectonic Modelling and its Application to Petroleum Geology. Swanson, M.T., 1986. Pre-existing fault control for Mesozoic
Norwegian Petroleum Society (NPF), Special Publication 1. basin formation in eastern North America. Geology, 14: 419-
Elsevier, Amsterdam, pp. 163-185 (this volume). 422.
Roberts, D. and Sturt, B.A., 1980. Caledonian deformation in Townsend, C , 1987. Thrust transport directions and thrust sheet
Norway. J. Geol. Soc. London, 137: 241-150. restoration in the Caledonides of Finnmark, north Norway. J.
Rosendahl, B.R., Reynolds, D J . , Lorben, P.M., Burgess, C.F., Struct. Geol., 9: 345-352.
McGill, J., Scott, D., Lambiase, J.J. and Derksen, S.J., Wernicke, B. and Burchfiel, B.C., 1982. Modes of extensional
1986. Structural expressions of rifting: lessons from Lake tectonics: J. Struct. Geol., 4: 105-115.
Tanganyika, Africa: sedimentation in the African Rifts. Geol. Wernicke, B., 1985. Uniform-sense normal simple shear of the
Soc. London, Spec. Publ., 25: 29-43. continental lithosphere. Can. J. Earth Sei., 22: 108-125.
R0nnevik, H.C., 1981. Geology of the Barents Sea. In: Petroleum Wernicke, B. and Axen, G.J., 1988. On the role of isostacy in
Geology of the Continental Shelf of North West Europe. the evolution of normal fault systems. Geology, 16: 848-851.
Institute of Petroleum, London, pp. 395-406. Voorhoeve, H. and Houseman, G., 1988. The thermal evolution
R0nnevik, H.C., Beskow, B. and Jacobsen, H.P., 1982. Structural of lithosphere extending on a low-angle detachment zone.
and stratigraphic evolution of the Barents Sea. ONS-82, Basin Res., 1: 1-9.
Norw. Pet. Soc, pp. 1-28 Ziegler, P.A., 1978. North western Europe: tectonics and basin
R0nnevik, H.C. and Jacobsen, .., 1984. Structural highs and development. Geol. Mijnbouw, 57: 589-626.
basins in the western Barents Sea. In: Petroleum Geology of

C.A. DENGO Exxon Exploration Company, P.O. Box 4279, Houston, TX77210-4279, U.S.A.
K.G. R0SSLAND Esso NorgeA/S, P.O. Box 60, N-4033Forus, Norway
Present address: Statoil, UND-LS, ST-FHD1, P.O. Box300, N-4001 Stavanger, Norway
 109

Tectonic framework and halokinesis of the Nordkapp Basin,

Barents Sea
L.N. Jensen and K. S0rensen

The Nordkapp Basin (NKB) is a deep Paleozoic basin on the Barents Shelf, initiated in a regional rift episode of Late
Devonian to Carboniferous age. The basin contains numerous major salt structures, but is surrounded by flat-lying platforms
with only minor Post-Paleozoic tectonism.
The Trollfjord-Komagelv Fault Zone (TKFZ) is a fundamental strike-slip zone in northern Finnmark, with a late reactivation
in Devonian to Visean time. The late dextral strike-slip is reported to be less than 5 km. We propose that the contemporaneous
rifting in the Barents Sea and the movements on the TKFZ were related. The rifting occurred north of the TKFZ, with the
Nordkapp Basin as the most prominent rift basin. The extension needed to create these basins was accommodated by dextral
strike-slip reactivation of the TKFZ.
A Late Carboniferous to Asselian age is inferred for the salt in the Nordkapp Basin. In Permian times both the platforms
and the basin were covered by a carbonate sequence of regional extent, followed in the Early Triassic by a thick sequence of
fine-grained clastic sediments. The even thickness of these two sequences demonstrates that no salt movements took place in the
Nordkapp Basin at that time.
The salt structures in the NKB are walls and diapirs, while salt pillows are only seen on the basin margins. The normal
(Trusheim) model for halokinesis, comprising a pillow stage with primary rim synclines followed by a diapiric phase with
secondary rim synclines, cannot be applied to the Nordkapp Basin. Instead, the whole basin evolved as secondary rim synclines
with closely spaced diapirs in Triassic times. Based on volume calculations of the secondary rim synclines and the salt we
estimate an original maximum salt thickness of more than 2 km in the southwestern part of the basin, and more than 4 km in
the northern part. Apparently, such abnormal salt sequences give rise to an unusual halokinetic development which deviates
from the Trusheim model, and an alternative model is proposed.

Introduction Barents Sea have been described in a number of

recent publications (Faleide et al., 1984; Gabrielsen
Since the opening of the Barents Sea for hydrocar- et al., 1984, 1990; R0nnevik and Jacobsen, 1984)
bon exploration the area has attracted much interest and only a brief summary focusing on the Nordkapp
and, although drilling results in the last 2 years have Basin will be given here.
been disappointing, there are still major basins in the In Late Devonian to Early Carboniferous time the
area that have scarcely been explored. The Nordkapp Barents Sea, including Bj0rn0ya and Svalbard, was
Basin is one of these (Fig. 1). affected by a regional rift episode, and a number of
Two exploration wells have been drilled on the N-S to NE-SW trending basins formed. The most
western flank of the Nordkapp Basin, finding sub- prominent of these, the Nordkapp Basin, is 350 km
commercial quantities of gas and oil (Fig. 2). A third long and 30-80 km wide, with a general NE-SW
well has been drilled in the central segment of the trend and a dog-leg shape. The basin is defined
basin, near the western margin (7228/2-1, Fig. 2). at base Cretaceous level (Fig. 1; Gabrielsen et al.,
The oil and gas found in all three wells indicates that 1990). At the Paleozoic level, the southwestern part
hydrocarbons have been generated and migrated in of the basin is a half-graben, while the central part is
the Nordkapp Basin and that seal failure may be the a graben (Figs. 2 and 3). Shallow flat-lying platforms,
major problem in the area. with only minor post-Paleozoic tectonism, surround
The Barents Sea covers an intracratonic basin with the Nordkapp Basin, and the basin is characterized
Devonian to Quaternary sediments over a basement by numerous major salt structures. There are no pub-
of Caledonian and pre-Caledonian age. The geolog- lications solely concerning the basin, but a number of
ical history and tectonic subdivision of the southern recent abstracts (Lind, 1987; Jensen and S0rensen,

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 109-120. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
110 L.N. Jensen andK. S0rensen

31 32 33 34" 35

Fig. 1. The main structural elements in the Barents Sea (from Gabrielsen et al., 1990).

1988; Jensen and Broks, 1988) and theses (Krokan, movements have recently been debated (Rice et al.,
1988; Bergendahl, 1989) have focussed on the basin 1989), and the late movements may be transtensional
history. with a normal fault component (cf. Gabrielsen and
The TYollfjord-Komagelv Fault Zone (TKFZ) is Faerseth, 1989, p. 61).
a fundamental fault zone in northern Finnmark Jenyon (1988, p. 316) has mentioned the possibility
(Fig. 1) which forms part of the tectonic frame- of salt diapirs developing without a preceding pillow
work for the Nordkapp Basin. The Proterozoic to phase and primary rim synclines. In this paper we
Devonian history of the TKFZ is one of large-scale document the existence of such diapirs and show that
dextral strike-slip movements (Kj0de et al., 1978) the whole Nordkapp Basin developed in this way
followed by a minor reactivation in Devonian to in THassic time. The salt in the Nordkapp Basin is
Visean time. The amount and sense of the strike-slip inferred to be of Late Carboniferous to Asselian age.
Tectonic framework and hahkinesis of the Nordkapp Basin, Barents Sea 111

Fig. 2. The main structural elements of the southwestern and central part of the Nordkapp Basin and adjacent platforms. The locations
of wells and seismic lines used in this study are shown.

Halokinesis started in Late Scythian time and was Late movements on the Trollfjord-Komagelv
reactivated in Tertiary time. The salt structures in Fault Zone and rifting in the Nordkapp
the Nordkapp Basin are walls, irregular massifs and Basin
diapirs, all piercing to Quaternary level. Salt pillows
are only seen on the basin margin. The normal Rifting in the Barents Sea
model for halokinesis (Thisheim, 1957), comprising A Late Paleozoic rift episode affected the North
a pillow stage with primary rim synclines followed Atlantic region including the Barents Sea (Hazel-
by a diapiric phase with secondary rim synclines, dine, 1984). In Svalbard this episode is well docu-
cannot be applied to the Nordkapp Basin. Instead, in mented, with Devonian and Carboniferous sedimen-
TYiassic time, the whole basin evolved as a secondary tation in grabens and half-grabens (Steel and Wors-
rim syncline with closely spaced diapirs. ley, 1984; Worsley and Aga, 1986). On Bj0rn0ya this
The aim of this paper is to present and dis- episode started in Late Devonian (Fammenian) time
cuss two aspects of the tectonic evolution of the with sedimentation in a half-graben east of a ma-
Nordkapp Basin: a model for the relation between jor fault (Gjelberg, 1981). The syntectonic sedimen-
Late Paleozoic rifting and late movements on the tary development is similar in both places, starting
Trollfjord-Komagelv Fault Zone and one for the with non-marine coal bearing sediments, followed by
unusual halokinesis in the basin. more marine influenced sequences and ending with
112 L.N. Jensen and K. S0rensen

7226/11-1 (PROJECTED)
"" 7200'
I i

LINE 2 6 1 5 - 8 5

REFLECTORS
1: SEA BOTTOM 2: BASE PLEISTOCENE 3: BASE TERTIARY 4: BASE CRETACEOUS 5: M.JURASSIC
6: N.BASE JURASSIC 7: L A D f i A N 8: N.BASE ANISIAN 9: BASE SMTTHIAN 10: TOP PERMIAN CARBONATES
11: BASE PERMIAN 12,13,14: INTRA CARBONFEROUS 15: BASEMENT

Fig. 3. Profile across the southwestern segment of the Nordkapp Basin. Reflectors are tied to the nearby well, 7226/11-1. The location
of the profile is shown in Fig. 2.

carbonate and evaporite dominated sedimentation in evaporites. The Paleozoic origin of the Harstad Basin
the Late Carboniferous. (Fig. 1) is uncertain, but we assume that the basin
When hydrocarbon exploration in the Barents Sea formed part of a Paleozoic trend through the TYoms0
started, a similar development was expected there Basin and Bj0rn0ya Basin. The salt in the TYoms0
(R0nnevik and Jacobsen, 1984; Faleide et al., 1984) Basin is believed to be of Late Carboniferous to
and this has now been confirmed by seismic mapping Asselian age as in the other Late Paleozoic basins.
and exploration drilling. The Late Paleozoic rift The minor basin on the southwestern margin of the
basins in the southern Barents Sea are mainly half- Loppa High (Fig. 4) has been confirmed by three
grabens, but grabens are also known. The Nordkapp exploration wells, and the stratigraphic development
Basin is the most prominent of these Late Paleozoic in this basin is similar to that on Bj0rn0ya.
rift basins in the Barents Sea, and the southwestern Gabrielsen (1984) discussed the importance of
segment of the basin is used as an example (Figs. 2 the TYoms-Finnmark Fault Zone (Fig. 1), but we
and 3). are not sure about the Late Paleozoic activity or
The Norsel High (Fig. 1) is a paleo high, with the existence of marked Late Paleozoic basins to
Permian and Carboniferous carbonate and evaporite the north of this fault. The fault zone is therefore
strata resting on a crystalline basement (Fig. 3). indicated by a dotted line on Fig. 4.
Southeast of the Norsel High, the Nordkapp Basin The on-shore position of the TYollfjord-Komagelv
is developed as a deep (12-14 km) half-graben with Fault Zone is clear, and has been mapped in detail
more than 6 km of pre-Permian sediments (Fig. 3). (Lippard and Roberts, 1987). Caledonian and pre-
The Permian carbonate sequence between reflector Caledonian activity on the fault zone is believed
10 and 11 (Fig. 3) is used as a reference level. A deep to be large-scale strike-slip movements (Kj0de et
reflector is seen on this line (2615-85, Fig. 3), and its al., 1978). In this paper we will only focus on the
position at top basement level has been confirmed post-Caledonian reactivation of the zone. Based on
by shallow drilling near the Finnmark coast. The age relations between dyke intrusions and activity on
sequences between the basement and the Permian the TKFZ, a Devonian to Visean reactivation with
carbonates are of Late Devonian to Permian age and (?dextral) strike-slip of less than 5 km is reported
we anticipate a stratigraphic development similar to (Kj0de et al., 1978; Gabrielsen and Faerseth, 1989).
that on Svalbard and Bj0rn0ya. The offshore continuation of the TKFZ is less
The known Late Paleozoic basins in the Barents clear, and has been discussed in a number of recent
Sea and their major boundary faults have a general publications. Our interpretation is in agreement with
N-S to NE-SW trend (Fig. 4), and they all contain that of Gabrielsen and Faerseth (1989, their fig. 1)
Tectonic framework and halokinesis of the Nordkapp Basin, Barents Sea 113

1 1 1 1

1
32
BJ0RN0YAir) PALEOZOIC RIFT BASINS H PALEO HIGH
-74

EXTENSION
74
0 ^^ TRANSTENSION
^ STRIKE-SLIP
?
->

y -73

73

-72

72-

'
71-

-70

16

Fig. 4. Paleozoic basins in the southwestern Barents sea. The proposed model is illustrated by arrows indicating areas of regional
extension in late Devonian and Carboniferous time. This rifting was accommodated by dextral strike-slip or transtension on the
Trollfjord-Komagelv Fault Zone.

and is shown in a simplified form in Fig. 4. The ing et al. (1985, their fig. 22f) who described a tear
offshore part of the fault zone is discontinuous and fault accommodating differential extension. An anal-
obscure in places, and the zone seems to die out to ogous example is the Garlock Fault in California,
the northwest near the Loppa High. as described by Davis and Burchfiel (1973, their
fig. 4).
Rifting related to the It is important to stress that the tectonic model
Trollfjord-Komagelv Fault Zone (Fig. 4) is partly based on published data for the
Figure 4 shows our model for the northwestern Ttollfjord-Komagelv Fault Zone, and that these data
termination of the Ttollfjord-Komagelv Fault Zone are under debate (cf. Rice et al., 1989). The model
and its relation to the Late Paleozoic rift basins in proposed will also work if the late movements on
the southwestern Barents Sea. The rifting needed to the TKFZ turn out to be larger (dextral shear sense)
create the Nordkapp Basin and other basins north or if they have a transtensional component. In fact,
of the TKFZ was accommodated as dextral strike- the model will work in all cases where the late
slip on the fault zone. This strike-slip decreases to movements on the TKFZ accommodate differential
the northwest and the fault zone dies out in an extension north and south of the zone.
area of regional extension west of the Loppa High
and the Hammerfest Basin (Fig. 4). This extensional Geologic history and halokinesis of the
termination of a major intracontinental strike-slip Nordkapp Basin
zone could be the reason why the oflshore part of the
zone is so difficult to define. Regional stratigraphic information and shallow
The inspiration for this model is taken from Read- drilling indicate that the Nordkapp Basin was ini-
ing (1980, his fig. 3d), who described the termination tiated in Late Devonian time. We expect the Late
of major intracontinental strike-slip zones, and Hard- Devonian and Early Carboniferous sedimentation in
114 L.N. Jensen and K. S0rensen

the basin to be dominated by clastic sequences, fol- shaped. However, some salt may still exist in the
lowed by carbonate and evaporite sedimentation in source layer and some may have been lost due
Late Carboniferous time. The geographical extent of to dissolution, extrusion and Late Tertiary erosion.
the Late Paleozoic rift basin was larger than that Assuming an overestimation of 25% or 50% and
of the Nordkapp Basin as defined at the base Cre- using the three different depth estimates, we end up
taceous level. The most important difference is in with a total salt volume in the Nordkapp Basin of
the southwestern part, where the Paleozoic basin ex- 22500 to 30000 km3 (Table 1). Of this salt volume,
tends further to the SW and terminates against the approximately 25% is in the southwestern segment
offshore part of the Trollfjord-Komagelv Fault Zone of the basin and 75% in the central segment west of
(Fig. 4). 32 (Tkble 1).
The areas of the Nordkapp Basin at Middle Juras-
Age of the salt in the Nord kapp Basin sic and Base Permian levels have been calculated
Using the Early Permian regional carbonate se- (Table 1). If we redistribute the above volumes of
quence as a reference level, it is possible to demon- salt in the basin using the area at Base Permian level,
strate that the salt in the Nordkapp Basin must be of the original average salt thickness is 1400-1800 m in
Late Carboniferous to Asselian age (Figs. 3 and 5). the southwestern segment and 2300-3100 m in the
Around the central part of the basin, a thin salt se- central segment of the basin (TM>le 1). The base of
quence of similar age is extending onto the platforms the salt is not horizontal. It is clearly faulted in the
30-60 km beyond the basin margins. We assume central part of the basin, and forms a half-graben in
that the lateral equivalent of the salt in the basin the southwestern part (Fig. 3). An original maximum
consists of interbedded sequences of carbonates and salt thickness of 2000 m in the SW part of the basin
anhydrite, as found in wells 7124/3-1 and 7226/11-1. and 4000-5000 m in the central part (cf. Jensen and
South of well 7124/3-1 (Fig. 2), seismic lines in- S0rensen, 1988) is therefore reasonable. This has
dicate the existence of deep salt pillows near the also been confirmed by two-dimensional calculations
basin boundary fault and the offshore part of the of the excess rim syncline volume.
TKFZ soles out in these salt pillows. In an early Bergendahl (1989) has confirmed these estimates
interpretation of the Nordkapp Basin, Faleide et al. for the SW part of the Nordkapp Basin, using both
(1984, their fig. 13) indicated the existence of two methods for volume calculations. Based on gravi-
salt layers. Based on our work, we see no need for metric modelling, Krokan (1988) also indicated an
two salt layers to explain the features seen in the original maximum salt thickness of 4000-5000 m in
Nordkapp Basin, and it is hard to imagine that an the central part of the basin, but he gave no salt
early salt layer could have survived the pre-salt tec- volume estimate.
tonic episode seen in Fig. 5. This tectonic episode is
anticipated to be of Early Carboniferous age.
TABLE 1
The amount of salt in the Nordkapp Basin Areal content, salt volume and original salt thickness in the
In order to estimate the original salt thickness in Nordkapp Basin west of 32
the southwestern and central part of the Nordkapp Areal content (km 2 )
Basin, volume calculations have been carried out.
Stratigraphic level: Middle Jurrasic Base Permian
Two methods were used; direct calculation of the
Southwestern segment 5507 4244
present salt volume and indirect calculation based on
Central segment 8271 7120
excess volume in the secondary rim synclines. These Total (W of 32) 13778 11364
methods are described by Seni and Jackson (1983)
and S0rensen (1986). Salt volume (km 3 )

Due to Late Tertiary erosion all diapirs in the Base salt depth estimate: A B C
Nordkapp Basin pierce to Base Quaternary level, Southwestern segment 8700 9200 9700
and have their maximum horizontal cross-sectional Central segment 25100 26400 27700
area at a level somewhere between Upper THassic Total (W of 32) 33800 35600 37400

and Base Quaternary. The depth to the base of the 25% overestimated 27700 28500 30000
50% overestimated 22500 23700 25000
diapirs is listed using three different depth estimates
(A,B and C in Table 1). As a first approximation, Original salt thickness 3 (m)
the present salt volume was calculated assuming Southwestern segment: 1400- 1800 m
vertical sides for the diapirs (total volume, Table 1). Central segment: 2300- 3100 m
This is an overestimation, because some diapirs have a
Thickness range for depth estimate A to C and 25% to 50%
a spherical top and some are clearly mushroom overestimated salt volume.
Tectonicframeworkand halokinesis of the Nordkapp Basin, Barents Sea 115

W E
B400 SP 8600 8800 9000 9200 9400 9600 9800
e>**

1
1 \ v^\ -1

=f=~
r r r \ < \ -, 3
r r r \ \>
r\ N ~f5
r r r\
2
r r r r
6 - 2 *
to

L
r r r rl 7
r r r r /
- ^ ^ - - / -
r r r /
3 r/ ^ ^ r: r 3
I
=*= = 3/
r^:__^^-"^
^- =
10
*""
/ ~"~~"~

---/ 7
r r r7\^^
4~ r r r (^^
g_ ^^^\/~~/ ^4

LINE 7225-- 8 5 0 km 6
REFLECTORS
1: BASE PLEISTOCENE 2: BASE TERTIARY 3: BASE CRETACECXIS 4: MJURASSC
5: (SLBASE JURASSIC 6: BASE AMSIAN 7: BASE SMTHAN 8: TOP PERMIAN CARBON/*TES
9: BASE PERMIAN 10: BASE SALT 11: VISEAN??

Fig. 5. Line drawing of a seismic section across the eastern margin of the Nordkapp Basin (line 7225-85, Fig. 2). From the known age
of the overlying carbonate sequence, a Late Carboniferous to Asselian age is inferred for the salt. A pre-salt tectonic episode is clearly
seen on this line.

Post-Carboniferous development lows are explained by Gabrielsen et al. (this volume).

of the Nord kapp Basin Near the salt diapir in Fig. 6, minor erosion of the
The Post-Carboniferous development of the Nord- Early Scythian sequence is seen, indicating that a
kapp Basin will be illustrated using a reconstruction salt pillow may have existed. All diapirs in the basin
of a typical seismic line (GNB-87-108) from the seems to have been initiated at the same time. The
central part of the basin (Fig. 6). We use a recon- diapirism was strongest from Late Scythian to Late
struction in two-way time instead of depth, to allow TYiassic time, and enormous secondary rim synclines
tie and sequence identification from our line. formed. Stratigraphic traps may have formed in these
Before halokinesis started in Late Scythian time rim synclines (Fig. 6). The reflectors in the secondary
the salt in the Nordkapp Basin was covered by a re- rim synclines diverge towards the diapirs and there
gional Early Permian carbonate sequence, Late Per- is no evidence of syn-diapiric sediment progradation
mian shale and chert sequences and Early Scythian into the basin.
shales (Fig. 6). The even thicknesses of these se- In Late TViassic time, the diapirism stopped, and
quences on the platforms and in the basin indicate the diapirs were covered by sediments (Fig. 6). There
that no differential subsidence or salt movement took was still some growth faulting at the basin margin,
place at that time. and the salt pillow also grew a little (Fig. 6). The
In the nearest wells on the Bjarmeland Plat- diapirism probably stopped because the gravitational
form, the Permian carbonate sequence is 540-800 m instability had disappeared and the salt source layer
thick, the Late Permian sequence is 220-410 m and had been exhausted.
the Early Scythian sequence is 820-1010 m. These
are present-day non-decompacted thicknesses. It is, Post-Triassic history
therefore, reasonable to assume that before haloki- The Jurassic sequences covered the diapirs and
nesis started, the salt in the Nordkapp Basin was there was only minor differential subsidence of the
covered by 1600-2200 m of sediments, comprising Nordkapp Basin at that time (Fig. 6). A regional
a rigid "lid" of carbonates and a rapidly deposited tectonic episode affected the Barents Sea in Late
shaly sequence. When 2000-4000 m of salt were Jurassic to Early Cretaceous time and evidence of
covered by such a sequence, a strong gravitational this episode is seen in and near the Nordkapp Basin.
instability must have existed immediately prior to the But the faults are small, and the episode had little
onset of diapirism. influence in the Nordkapp Basin area.
A 1000-1500 m thick Cretaceous sequence cov-
Triassic diapirism ered the area. There are no indications of differential
Diapirism commenced in Late Scythian time, and subsidence or salt movement in the Nordkapp Basin
the whole Nordkapp Basin developed as one major at that time (Fig. 6). From well information, we
rim syncline (Fig. 6). At the same time salt pillows know that 1000-1500 m of Early Tertiary sediments
and growth faults formed at the basin margin (Figs. covered the area (Fig. 6) and the weight of these
5 and 6). The detailed development of these salt pil- post-Ladinian sequences probably recreated a gravi-
116 L.N. Jensen and K. S0rensen

W
SPieo TODAY GNB 87 108 SP200

- 0

|

r r r r r rt-2 -1000-1500M OF NEOGENE

r r r r r " UPLIFT AND EROSION
r r r r r r
r r r r r rj 3 ~ - MINOR SALT MOVEMENT ? ->
r r r r r r
r r r r r rt K
r r r r r r 4

OLIGOCENE(30m.y)

^ e ^
EARLY TERTIARY:
- 1 0 0 0 - 1 5 0 0 M OF PALEOGENE
SEDIMENTS
- REACTIVATION OF DIAPIRS ANO
THE NORDKAPP BASIN
- MAXIMUM SUBSIDENCE
- HC GENERATION AND MIGRATION

BASE TERTIARY(66m.y)

/ / / /
r 1 - LATE JURASSIC TECTONISM

r r r r r r> - NO DIAPIRISM
r r r r r r
r r r r r r| - NO DIFFERENTIAL SUBSIDENCE

r r r r r rl OF THE NORDKAPP BASIN
r r r r r r
r r r r r 1

M.JURASSIC (165m.y)

r r r r r r| - MINOR DIAPIRISM

| - MINOR DIFFERENTIAL SUBSIDENCE


OF THE NORDKAPP BASIN

ANISIAN(240m.y)

1 $ - DIAPIRISM AND RAPID SUBSIDENCE

r OF SECONDARY RIM SYNCLINES
DIAPIR *
r r r r r r rt ^ - SALT PILLOWS AT BASIN MARGIN
h
- GROWTH FAULTING

SMITHIAN(245m.y)
- RAPID SUBSIDENCE
.' ' "i
* - CLASTIC SEDIMENTS
r ^ SALT MOVEMENT, BUT NO
\ \ ~ -^>lliL-r
^ r r r
r r cr
^ PRIMARY RIM SYNCLINES

LATE PERMIAN(250m.y)
6 - EARLY CARBONIFEROUS RIFTING
- LATE CARBONIFEROUS SALT (2-4000M)
1
- LATE CARBONIFEROUS LATE PERMIAN
REGIONAL CARBONATE PLATFORM
2 h-
i
km 10

Fig. 6. Trie geological development and halokinesis of the Nord kapp Basin illustrated by a reconstruction of seismic line GNB-87-108.
The location of the line is shown in Fig. 2. Further details are explained in the text.
Tectonicframeworkand halokinesis of the Nordkapp Basin, Barents Sea 117

tational instability in the Nordkapp Basin. A regional Bishop (1978) later criticized this buoyancy model
compressive episode affected the area in Early Ter- and contributed to the understanding of piercement
tiary time, and the combined effect of gravitational diapirism. Jenyon (1988, p. 316) mentioned a pos-
instability and compression caused some reactivation sible situation in which salt diapirism may occur
of the diapirs in the Nordkapp Basin. Clear evidence without a pillow phase and the formation of primary
of this reactivation is seen in the southwestern part rim synclines. Such a situation would be an exception
of the basin, where Early Tertiary sediments are to the general TYusheim model (Fig. 7). We believe
preserved. that the whole Nordkapp Basin developed in this way
In Late Tertiary time a regional episode of uplift in TYiassic times (Fig. 8).
and erosion affected the Barents Sea (Nyland et al., The driving force for the diapirism in the Nord-
this volume). At the location of the reconstructed kapp Basin must have been a strong gravitational
line, the erosion was of the order of 1200-1500 m, instability, but the triggering mechanism is less clear.
and all the Early Tertiary sediments were removed Near the diapir in Fig. 6, there is a minor erosion of
(Fig. 6). The Neogene erosion probably removed the Early Scythian sequence, and a short-wavelength
some of the salt. The area then subsided and was pillow or "bulge" may have existed (Fig. 6). This
covered by 100-200 m of Pleistocene sediments and model is shown in Fig. 8 (diapir C). But this phe-
300-400 m of water (Fig. 6). nomenon is only seen at a few diapirs in the Nord-
kapp Basin and there are no indications of related
A halokinetic model for the Nordkapp Basin primary rim synclines.
TYusheim (1957) introduced the term halokinesis A tectonic episode may have been the triggering
and presented a general model for gravity-driven mechanism for the diapirs. A minor growth faulting
salt tectonics (Fig. 7). His model is now widely ac- episode of Late Scythian age is known from the
cepted and seems to apply to many of the worlds Bjarmeland Platform, and it probably also affected
salt basins (Halbouty, 1979; Seni and Jackson, 1983; the Nordkapp Basin. Because of the strong gravita-
Jackson and TMbot, 1986; Jaritz, 1987). Trusheim's tional instability in the Nordkapp Basin at the onset
model comprises a salt pillow phase with related of this extension, salt may have been injected into
primary rim synclines, followed by a diapiric phase the faults and developed directly into a piercement
with formation of secondary rim synclines (Fig. 7). (diapirs A and B, Fig. 8). Jackson and TMbot (1986,
their fig. 5f) and Jenyon (1988) both mention such a
mechanism.
On the Finnmark Platform, prograding sequences
of Early Scythian age are seen from the Finnmark
coast to the margin of the Nordkapp Basin. These
sequences are 200-500 m thick and prograde towards
the north and northwest. The progradational pattern
has not been identified in the Nordkapp Basin, but
the influx of sediments from the Finnmark Platform
may have contributed to the initiation of diapirism.
Such a triggering mechanism has been suggested by
Bishop (1978, fig. 3), Halbouty (1979) and Woodbury
et al. (1980, fig. 3).
Each of the above mechanisms or a combination
of them, may have triggered the unusual diapirism
in the Nordkapp Basin. Unfortunately the intruding
diapirs and later erosion together have destroyed
the evidence for these triggering mechanisms, and a
diapir "frozen" in the initial phase does not exist in
the Nordkapp Basin.
In the absence of clear evidence for the triggering
mechanism within the Nordkapp Basin an analogous
example can help. We have searched for published
analogues from salt basins around the world, but
Fig. 7. The Trusheim (1957) model for halokinesis. The salt-
tectonic development of the Nordkapp Basin clearly deviates have been able to find only two. The Louann salt in
from this model (see Figs. 6 and 8). PRS = primary rim the Gulf of Mexico is an analogue, but only with re-
syncline; SRS = secondary rim syncline. spect to salt thickness and the closely spaced diapirs
118 L.N. Jensen and K. S0rensen

+*++++++++**. *+++++++++++
SRS /fTTTv - ~~A r r r r r r f\
/ I
/

____ ^.
]

! SRS

rr r r r r r
|

r r r r r r r rt [
rrrrrrrrf
I

3
rrrr r r r

^ ^ ^^ I I II I III
^^ -^frrrrrrr
^/ .

LADINIAN

E.TRIASSIC

I I I I I I I
r_T_j
?^/*^
r r r r r r r rsr r j > f f
r r r r r r r r (V ^
L " - ^\<
"
^

DIENERIAN

Fig. 8. A halokinetic model for the Nordkapp Basin. Details are explained in the text.

on the outer shelf (Seni and Jackson, 1983, fig. 1). on the TYollfjord-Komagelv Fault Zone. This ge-
This is because the Louann salt appears to have netic relationship may also explain why the offshore
moved laterally under the weight of the prograding continuation of the fault zone is so difficult to define.
Tertiary sediments (Halbouly, 1979; Woodbury et al., (2) The TYollfjord-Komagelv Fault Zone dies out
1980). The salt in the Nordkapp Basin, however, is in an area of regional extension west of the Loppa
confined to a rift basin and has not moved laterally as High.
is the case for the Louann salt. The Gulf of Mexico (3) The salt in the Nordkapp Basin is of Late
is, therefore, not an exact analogy. Carboniferous to Asselian age, and only one salt
The second analogous example is the North layer is needed to explain the features seen in the
Caspian Depression in the U.S.S.R. Only a few basin.
translated publications exist for this basin. According (4) An original salt volume of 22500-30000 km3
to Kapustin et al. (1981), 2-3.5 km of Kungurian was deposited in the Nordkapp Basin. Based on
salt were deposited in this rift basin. A profile pub- this volume we estimate an original maximum salt
lished by Avrov et al. (1986) may indicate that the thickness of more than 2 km in the southwestern
closely spaced diapirs in the North Caspian Depres- part and more than 4 km in the northern part of the
sion developed in the same way as those in the basin.
Nordkapp Basin. The North Caspian Depression is (5) During TYiassic times the whole Nordkapp
a prolific oil-producing basin and may be the only Basin developed as a number of secondary rim syn-
exact analogy to the Nordkapp Basin. clines related to closely spaced diapirs. Evidence
of a preceding salt pillow phase with primary rim
Conclusions synclines has not been found. The Nordkapp Basin
is, therefore, a rare exception to Trusheim's (1957)
The tectonic and halokinetic models for the Nord- halokinetic model.
kapp Basin can be summarized as follows: The halokinetic model for the Nordkapp Basin
(1) There may be a genetic relation between the has implications for the hydrocarbon exploration
Late Paleozoic rift basins in the Barents Sea and in the basin. It is important to realize that we
the contemporaneous reactivation of the TYollfjord- cannot expect to find a number of hydrocarbon
Komagelv Fault Zone. The extension needed to cre- trap types related to salt pillows and primary rim
ate these basins was accommodated as strike-slip synclines (cf. Halbouty, 1979, fig. 6-1) because this
Tectonicframeworkand halokinesis of the Nordkapp Basin, Barents Sea 119

phase of halokinetic development is absent in the Harding, T.P, Vierbuchen, R.C. and Christie-Blick, H.N., 1985.
Nordkapp Basin. Moreover, the Late Tertiary erosion Structural styles, plate-tectonic settings, and hydrocarbon
traps of divergent (transtensional) wrench faults. In: K.T
has removed the possibility of finding hydrocarbon Biddle and H.N. Christie-Blick (Editors), Strike-slip De-
traps over the crest of the diapirs. We are, therefore, formation, Basin Formation and Sedimentation. Soc. Econ.
left with traps related to the flanks of the diapirs and Paleontol. Mineral., Spec. Publ., pp. 52-77.
the secondary rim synclines. A number of such traps Hazeldine, R.S., 1984. Carboniferous North Atlantic paleogeog-
have already been mapped. raphy: stratigraphic evidence for rifting, not megashear or
subduction. Geol. Mag., 121: 443-463.
Jackson, M.P.A. and Talbot, C.J., 1986. External shapes, strain
Acknowledgements rates and dynamics of salt structures. Geol. Soc. Am. Bull.,
97: 305-323.
We thank the management of Statoil Exploration Jaritz, W, 1987. The origin and development of salt structures in
Northwest Germany. In: I. Lerche and J J . O'Brien (Editors),
for permission to publish this paper and we acknowl- Dynamical Geology of Salt and Related Structures. Academic
edge the helpful comments from Anthony Spencer, Press, New York, N.Y., pp. 479-493.
Elin Kjemtrup and two anonymous referees. Jensen, L.N. and S0rensen, K., 1988. Geology and salt tectonics
of the Nordkapp Basin, Barents Sea. Abstract, 18. Nord.
Geol. Vinterm0de, K0benhavn, 1988, Danm. Geol. Unders.,
References pp. 187-188.
Jensen, L.N. and Broks, T.M., 1988. Late movements on the
Avrov, V.P., Balgimbayev, M.B., Krylov, N.A. and Sagingaliyev, Trollfjord-Komagelv Fault Zone and rifting in the Nordkapp
A.B., 1986. Possibilities for discovery of new oil fields in sub- Basin. VI Annu. TSGS Meet., Inst. Geol., Univ. Oslo, Intern
salt carbonate sediments of the North Caspian Depression. Skriftser., 54, pp. 24-25 (abstract).
Translation in: Pet. Geol., 22(9): 391-399. Jenyon, M.K., 1988. Some deformation effects in a clastic
Bergendahl, E., 1989. Halokinetisk utvikling av Nordkappbas- overburden resulting from salt mobility. J. Pet. Geol., 11:
sengets s0rvestre segment. Unpublished Cand. Scient. Thesis, 309-324.
University of Oslo, 120 pp. Kapustin, I.N., Kiryukhin, L.G., Payrazyan, V.V. and Ras-
Bishop, R.S., 1978. Mechanism for emplacement of piercement myshlyayev, A.A., 1981. Geologic history of the North Caspian
Diapirs. Am. Assoc. Pet. Geol. Bull., 62: 1561-1583. Depression and time of formation zones of oil-gas accumula-
Davis, G.A. and Burchfiel, B.C., 1973. Garlock Fault: an in- tion. Translation in: Pet. Geol., 22(10): 483-485.
tracontinental transform structure, Southern California. Geol. Kj0de, J. Storetvedt, K.M. Roberts, D. and Gidskehaug, A.,
Soc. Am. Bull., 84: 1407-1422. 1978. Paleomagnetic evidence for large-scale dextral move-
Faleide, J.I., Gudlaugsson, S.T. and Jacquart, G., 1984. Evolution ment along the Trollfjord-Komagelv Fault, Finnmark, north
of the Western Barents Sea. Mar. Pet. Geol., 1: 123-150. Norway. Phys. Earth Planet. Inter., 16: 132-144.
Gabrielsen, R.H., Faerseth, R., Hamar, G.P. and R0nnevik, H., Krokan, B., 1988. Et gravimetrisk Studium av Nordkappbassen-
1984. Nomenclature of the main structural features on the get. Cand. Sei. Thesis, Univeristy of Oslo, 174 pp (unpub-
Norwegian Continental Shelf north of the 62nd parallel. In: lished).
A.M. Spencer et al. (Editors), Petroleum Geology of the Lind, E., 1987. The Nordkapp Basin, a major salt feature in the
North European Margin. Graham and Trotman, London, pp. SW Barents Sea. 4th Annu. Meet., TSGS Group, Nor. Geol.
41-60. Tidsskr., 67: 435 (abstract).
Gabrielsen, R.H., Faerseth, R., Jensen, L.N., Kaiheim, J.E. and Lippard, S.J. and Roberts, D., 1987. Fault systems in Caledonian
Riis, F, 1990. Structural elements of Norwegian Continental Finnmark and the southern Barents Sea. Nor. Geol. Unders.
Shelf. Part I: The Barents Sea region. Norwegian Petroleum Bull., 410: 55-64.
Directorate, Bull., 6, 33 pp. Nyland, B., Jensen, L.N., Skagen, J., Skarpnes, O. and Vorren,
Gabrielsen, R.H., 1984. Long-lived fault zones and their in- T, 1992. Tertiary uplift and erosion in the Barents Sea;
fluence on the tectonic development of the Southwestern magnitude, timing and consequences. In: R.M. Larsen, H.
Barents Sea. J. Geol. Soc. London, 141: 651-662. Brekke, B.T. Larsen and E. Talleraas (Editors), Structural and
Gabrielsen, R.H. and Faerseth, R.B., 1989. The inner shelf of Tectonic Modelling and its Application to Petroleum Geology.
North Cape, Norway and its implication for the Barents Shelf- Norwegian Petroleum Society (NPF), Special Publication 1.
Finnmark Caledonide boundary. A comment. Nor. Geol. Elsevier, Amsterdam, pp. 153-162 (this volume).
Tidsskr., 69: 57-62. Reading, H.G., 1980. Characteristics and recognition of strike-
Gabrielsen, R.H., Kl0vjan, O.S., Rasmussen, A. and St0lan, slip fault systems. Spec. Publ. Int. Assoc. Sedimentol., 4:
T, 1992. Interaction between halokinesis and faulting: struc- 7-26.
turing of the margins of the Nordkapp Basin, Barents Sea Rice, A.H.N., Fearn, P. and Townsend, C , 1989. Fault patterns
region. In: R.M. Larsen, H. Brekke, B.T Larsen and E. in Northeast Finnmark and its implications for the basement
Talleraas (Editors), Structural and Tectonic Modelling and structure of the Southern Barents Sea. Paper presented
its Application to Petroleum Geology. Norwegian Petroleum at Norwegian Petroleum Society (NPF) Workshop, 18-20
Society (NPF), Special Publication 1. Elsevier, Amsterdam, October 1989, Stavanger.
pp. 121-131 (this volume). R0nnevik, II. and Jacobsen, H.P., 1984. Structural highs and
Gjelberg, J.G., 1981. Upper Devonian (Famennian)-Middle Car- basins in the western Barents Sea. In: A.M. Spencer et al.
boniferous succession of Bj0rn0ya. Nor. Polarinst. Skr., 174, (Editors), Petroleum Geology of the North European Margin.
67 pp. Graham and Trotman, London, pp. 98-107.
Halbouty, M.T., 1979. Salt Domes, Gulf Region, United States Seni, S.J. and Jackson, P.A., 1983. Evolution of salt structures,
and Mexico, Gulf Publishing Co., Houston, Texas, 2nd ed., east Texas diapir province, Parts 1 and 2. Am. Assoc. Pet.
561 pp. Geol. Bull., 67: 1219-1244; 1245-1274.
120 L.N. Jensen andK. S0rensen

Steel, R.J. and Worsley, D., 1984. Svalbard's post-Caledonian Trusheim, F., 1957. ber Halokinese und ihre Bedeutung fr
strata an atlas of sedimentational patterns and pale- die strukturelle Entwicklung Norddeutschlands. Z. Deutsch.
ogeographic evolution. In: A.M. Spencer et al. (Editors), Geol. Ges., 109: 111-151.
Petroleum Geology of the North European Margin. Graham Woodbury, H.O., Murray, I.B. and Osborne, R.E., 1980. Diapirs
and Trotman, London, pp. 109-135. and their relation to hydrocarbon accumulation. In: A.D.
S0rensen, K., 1986. Rim syncline volume estimation and salt Miall (Editor), Facts and Principles of World Petroleum
diapirism. Nature, 319: 23-27. Occurrence. Canadian Society of Petroleum Geologists, pp.
Worsley, D. and Aga, O., 1986. The Geological History of 119-142.
Svalbard. Statoil, 121 pp.

L.N. JENSEN Statoil AIS, P.O. Box 300, N-4001 Stavanger, Norway
K. S0RENSEN Statoil A IS, St. Annce Plads 13, DK-1298 K0benhavn K, Denmark
 121

Interaction between halokinesis and faulting: structuring of

the margins of the Nordkapp Basin, Barents Sea region

R.H. Gabrielsen, O.S. Kl0vjan, A. Rasmussen and T. St0lan

Salt is found in three alternative structural settings in the southwestern Barents Sea region: (1) as domes and pillows in the
platform areas, (2) as pillows, stocks and walls in some basins, and (3) as pillows along some graben margins.
In the Nordkapp Basin there is clear evidence that halokinesis has interacted with faulting during structuring of the graben
margins. Furthermore, periods with salt pillow growth are associated with condensed sedimentary sequences in the footwalls
and expanded sequences in the hanging walls relative to the deep faults with which the pillows are associated. In general, this
suggests flow of salt towards the basin margins.
Some salt pillows are divided into subparallel elongated structures where they overlie deep faults. The crestal parts of the salt
pillows are frequently faulted, and in some cases characterized by reverse fault geometries. The following sequence of events
is proposed to explain this configuration: (1) salt flow directed towards basin margins; (2) continued deep faulting below the
salt pillows associated with continued basin subsidence and basinward rotation of the basin margin areas; and (3) reactivation of
crestal faults under differential compaction.

Introduction The tectonic setting and history of the

Nordkapp Basin
Since the early days of exploration in the Barents
Sea continental shelf, the presence of large amounts The Nordkapp Basin is an ENE-WSW-trending
of evaporites in some of the basins of this area has structure located between 713(yN, 25 and
attracted attention (Hinz and Weber, 1975; R0nnevik 7330'N, 34. It is more than 100 km long and
et al., 1975; Hinz and Schlter, 1978). From the has a width varying between 30 and 80 km. A kink
early works it soon became clear that there is much in the basin margin at about 7230'N, 2830 gives
contrast between the different basins as regards the the structure the shape of a dog-leg. The basin is
amount of evaporites and, consequently, tectonic bordered by the Nysleppen, Mas0y and Thor Iversen
styles. The major salt deposition seems to have been Fault Complexes separating it from the Finnmark
concentrated in the TVoms0 and Nordkapp Basin Platform to the southeast, and the Bjarmeland Plat-
areas (0vreb0 and TMleraas, 1976, 1977; R0nnevik, form to the northeast. Although the early (i.e., pre-
1981; R0nnevik et al., 1982; Faleide et al, 1984), late Carboniferous) history of the basin is poorly
although several basins of Late Palaeozoic age in the known, it is likely that the major fault trends of
Barents Sea continental shelf are known to contain the area were established by late Devonian to early
evaporites (Jensen and S0rensen, this volume). Carboniferous times (R0nnevik et al., 1982).
Salt is found in three different settings in the The southeasternmost part of the Nordkapp
southern Barents Sea: Basin is delineated to the northwest by the Norsel
- domes, pillows and sheets in platform areas High, which is a part of the Bjarmeland Platform
- pillows, stocks and walls in the basins (Gabrielsen et al., 1990). Defined on the pre-Permian
- pillows along the graben margins. levels, the southwestern part of the Nordkapp Basin
The present study concentrates upon a set of is a half-graben, whereas sections across its central
pillow-like salt structures which are found along and northern parts show a more symmetrical struc-
the northwestern and southeastern margins of the ture.
Nordkapp Basin, i.e., along the Nysleppen and the A NW-SE-oriented cross-section through the
Mas0y Fault Complexes (Figs. 1 and 2). graben shows that the Nordkapp Basin has a pro-
nounced axial development of large salt stocks,

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 121-131. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
122 R.H. Gabriehen, OS. Kl0vjan, A. Rasmussen and T. St0lan

75H

70

Fig. 1. Structural elements of the Barents Sea. From Gabrielsen et al. (1990). 1 = Nysleppen Fault Complex; 2 = Mas0y Fault
Complex; 3 = Thor Iversen Fault Complex. The location of the lines (Figs. 3-6 and 8) are indicated in the inset.

whereas the basin margin zones are characterized by large-scale differential subsidence took place in the
extensional faults associated with salt pillows (Fig. 2). Nordkapp Basin at this point in time, but there are
Based upon volume calculations, the original thick- indications that salt movements accelerated by mid-
ness of the salt has been estimated to about 2000 Triassic time, followed by decreasing subsidence in
metres in the southern, and up to 5000 metres in the the late THassic. Tectonic episodes in late Jurassic to
northern part of the basin (Jensen and S0rensen, this early Cretaceous and Tertiary times reactivated the
volume; Bergendahl, 1989). basin. The latter episodes also caused reactivation of
The evaporites of the Nordkapp Basin are believed the salt structures.
to be of late Carboniferous to early Permian (As-
selian) age, being deposited in a broad epeirogenic Master fault complexes and associated
basin (e.g., R0nnevik and Jacobsen, 1984). Early salt pillows
works suggested that the salt structures were rooted
at two different levels (Faleide et al., 1984), but The Nysleppen Fault Complex
new data do not seem to confirm this (Jensen* and The Nysleppen Fault Complex (Gabrielsen et al,
S0rensen, this volume). 1990) delineates the Nordkapp Basin against the
The relatively uniform thickness of the Permian Bjarmeland Platform, and locally against the Norsel
and lower TYiassic sequences demonstrates that no High. Regionally, the fault complex trends NE-SW
Interaction between halokinesis and faulting: structuring of the margins of the Nordkapp Basin, Barents Sea region 123

Fig. 2. Regional NW-SE profile through the Nordkapp Basin. Based on Gabrielsen et al. (1990). Same location as Fig. 8, see Fig. 1.

to ENE-WSW, although a N - S to NNE-SSW strik- In the area in which the Nysleppen Fault Complex
ing segment northeast of the Norsel High is found follows a northerly trend, the fault zone is associ-
(Fig. 1). The structure can be traced northeastwards ated with a major salt pillow. At base salt level,
to approximately 30 in the form of a flexure two large faults are identified (Fig. 3), whereas the
defined at intra-lower Cretaceous level. It should fault pattern at the shallower levels is dominated by
be noted that the Nysleppen Fault Complex lies one fault. At the position at which the deep faults
along the continuation of the TYoms-Finnmark Fault transect the base of the salt, the salt may be di-
(Fig. 1), and may represent a segment of a zone of vided into two or even three almost isolated smaller
weakness which has wider regional tectonic implica- pillows. Along its strike, the faulted salt pillow asso-
tions. ciated with the Nysleppen Fault Complex accordingly
As defined at Palaeozoic levels, the Nysleppen includes all the characteristic styles described by
Fault Complex consists of one or more subparallel Jenyon (1986a, b), namely those structures in which
normal faults with an accumulated throw in the or- the faults affect only the top of the salt, those in
der of 500 milliseconds (TWT) or more. At the lower which only the base of the salt is faulted and those
Cretaceous level, the faulted zone is less broad, and in which the whole salt sequence is influenced by
the throws are concentrated into one single, or two faulting.
closely spaced faults. The accumulated throw is con- In contrast to a section to the south, which has
siderably less at lower Cretaceous than at Palaeozoic a much smaller salt structure in the footwall, the
levels. Because of contrasts in deformational style, footwall above the salt in the northern part of the
the different number of faults and the lack of fault salt pillow is strongly rotated towards the platform
plane correlation between the deeper and shallower side (compare Figs. 3 and 4). It can be demonstrated
levels, it is natural to propose that an extensional that a gradual rotation towards the platform takes
detachment is present in the lower Cretaceous se- place as the salt thickness gradually increases. Fur-
quence (Fig. 3). thermore, in areas where the salt is thin or absent,
124 R.H. Gabriehen, OS. Kl0vjan, A. Rasmussen and T. St0lan

Fig. 3. NW-SE-oriented section through the Nysleppen Fault Complex (see Fig. 1 for location). Note the faults on the base salt level
and the rotation of the footwall towards the platform side above the salt. The colour code is as in Fig. 4.

Legend

4 km

Fig. 4. E-W-oriented section through the Nysleppen Fault Complex (see Fig. 1 for location), in an area where the salt pillow is small.
Note the lack of rotation of the sequence above the salt pillow.
Interaction between halokinesis and faulting: structuring of the margins of the Nordkapp Basin, Barents Sea region 125

Fig. 5. E-W-oriented section through the Mas0y Fault Complex (see Fig. 1 for location). Note the brittle deformation of the carbonate
sequence (green) immediately above the salt layer. The colour code is as in Fig. 4.

Fig. 6. E-W-oriented section through the Mas0y Fault Complex. Note apparent reverse fault at the crest of the salt pillow. The colour
code is as in Fig. 4.
126 R.H. Gabriehen, O.S. Kl0vjan, A. Rasmussen and T. St0lan

there seems to be a tendency for no detachment to sition of the salt, it is reasonable to assume that
occur between the shallower and the deep faults. a basin was initiated as a half-graben in Late De-
These two observations suggest that the strong vonian to Early Carboniferous times (R0nnevik et
rotation of the footwall fault block, as well as de- al., 1982; Gabrielsen et al, 1990), possibly cover-
velopment of the fault detachment surface in the ing a wider area than the present Nordkapp Basin
Cretaceous part of the sequence in the hanging wall, (Jensen and S0rensen, this volume). Shortly after
are governed by halokinetic processes. deposition of the carbonates a slight subsidence in
Locally, the fault planes which are defined at the present Nordkapp Basin area can be recorded
Mesozoic and Cenozoic levels are complex and ir- (Scythian). The timing of this is very close to the
regular and are characterized by splays cutting into initiation of the halokinesis in the central part of the
the footwall as well as into the hanging walls. In such basin. From this point in time subsidence prevailed in
areas apparent folds and locally strongly tilted fault the Nordkapp Basin and lasted well into the Creta-
blocks are also observed. This indicates a late, local ceous (end Albian). However, the subsidence pattern
inversion along the Nysleppen Fault Complex. varied considerably within the basin, probably de-
pending upon the local salt movements. According
The Msoy Fault Complex to Jensen and S0rensen (this volume), the salt em-
This fault defines the southeastern margin of piercement peaked in Middle Triassic, and slowed
the Nordkapp Basin between 7127'N, 2440 and down considerably towards the end of the Triassic.
7235'N, 2840 (Gabrielsen et al., 1990). The main The salt was then reactivated in Tertiary time. This
general picture was confirmed in our analysis, even
trend of the structure is ENE-WSW, but the north-
though it may be added that slight, symmetrical sub-
ernmost branch turns in a more northerly direction.
sidence took place in the basin in areas which were
Except for the opposite sense of throw, the geometry
less affected by halokinesis, also during the Early and
of the Mas0y Fault Complex is generally similar to
Middle Jurassic.
that of the Nysleppen Fault Complex. However, as
the southwestern segment of the Nordkapp Basin As seen from slight thinning of the Triassic se-
is developed as a half-graben at pre-Permian levels, quences along the basin margins where salt is present
with the larger subsidence along its northwestern (e.g., in the Mas0y Fault Complex; Fig. 5), salt move-
margin, the throws recorded along the Mas0y Fault ments were locally initiated along the graben margins
Complex are smaller. simultaneously with the activation of the major salt
Mas0y Fault Complex is also associated with salt stocks in the central part of the Nordkapp Basin.
pillows with their long axes oriented parallel or sub- This effect is illustrated in Fig. 8 displaying two
parallel to the basin margin (Figs. 1 and 5). It should depth-converted and restored sections across one of
be noted that even here a set of deep faults divide the major salt pillows within the Mas0y Fault Com-
the major salt pillow into two separate structures plex (Fig. 1). The sections are flattened on the lower
in a manner similar to that of the Nysleppen Fault THassic and base upper Jurassic levels, respectively,
Complex. Furthermore, the layers situated at the top and demonstrate that uplift above the crest of the
of the salt are broken into straight segments, some of salt pillow was significant in both these time inter-
which are separated by faults with reverse geometries vals. Since large-scale, secondary horizontal flow in
(Fig. 6). the sediments of the sequences above the salt is con-
sidered to be very unlikely, it is implied that the flow
of salt towards the basin margin was initiated early
Timing of tectonic events in the in the basin history, and that halokinesis affected the
Nordkapp Basin surface topography. The thinning of some units at
the crestal points of the pillows relative to that on
NW-SE and N-S-oriented key lines through the the platform indicate that salt has not only flowed
Nordkapp Basin have been depth-converted, back- towards the basin margin from the basin, where it
stripped and restored with the use of the program competed with the central stocks and walls, but also
ECHO/PAL (Welldrill 1988). Decompaction was not from the platform areas.
performed. The assumptions and constraints affili- Thinning of the Scythian sequence associated with
ated with this method are reviewed by Gabrielsen et salt pillows is not seen everywhere (e.g., in parts
al. (1986). The general results for the timing of the of the Nysleppen Fault Complex; Fig. 3). On the
tectonic events of the basin area are summarized in contrary, a pronounced thinning of the Ladinian to
Fig. 7. Middle Jurassic sequences, if recorded above this
Although the present interpretation gives very lit- salt structure, suggests that flow in the pillows took
tle information on the basin history prior to depo- place at different points in time, and that the salt in
Interaction between halokinesis and faulting: structuring of the margins of the Nordkapp Basin, Barents Sea region 127

MS0Y FAULT]
BREKKE & RIIS 1987 NYSLEPPEN NORDKAPP
COMPLEX;
FAULT COMPL BASIN; NORTH
|TECT| ACTIVE NORTHERN
PHASE
PROCESS AREA |(N-S)SEGMENT|

Deep
erosion
Uplift

Intrusions

i)
Tromso ?

Basin
? MILD INVERSION?
Doming
Faulting
Thrusting
& faulting Harstad
Basin &
West of
Senja R.
V7.
Block
faulting

Transg.
Uplift
Transg.

Transg.
of basem.
Loppa
High area m 1 m
his
Block
faulting
DEPOSITION OF EVAPORITES

Transg.
of basem.
his
Tromso
Basin
Salt ? LEGEND
Block
^ Halokinesis
faulting
Subsidence
Inversion

Fig. 7. Timing of the main tectonic events in the Nordkapp Basin, the Nysleppen and Ms0y Fault Complex. The table is based upon
results from back-stripping and restoration of one regional seismic line of which parts are shown in Fig. 8 and the lines are shown in
Figs. 3-6. For locations see Fig. 1. Data from Brekke and Riis (1987; ii) and Jensen and S0rensen (this volume; /) have been used in
addition to the present data. The time scale is from Harland et al. (1982).

the marginal pillows moved in several stages as basin ders of the footwall fault blocks frequently stand out
subsidence proceeded. as the most elevated points as defined on the base
As already pointed out in the descriptions of the of the Upper Jurassic and shallower levels. As seen
Nysleppen and Mas0y Fault Complexes, the shoul- in Fig. 5, the strata on both sides of the high area
128 R.H. Gabriehen, O.S. Kl0vjan, A. Rasmussen and X St0lan

m
^ "7^>' 2500
\ AA A* A \
^ ^ x x^ /N^/ N/^ / ^x / x^/ A^ A^ / w ^

^ ^ /
^^
^ \ AAj

5000

Fig. 8. (a) Restored, depth-converted section at the lower Triassic levels across the southern margin of the Nordkapp Basin. Note slight
thinning above salt pillow. Shaded sequence, Base Dienerian-Base Smithian; hatched sequence, salt. The section is not decompacted.
The depth scale is in metres. For location see Fig. 1. (b) Restored, depth-converted section at the pre-Late Jurassic levels across the
southern margin of the Nordkapp Basin. Note thinning in the footwall above the salt pillow. Base Upper Jurassic-Base Ladinian;
hatched sequence, salt. The depth scale is in metres. The section is not decompacted. For location see Fig. 1.

laps onto this reflector, suggesting that this point some cases faults at these levels are characterized by
was elevated also in the Late Jurassic and Early reverse geometries.
Cretaceous. It is well documented that salt commonly tends to
Due to the heavy halokinetic effects, to decipher flow towards fault zones (Jenyon 1986a, b, 1988). As
the regional stress situation at any time from the fault zones are associated frequently with differences
present data from Nordkapp Basin and the adjacent in sediment thicknesses, e.g., along basin margins, it
fault complexes would not be trivial. Nevertheless, is reasonable to ascribe this coincidence to primary
we do suggest that the development of the Nordkapp lateral thickness, or lateral density contrasts. This is
Basin took place under mild extension. There are, indeed the case for the Nysleppen and Mas0y Fault
however, indications of Tertiary regional inversion, Complexes.
as also mentioned by Jensen and S0rensen (this It is noted that halokinesis may be triggered at
volume). shallow burial depths and salt has been reported to
flow under as little as 150-200 metres of overbur-
A model for the structuring of the basin den where overlain by dense rocks such as anhy-
shoulders drites and carbonates (Brink, 1974). Furthermore,
low-temperature creep in salt is shown to be com-
The following observations should be taken into mon under the presence of water (Richter-Bernburg,
account when developing a model of the salt struc- 1970; TMbot and Jarvis, 1984; Tklbot and Jackson,
tures along the margins of the Nordkapp Basin: 1987).
- the pillows are found along the basin margins With this in mind, and based on the observations
and have their long axes are oriented subparallel to of the geometric relations of the salt pillows asso-
the margins; ciated with the Nysleppen and Mas0y Fault Com-
- the pillows are detached from the salt in the plexes, we propose the following model for the de-
central parts of the basin; velopment of the pillows.
- the pillows are frequently seen to be divided Flow of salt towards the margins of the Nord-
into separate elongated structures where deep faults kapp Basin started shortly after deposition of the
affect the base of the salt; carbonates (Scythian). A high water content may
- the top of the salt may be faulted and in have facilitated the flow and, most likely, the driving
Interaction between halokinesis and faulting: structuring of the margins of the Nordkapp Basin, Barents Sea region 129

tion of salt or a combination of these) took place.

At this stage continued activity was seen along the
deep faults below the salt pillows. This process was
probably followed by rotation towards the basin.
This would cause thinning of the salt pillows directly
above the rotational axis and subdivision of the salt
into two separate structures (Fig. 9b).

Stage 3. The local accumulation of salt, together

with the regional subsidence of the basin floor, would
accentuate the contrasts in sediment thickness be-
tween the basin and the platform areas and, hence,
differential compaction would increase with a com-
paction maximum in the basin area. This may have
triggered the breakthrough in the central areas of the
salt pillows, reactivating the pre-existing crestal faults
as salt flowed from the highly pressurized basin areas
towards the low-pressure platforms. It is expected
that this would have lifted the crestal parts of the
pillows now situated on the platform margins. Dur-
ing continued burial, salt may have escaped through
the crestal faults developing a reverse fault geometry
(Fig. 9c).
A plastiline model can be used to illustrate the
main stages in this development. The plate of choco-
late in Fig. 10a illustrates the break-up of the stiff
Fig. 9. Schematic model for development of salt pillows along carbonate lid, simultaneously with separation of the
the margins of the Nordkapp Basin. (1) Stiff layer on top of
the flowing salt responds with brittle deformation and crestal
salt pillow into two structures above the fault at
collapse. Note extensional separation along the faults. (2) Ro- the base of the salt. In Fig. 10b plastiline has been
tation towards the basin axis, thinning above the rotational axis squeezed out through the faults at the crest of the
and subdivision of the salt into two separate structures. (3) pillow during compaction following continued sedi-
Break-through of salt in the crestal areas of the salt pillow,
following burial and escape of salt through the crestal faults.
mentation.
It might be suggested that the pronounced up-
lift of the footwall above the salt is associated with
mechanism was differential loading along the graben fault-block rotation of "domino-type" or "bookshelf-
margin. It is, however, not known whether active type" fault blocks (Thompson, 1960; Mandl, 1986;
faulting took place along the master fault systems or Yielding, 1990) decoupling in the salt layer or, alter-
whether the basin margin was developed as a flexure natively, uplift according to the "cantilever model"
above a deeper fault zone. (Roberts and Yielding, 1991). However, the book-
shelf model in its basic form requires a system of ro-
The development of the salt structures and their tating blocks (Wernicke and Burchfiel, 1982; Sclater
associated sedimentary sequences along the basin and Celerier, 1989; Sclater and Shorey, 1989), which
margins may be illustrated by a three-stage model is not found in the present area. Even more serious,
(Fig. 9). the uplift is exclusively associated with the salt pil-
lows (Fig. 4), and is not observed in the absence of
Stage 1 is the initial pillow stage. Figure 9a shows salt pillows (Fig. 5). It is felt that these two observa-
how the stiff layer of carbonates capping the flowing tions rule out both the alternative models for uplift
salt may have responded by brittle deformation and of the basin shoulders.
crestal collapse. This process would be accompanied
by extension of the brittle top layer and accordingly Conclusions
extensional separation along the faults.
The salt accumulations along the margins of the
Stage 2. As salt continued to flow towards the Nordkapp Basin demonstrate that salt flow towards
graben margins, subsidence in the central parts of the faulted areas has taken place. It is suggested
the basin (either due to active stretching, retrac- that the margin-directed flow of salt started early
130 R.H. Gabriehen, O.S. Kl0vjan, A. Rasmussen and T. St0lan

Fig. 10. Plastiline models which illustrate the main stages in the development of the salt pillow associated with the Mas0y Fault
Complex, (a) Break-up of the stiff cap lid due to doming, (b) Plastiline is squeezed out through the faults at the crest of the pillow
simultaneously with the separation of the pillow into two structures above the "fault" at the base of the plastiline. The effects in (b)
happen during vertical compression, which mimics compaction.

in the history of the basin and that it has contin- text, and Dr. C. Needham, Fina Exploration and an
ued periodically throughout most of its subsequent anonymous referee whose comments and suggestions
development. have been of great help.
Generally, the marginal salt accumulations are
elongated pillows, locally subdivided into several,
References
subparallel structures, and associated with deep
faults. Rotation toward the platform of the sedimen- Bergendahl, E., 1989. Halokinetisk utvikling av Nordkappbas-
tary sequences above the salt pillows in the footwalls sengets s0rvestre segment. Cand. Scient. Thesis, University of
is common, and the amount of rotation can be di- Oslo, 120 pp. (unpublished).
rectly correlated with salt thickness. Accordingly, the Brekke, H. and Riis, F., 1987. Tectonics and basin evolution of
the Norwegian shelf between 62 and 72. Nor. Geol. Tidsskr.,
elevation of the basin margins is not associated with
67: 295-322.
rotational fault blocks nor with "cantilever" effects. Brink, A.H., 1974. Petroleum geology of the Gabon Basin. Am.
Some pillows (notably those of the southern basin Assoc. Pet. Geol. Bull., 58: 216-235.
margin) have crestal faults with reverse geometries. Faleide, J.I., Gudlaugsson, S.T. and Jacquart, G., 1984. Evolution
These structures can be explained by a simple three- of the western Barents Sea. Mar. Pet. Geol., 1: 123-150.
stage model involving: (1) salt flow directed towards Gabrielsen, R.H., Ekern, O.F. and Edvardsen, A., 1986. Struc-
tural development of hydrocarbon traps, Block 2/2, Norway.
faults along the basin margins; (2) continued deep In: A.M. Spencer et al. (Editors), Habitat of Hydrocarbons
faulting below the salt pillows associated with con- on the Norwegian Continental Margin. Norw. Petrol. Soc,
tinued basin subsidence and basinward rotation; and Graham and Trotman, London, pp. 129-141.
(3) reactivation of crestal faults under differential Gabrielsen, R.H., Faerseth, R.B., Jensen, L.N., Kaiheim, J.E. and
compaction. Riis, F , 1990. Structural elements, Barents Sea, Norwegian
Shelf. Norwegian Petroleum Directorate Bull., 6: 33 pp.
The fault complexes generally show evidence of Harland, W.B., Cox, A.V., Llewellyn, P.G., Pickton, C.A.G.,
reactivation and, locally, of inversion. Even though Smith, A.G. and Walters, R., 1982. A Geologic Time Scale.
reverse fault geometries are closely associated with Cambridge University Press, Cambridge, 131 pp.
deformation of crestal areas of salt pillows, other Hinz, K. and Schlter, II.-U., 1978. The geological structure of
geometries are not. It is therefore recognized that the western Barents Sea. Mar. Geol., 26: 199-230.
Hinz, K. and Weber, J., 1975. Zum geologischen Aufbau
some inversion structures are difficult to explain
des Norwegischen Kontinentalrandes und der Barents-See
by halokinesis, and that these may indicate a late nach reilexionsseismische Messungen. Compendium 75/75,
phase of reactivation in the Nordkapp Basin area Ergnzungsband der Zeitschrift Erdl und Kohle, Erdgas,
associated with regional tectonic events. Petrochem., 3-29.
Jensen, L.N. and S0rensen, K., 1988. Geology and salt tecton-
ics of the Nordkapp Basin, Barents Sea (abstract). In: K.
Acknowledgements Binzer, I. Marcussen and P. Konradi (Editors), 18. Nordiske
Geologiske Vinterm0de, Abstracts. Danm. Geol. Unders.,
The authors are indebted to Norsk Hydro a.s., Copenhagen, pp. 187-188.
Jensen, L.N. and S0rensen, K., 1992. Tectonic framework and
Harstad for support and permission to publish this halokinesis of the Nordkapp Basin, Barents Sea. In: R.M.
paper. We would also like to thank Dr. Ron J. Larsen, H. Brekke, B.T. Larsen and E. Talleraas (Editors),
Steel, Norsk Hydro a.s. who corrected the English Structural and Tectonic Modelling and its Application to
Interaction between halokinesis and faulting: structuring of the margins of the Nordkapp Basin, Barents Sea region 131

Petroleum Geology. Norwegian Petroleum Society (NPF), Norwegian Petroleum Society, Graham and Trotman, London,
Special Publication 1. Elsevier, Amsterdam, pp. 109-120 (this pp. 98-107.
volume). R0nnevik, H.C., 1981. Geology of the Barents Sea. In: L.V.
Jenyon, M.K., 1986a. Salt Tectonics. Elsevier, Barking, 191 pp. Illing and G.D. Hobson (Editors), Petroleum Geology of
Jenyon, M.K., 1986b. Some consequences of faulting in the the Continental Shelf of Northwest Europe. Institute of
presence of a salt rock interval. J. Pet. Geol., 9: 29-52. Petroleum, London, pp. 395-406.
Jenyon, M.K., 1988. Some deformation effects in a clastic R0nnevik, H.C., Bergsaker, E.I., Moe, A., 0vreb0, O.,
overburden resulting from salt mobility. J. Pet. Geol., 11: Navrestad, T and Stangenes, J., 1975. The geology of the
309-324. Norwegian Continental Shelf. In: A.W. Woodland (Editor),
Lind, E., 1986. The Nordkapp Basin, a major salt feature in Petroleum and the Continental Shelf of North-West Europe,
the SW Barents Sea. Abstract Volume, 4th Annu. Meet., Vol. 1. Institute of Petroleum, London, pp. 117-129.
Tectonics and Structural Geology Studies Group, Geological R0nnevik, H.C., Beskow, B. and Jacobsen, .., 1982. Structural
Society of Norway, Tectonics and Deposition in Sedimentary and stratigraphic evolution of the Barents Sea. Can. Soc. Pet.
Basins, Stavanger, p. 23 (abstract). Geol. Mem., 8: 431-440.
Mandl, G., 1986. Tectonic deformation by rotating parallel faults: Sclater, J.G. and Celerier, B., 1989. Errors in extension mea-
the "bookshelf" mechanism. Tectonophysics, 141: 277-316. surements from planar faults observed on seismic reflection
0vreb0, O. and Talleraas, E., 1976. The structural geology of the lines. Basin Res., 1: 217-221.
Troms area. In: Exploration Geology and Geophysics. Off- Sclater, J.G. and Shorey, M.D., 1989. Mid-Jurassic through mid-
shore North Sea Conf., Article G/TV-6, Norwegian Petroleum Cretaceous extension in the Central Graben of the North
Society. Sea part 2: estimates from faulting observed on a seismic
0vreb0, O. and Talleraas, E., 1977. The structural geology of reflection line. Basin Res., 1: 201-215.
the Troms area, Barents Sea. Geol. J., 1: 47-54. Richter- Talbot, C.J. and Jackson, P.A., 1987. Internal kinematics in salt
Bernburg, G., 1970. Post-depositional structures in salt. Proc. diapirs. Am. Assoc. Pet. Geol. Bull., 71: 1068-1093.
Symp. Geol. Technol. Gulf Coast Salt, Louisiana State Univ., Talbot, C J . and Jarvis, R.J., 1984. Age, budget and dynamics of
Baton Rouge, La., pp. 145-146. an active salt extrusion in Iran. J. Struct. Geol., 6: 521-533.
Roberts, A. and Yielding, G., 1991. Deformation around basin- Thompson, G.A., 1960. Problem of late Cenozoic structure of
margin faults in the North Sea/mid Norway Rift. In: A. the Basin Ranges. Proc. 21st Int. Geol. Congr., Copenhagen,
Roberts, G. Yielding and B. Freeman (Editors), The Geom- 18: 62-68.
etry of Normal Faults. Geol. Soc. London, Spec. Publ., 56: Wernicke, B. and Burchfiel, B.C., 1982. Modes of extensional
61-78 tectonics. J. Struct. Geol., 4: 105-115.
R0nnevik, H. and Jacobsen, H.P., 1984. Structural highs and Yielding, G., 1990. Footwall uplift associated with Late Jurassic
basins in the western Barents Sea. In: A.M. Spencer et al. normal faulting in the northern North Sea. J. Geol. Soc.
(Editors), Petroleum Geology of the North European Margin. London, 147: 219-222.

R.H. GABRIELSEN Geological Institute, Department A, University of Bergen, Allegaten 41, 5007 Bergen, Norway
Present address: Norsk Hydro AIS, Forskningssenteret, N-5020 Bergen, Norway
O.S. KL0VJAN Norsk Hydro AIS, Harstad, P.O. Box 31, 9401 Harstad, Norway
A. RASMUSSEN Norsk Hydro AIS, Harstad, P.O. Box 31, 9401 Harstad, Norway
T. ST0LAN Norsk Hydro AIS, Harstad, P.O. Box 31, 9401 Harstad, Norway
 133

Seismostratigraphic study of a synrift megasequence, from

the Barents Sea# northern Norway
K. Gerdes, J. Hurst and R. Jeans

This paper discusses the development and content of a package of late Jurassic-early Cretaceous synrift sediments from the
Bj0rn0ya Basin. The discussion is based on the result of a detailed structural and seismostratigraphic interpretation of seismic
data, limited well data from the area and models for synrift sedimentation developed from extensive field studies in well-exposed
rift zones.
Structural maps were constructed at two scales to elucidate the first order (regional) and second order (local) structural
controls on deposition. The approximate water depths and inclination of depositional surfaces were estimated using backstripped
sections tied to well data. These techniques produced an estimate of the basin architecture during deposition of the synrift
megasequence.
The synrift megasequence was then subdivided into five seismic sequences. This subdivision depended largely on a high
resolution seismic survey of limited areal extent. Each sequence was then subdivided into seismic facies units and isochron and
facies maps were constructed for each sequence. A correlation between seismic facies and lithology was made using the only
available well in the area. The resultant maps were then interpreted using information on the controls on synrift sedimentation
gleaned from extensive fieldwork in East Africa, Greece and the Gulf of Suez and well-constrained subsurface examples.
Depositional models for each sequence were generated and areas most likely to collect reservoir prone sediments were
highlighted. The main critical factors concerning the predictive nature of the study are the quality of the tie between lithology
and seismic character and the assumption of a coarse grained sediment supply.

Introduction to these data the site survey shot prior to drilling

of the first well was of sufficiently high resolution
This paper deals with the development and con-
to clearly image the external geometry and internal
tent of the late Jurassic-early Cretaceous synrift
character of the synrift megasequence and allowed a
megasequence, found in the Bj0rn0ya South area of
seismostratigraphic subdivision of the synrift megase-
the Barents Sea, offshore northern Norway, (Fig. 1).
quence. The area of the case study block covered by
The aims of the study were twofold.
these dip aligned surveys is termed the 'detailed
(1) To develop an integrated analysis of the inter-
study area' in this paper (Fig. 2).
play between extensional faulting and the develop-
A new detailed structural map for the whole
ment of the synrift megasequence sediment wedges.
case study block was also constructed using these
(2) To document the seismic facies within the data. The well results, new structural map and
synrift megasequence and infer from their extent seismostratigraphic subdivision of the synrift sec-
and distribution the likely presence and variation in
tion form the database of this study. The principles
reservoir facies.
and concepts developed during a research project on
The study presented in this paper was undertaken synrift sedimentation were used to interpret these
as part of the evaluation of the hydrocarbon prospec- data and generate the depositional models presented
tivity of a Barents Sea licence block (Fig. 1). The in this paper.
primary exploration target in this area was a pre-rift
play. The synrift megasequence was recognised as a
secondary target. The synrift sediment packages were Regional setting
poorly imaged on the east-west aligned seismic data
available prior to the block being licensed. Once the The regional setting of the Barents Sea and its
licence was awarded, BP's partners made available sequence stratigraphy has been described in detail in
dip aligned surveys of higher resolution which they the literature (R0nnevik and Jacobsen, 1983; Faleide
had previously acquired over the block. In addition et al., 1984) and only a brief summary will be given

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 133-152. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
 s

BASIN
REGIONA
L TECTONIC
S MEGASEQUENC
E GEOCHRONOLOG
Y

PRERIF
T
UPPE
R
COARSE
CLASTIC
S

<ir
zo
On-

100km
13 CASE STUDY
I
Si

BLOCK
Fig. 1. Regional location map and Barents Sea sequence stratigraphy.
Seismostratigraphic study of a synrift megasequence,fromthe Barents Sea, northern Norway 135

19 20' 19 40'
72 35'

72 00'
Fig. 2. Case study block schematic regional structural map with area of detailed study highlighted.

here. During the late Jurassic to early Cretaceous section from parts of the Bj0rn0ya Basin. This allows
period, an extensional regime developed in the Bar- the synrift megasequence to be imaged on the flanks
ents Sea area in response to the initial opening of of the basin at relatively shallow depths.
the Atlantic Ocean south of the Charlie Gibbs Frac- The case study block consists of a series of fault
ture Zone. The resultant multiphase rifting led to block terraces developed during the Mesozoic on
the development of the Troms0 and Bj0rn0ya Basins the eastern flank of the Bj0rn0ya Basin. Tb the
(Fig. 1). A series of fault block terraces developed southeast of these terraces lies the Mesozoic rift
on the flanks of the Bj0rn0ya Basin upon which a flank of the Loppa Ridge, to the northwest of the
synrift megasequence of variable thickness was de- terraces the fault blocks are buried beneath large
posited. These sediments were then buried beneath thickness (5 seconds +) of post-rift sediments. The
a great thickness of post-rift deposits, during the detailed study area covers a segment of one of these
mid-Cretaceous to early Tertiary. Recent uplift and fault terraces, within which the synrift megasequence
erosion has removed between 1.0-2.0 km of post-rift is particularly well developed and imaged. The well
136 K. Gerdes, J. Hurst and R. Jeans

drilled in the block was situated on the crest of this The well was situated on the crest of the tilted
tilted fault block terrace. fault block terrace and indicated that even at the
highest structural point on the fault terraces, the syn-
Database rift sediments were fully marine. The interpretation
of these data, therefore, relies heavily on information
Approximately 2720 km of seismic data were avail- gleaned from the research and development project
able for the study. These data were at a horizontal concerned with synrift depositional models. Hence,
scale of 1:25 000 and a vertical scale of 10 cm: 1 s the final models incorporate the results of observa-
and are summarised in Table 1. The seismic database tions made at a seismic scale during extensive field
is shown in Fig. 3. studies of well exposed rift systems from around the
The well in the study area was drilled to test a world.
deeper, pre-rift target and penetrated only the distal A number of features intrinsic to the database
part of the synrift megasequence wedge (Fig. 4). No control the degree of confidence which can be placed
other wells have been drilled in the licence block. on the resultant seismostratigraphic interpretation.

TABLE 1

Summary of seismic database

Survey Year Line Line Spacing Alignment Stacked Migrated

(km) (km)
Western 3 1984 700 4.5 N/S-E/W V V
Geco a 1984 680 4.5 N/S-E/W V
Fin a 1986 360 2.5 (6) dip (strike) V
Mobil 1985 250 4 (8.15) dip (strike) V
Geoteam 1987 180 0.3 (0.5) dip (strike) V
Regional pre-1984 550 10 various V V
a
Western and Geco surveys shot to interleave and give an overall line spacing of 2.25.

H 72 30'

172 10'

Fig. 3. Detailed study area seismic database map.

 a) b)
A SEQUENCE SEISMIC SEQUENCE CORRELATION
THICKNESS USED IN THIS STUDY
km

NW SE
280 270 260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100
i ^ L ^ l ^ J ^ U . *k*U

58m

v. mino
r
sandstones

Synrift
Megasequence
yellow-brn/brn-blk
carbonaceou
s
20m mudston
e
3

2
36m
dusky yellow brn
bioturbated silty
mudston e
loc. glauconiti
c 1
at base

Cored section
SEISMIC LINE 409

Fig. 4. (a) Seismic/well character tie. (b) Well results with seismic correlation.
138 K. Gerdes, J. Hurst and R. Jeans

A summary of the main limiting features of the majority of the lines in this survey only cover the
database is given below. thinnest half of the synrift wedge. Hence, the data
(1) Using well data the parallel continuous re- density available to tie across the thicker sections
flector character observed on the seismic data was of the wedge nearest the major bounding fault is
correlated with mudstones indicative of a low energy substantially less than that indicated by the average
marine environment over the fault block crest (see line spacings and survey specifications.
Fig. 4). The megasequence interval in the well was
mud dominated. The presence of coarse grained,
sand-prone deposits within the synrift sequences al- Methodology
luded to in this study, is based on the assumption
that the change in seismic character observed, from The main aim of this study was to investigate
parallel continuous events to more disturbed events, whether the expanded seismic database could be
can be correlated with a change in lithofacies from used to highlight areas of the synrift megasequence
low energy mudstones to coarser grained sand-prone which were likely to be sand-prone. This problem
deposits. was addressed using two independent lines of in-
(2) Data from six surveys was used for the seis- vestigation: the expanded seismic database was used
mic interpretation (see Fig. 3). Three surveys were to construct a more detailed structural map of the
aligned approximately along the dip to the main entire case study block (Fig. 2) to determine whether
structural trend, one of which was a site survey the synrift structure could highlight areas likely to
of high resolution. The other surveys were aligned focus sand-prone sediments. Secondly, in the de-
north-south, east-west. Correlating from survey to tailed study area the synrift megasequence was sub-
survey, in the absence of consistent reprocessing of divided into sequences and seismic facies units and
all the datasets, introduced a variable amount of these were used in conjunction with seismic charac-
uncertainty. ter/facies correlation from the nearby well, to infer
A short sensitivity test illustrated that the error in areas which might be sand-prone.
fault positioning between the east-west aligned data The seismic interpretation was completed on sec-
and the dip-aligned data was as much as 1 kilometre. tions with a horizontal scale of 1:25 000 and a
The convention used throughout the mapping was vertical scale of 10 cm: 1 s. Structure maps were
that, in the event of a positioning discrepancy, the constructed at a scale of 1:100000 and 1:50000 for
position on the dip-aligned survey was given the most the case study block (Fig. 2). Two-way time maps to
weight. the top and base of the synrift megasequence were
(3) The intrinsic variability between the surveys then constructed in the area where the two bound-
meant that the correlation of seismic character from aries had been correlated (Fig. 6). The interpretation
survey to survey was made with caution. The site concentrated on faults which cut the top and basal
survey data offered the highest resolution by far and, reflectors of the megasequence sediment wedges. No
hence, these were the data predominantly used for attempt was made to correlate fault trends beyond
the purpose of seismic sequence definition. the eastern seismic subcrop of the megasequence
(4) The disposition of the seismic lines and faults where the unit is thin or absent.
in the area prevented the tying of synrift reflectors on The synrift wedge on the main fault block of
the main fault block to the wedge shaped sequences interest was then divided into five seismic sequences
on the higher fault blocks. The same problem meant bounded by unconformities using seismostratigraphic
that these reflectors could not be directly tied across principles. The seismic picks are shown in Fig. 11.
the lateral fault on the main fault block (see Fig. 5). Each unconformity was a prominent onlap surface.
The exact position of this lateral fault is obscured, The subdivision of the megasequence was restricted
on the few lines that cross it, by an area of disturbed to the main fault block.
data probably caused by gas escape. Hence, there Isopach maps were then constructed for each seis-
is no independent method available for estimating mic sequence (Figs. 12, 13, 15 and 17). The seismic
the timing and amount of throw on this fault at character within each sequence was then annotated
particular periods during the development of the along each survey line at a scale of 1:25000. The
synrift sequences. amplitude, frequency and continuity of the seismic
(5) Although the site survey was of highest reso- events within each sequence were described and fa-
lution and covered the area of main interest, it was cies areas were drawn onto the isopach maps. Within
not acquired with the present type of study in mind. each facies area the seismic character was broadly
The strike lines are concentrated in the area where similar, making a qualitative allowance for variations
the synrift megasequence is thin or absent and the in data quality between surveys.
Seismostratigraphic study of a synrift megasequence, from the Barents Sea, northern Norway 139

SEISMIC UNE 301

a) km

2100 2050 2000 1950 1900 1850 1800 1750 1700 1650 sp
1 I J_ ! L J I I

mmm
!"L iiNltlllilltilllHiili mm ill! I BI I

UJ

NE
sw
Fig. 5. (a) Main NW-SE lateral fault crossing the main fault block. Noise on section on the fault position thought to be caused by
escape of gas into the shallow section. Exact number and location of faults within the disturbed zone unknown, (b) Collapse graben at
the apex of broad antiform which is thought to control the area of non-deposition of the synrift megasequence on the main fault block.

Depth maps to top and base synrift and a prelim- the reconstruction of the paleoenvironment during
inary isopach map of the megasequence were also the deposition of the synrift megasequence.
constructed, using velocity analyses obtained from
the BP Norway office and data from the well. The Structural mapping and regional setting
time and depth to base synrift surfaces were restored
to their original orientation by removing the effects Regional setting
of post depositional tilting (see section 4). A com- Figure 2 illustrates that the extensional trend
posite regional line was then backstripped to aid with changes azimuth sharply in the centre of the Case
140 K. Gerdes, J. Hurst and R. Jeans

72 30

72 15'
20 00'
Fig. 6. Two-way time map to base synrift.

Study Block. Tb the south of this central zone, ex- The transfer zone appears to have acted as a con-
tension is taken up along faults trending NNW-SSE duit for the escape of gas into the shallow section.
which turn to a truer NW-SE orientation as they The presence of the transfer zone produces a mor-
approach this zone. North of the zone the main phological ramp which dips downwards along the
extensional faults trend NE-SW, becoming closer to fault terraces from the transfer zone. TWo cusps (or
NNE-SSW farther north. This latter set includes the back basins) were mapped in the footwall of the main
main fault bounding the pre-rift target. Both sets of fault block.
faults lose displacement as they approach this central The first order set of faults north of the transfer
area and on some lines the displacement on them zone is cross-cut by a few faults which have the
is dispersed through a system of splays, although same approximate trend as the faults south of the
this was often difficult to map in detail due to poor transfer zone and throw down to the southwest.
data quality. The area in which both sets of exten- Only two of these faults cut both the top and basal
sional faults lose their displacement is interpreted reflectors of the megasequence, but they may have
as being the location of a transfer zone. The zone acted as important controls on the deposition of
itself is elevated above the fault terraces, is about the megasequence in the area of interest. The most
six kilometres in width and lies at the approximate important of these is the fault which effectively
intersection of the fault terrace and the northwest- bisects the fault block. This fault was seen on two
erly extension of the Tfrollfjord-Komagelv fault trend seismic lines, whereas the fault of similar trend to the
from the Hammerfest Basin. southwest was seen on only one line (Fig. 5a).
Seismostratigraphic study of a synrift megasequence,fromthe Barents Sea, northern Norway 141

Two-way time map to base synrift were obtained from a series of measurements made
The base synrift two-way time map (Fig. 6) shows on the seismic database.
the main fault block in closer detail and a number of
structural features particular to the area around the
well location. The fault block crest includes an area Controls on synrift sedimentation
of around 12 km in length and up to 2 km in width
Introduction
where the synrift sediment wedge is absent. Uplift
of the crest appears to have been accentuated along Studies in presently active rift systems have indi-
this zone with respect to the rest of the block crest as cated the dominant role which structurally controlled
evidenced by the increased separation between the topography plays in the distribution of sand-prone
fault block crest and the top of the sediment wedge sediments, (Crossley, 1984). Such topography con-
on its hanging wall. The fault block crest shows trols the location of sediment entry points into the
evidence of what may be minor footwall collapse rift system and also affects erosion and, hence, sedi-
features on the extreme northwestern edge on some ment production potential. Field studies indicate that
lines. sediment derived from the flanks of the rift system
The fault block surface dips down into the major gains access to the low lying rift axis predominantly
fault with a relief of up to 400 m from the crest via areas such as transfer zones and footwall cusps or
of the block. This trend is interrupted by a lateral back-basins. Examples of this structural control on
fault which bisects both the main fault block and the access of rivers to the rift system and location
the main footwall cusp (Fig. 6). The throw on this of sediment deposition in a rift are given in Figs. 7
fault is poorly constrained. The interplay of the fault and 8 from the Gregory Rift of Kenya and the Gulf
block rotation and this lateral fault define the first of Suez, respectively. A number of different types
order morphology of the base synrift surface. This is of transfer zone have been identified. The nature of
further modified by a set of structural features which the associated sediment bodies varies with the scale
only affect the base synrift surface. and style of topography generated within the transfer
South of the lateral fault, the morphology of the zone.
base synrift is that of the broad antiform with the
apex located beneath the area in which the synrift
deposits are absent. This is shown in detail in Fig. 5b
which shows a broad updoming centred around a
small collapse graben.
In addition to this broad doming there is also a
minor antiform, which runs subparallel to the main Elgeyo
Escarpment
fault trend with an average amplitude of 10 m, on the
southern half of the fault block. This is bounded to
the southeast by a small fault down to the northwest
which runs irregularly along the length of the fault Lake Baringo
NORTH
block mapped. Lake Bogoria
The origin of these minor features is uncertain.
They further enhance the relative relief on the base
synrift surface and emphasise the morphological de- 100km

pression which runs beneath the main fault scarp

southwest of the lateral fault. This area should pref-
erentially have collected the earliest sediments shed
) Uganda / Kenya
along the fault block axis.
The base synrift surface should approximate to
the topography upon which the initial synrift sedi-
ments were deposited after allowing for the effects of
any post-depositional deformation. An approximate Tanzania
'Indian
Ocean
Lake Magadi
correction for the effects of post-depositional tilting
was made by tilting the base synrift surface down
to the southeast by between 3 and 4 and down to Lake Natron

the northeast by approximately 0.9 before flattening Fig. 7. Role of transfer zones on drainage patterns: Gregory
onto the top synrift reflector. These rotation angles Rift, Kenya.
142 K. Gerdes, J. Hurst and R. Jeans

the zone. The fault blocks within the fault terrace

north of the transfer zone have a regional dip down
to the northeast (see Fig. 6). The cuspate faults
southeast of the main fault block may also have had
an effect on sediment input. If the sediment transport
direction were predominantly axial then these cusps
would have controlled any additional lateral input
from higher fault blocks.
Superimposed on these extensional geometries are
additional transfer elements which consist of NW-
SE cross-cutting faults. These are more difficult to
identify on the data, but are shown to partition the
fault blocks in Fig. 6. These faults laterally offset
the main NE-SW extensional trends and the largest
throw on the main fault occurs at the point of
intersection of the two trends.
In addition to these first order structural controls
on sedimentation, more subtle second order effects
exist which modify the details of the depositional
surface. The doming of the pre-rift deposits accentu-
ates the relative relief on the main fault block and
clearly affects the youngest sequence. It apparently
controls the position and extent of the area of non-
deposition at the fault block crest. This doming may
have been caused by diapirism at deeper levels or by
0 1.5 Kms minor compression associated with the transfer zone.
Either of these processes may have created the slope
Fig. 8. Role of transfer zones and footwall cusps in distribution instability at the fault block crest which led to the
of sand prone sediments: Gulf of Suez, Egypt. (1) Lateral sand
distribution down the footwall slope enters the graben at transfer slumps down the hanging wall seen to affect seismic
points, where the throw on the main fault is lost. (2) Lateral sequence 3 (see below).
sand input is focused at the intersection of cuspate faults with
the main footwall. Provenance
The megasequence is absent in an extremely lim-
ited area on the main fault block. Hence internal
Structural and depositional controls of sediment production within the block itself can be
the synrift megasequence considered to be minimal and the dominant prove-
nance areas for synrift sediments deposited on the
Structural fault terrace are the rift flanks lying to the east of the
The large scale schematic structural map (Fig. 2) fault terrace.
shows two extensional fault trends, one NW-SE and The subcrop of pre-rift sequences was clearly ob-
the other NE-SW. The throws on both sets of faults served on seismic data from the rift flanks. This
die out into a complex transfer zone in the centre truncation appears mainly to affect sequences of
of the case study block. It was not possible to show mid-Jurassic age. The available well data from the
during this study whether the two trends separated region indicate that these sequences largely comprise
by the transfer zone were active simultaneously. The highly porous quartz sandstones (Olaussen et al.,
true extension direction was probably east-west, and 1984). Hence, at least the earliest sediments of the
hence both sets of faults would have had an amount synrift megasequence should include the products of
of oblique motion along them. The detailed study reworking of these quartz sandstones.
area is located north of the transfer zone, and the
surface morphology generated in the proximity of the Depositional surfaces
transfer zone during the rifting was almost certainly The base synrift surface approximates to that on
an important first order control on sedimentation. which the synrift sediments were deposited, after
The transfer zone itself would have formed a positive allowing for the effects of post-depositional deforma-
topographic feature during synrift deposition. The tion. Whether deposition is subaerial or submarine,
faults leading from it increase in throw away from sediment pathways developed on this surface would
Seismostratigraphic study of a synrift megasequence,fromthe Barents Sea, northern Norway 143

72 30' ^_ T -
DEPOSITIONAL SURFACE AT BEGINNING OF
SEQUENCE ONE WITH SEDIMENTARY CONDUITS
(THICKNESS OF ARROW INDICATES
RELATIVE IMPORTANCE
OF CONDUIT)
0 3

72 20' L
20 00'
Fig. 9. Sedimentary conduits for seismic sequence 1.

have focused the sediment into the morphological Sequence mapping and depositional
deeps, i.e., areas A and B in Fig. 9. The sediments models
would have been further ponded into these areas by
Synrift megasequence
the second order affect of the small NE-SW oriented
structure approximately 4 km from the crest of the The base of the megasequence was dated from the
fault block (see Fig. 6). The preliminary well results well as being of late-Jurassic age. It overlies an angu-
indicated that the lowermost synrift sediments on lar unconformity which is the boundary between the
the fault crest were glauconitic and bioturbated and pre- and syn-rift (Fig. 11). This boundary is marked
were interpreted as being submarine. Since this area by a seismic event consisting of a high amplitude,
is the shallowest location on the fault terrace, it fol- moderate frequency, generally continuous positive
lows that the deposition down-dip from this location peak. It became more difficult to distinguish this
was also marine, passing further down-dip to deep event in the vicinity of the main boundary fault,
marine. Any coarse sediments present in this setting where it was often poorly migrated and partially ob-
would have to have been introduced by mass flow scured by migration and other noise from the fault
processes given the depths of water calculated from and footwall.
the relative relief between the top and base of each The top of the megasequence was dated as be-
sequence, i.e., between 200 and 400 m. ing early Cretaceous. This angular unconformity is
Sediments would have been introduced axially at recognised as a major onlap surface (Fig. 11). The
point A (Fig. 9) at the onset of rifting, but the timing defining seismic event followed was a strong, contin-
of sediment input from the upper fault block via the uous, positive peak of moderate amplitude and high
footwall cusp shown in Fig. 9 cannot be determined. frequency.
Any lateral input of sediment would obviously have The two-way-time isopach map for the synrift
postdated the filling of the footwall cusp itself. Fig- megasequence is shown in Fig. 12. The fault terraces
ure 10 assumes that the footwall cusp was filled prior on all the isopach maps have been divided into zones
to the deposition of seismic sequence 2. dependent upon the presence or absence of the
megasequence penetrated at the well. In Zone A the
megasequence is very thin or absent, in Zone B small
wedge shaped bodies exist, but they are isolated from
144 K. Gerdes, J. Hurst and R. Jeans

72 30

72 20' L
20 00'
Fig. 10. Sedimentary conduits for seismic sequence 2.
SEISMIC LINE 401

NW SE
260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20
J_J l_J I I I I I I I I I I I I I I I I I I I L_L

h- DOWNLAPS
IN MOUND

UNDULATING PARALLEL DISTURBED

SURFACE - MAYBE DISCONTINUOUS RUMBLY
CAUSED BY SLUMPING

Fig. 11. Synrift megasequence, sequence boundaries and seismic facies used in this study. Reflector characters and morphological
features referred to in the text are annotated.

correlatable deposits. It is not yet proven, but it extensive thickness of post-rift sediments. Within
is likely, that all these bodies are contained within Zone C there are two main depocentres on the main
the megasequence. Zone C includes the main fault fault block and a small depocentre in the footwall
block and footwall cusp within which the sediment cusp. The two depocentres on the main fault block
bodies can be correlated as being part of the synrift shown in Fig. 12 are separated by the lateral fault.
megasequence. Finally, Zone D covers the area to The megasequence is absent from a triangular area
the northwest of the main fault block of interest approximately 2 x 3 km on the upthrown side of the
where any synrift deposits present are obscured by lateral fault and also from a lens shaped area about 6
Seismostratigraphic study of a synrift megasequence, from the Barents Sea, northern Norway 145

72 30H

72 20'4

Fig. 12. Isopach map of synrift megasequence.

km south of the lateral fault. This second area has a Seismic sequence 1
maximum width of nearly 2 km centred over the apex
of an elongate antiform in the pre-rift. The basal unconformity of this sequence is the
Both depocentres hug the base of the fault scarp. base synrift reflector in Fig. 11, and was dated from
The southern depocentre is thicker and more lat- the well as being of late-Jurassic age. Seismic se-
erally continuous than the northern one. They are quence 1 is correlated with the lowermost part of
centred over the morphological deeps mapped on sequence C in the well (Fig. 4).
the base synrift surface (Fig. 6) and achieve maxi- The isopach map of this unit is shown in Fig. 13.
mum thickness of 500 m in the north and over 600 m Two depocentres run along the base of the main
in the south. boundary fault and are separated by the main lateral
The distribution of the megasequence is shown fault cutting the fault block (Fig. 6). The sequence
in Fig. 12. On the higher terrace east of the main achieves a maximum thickness of over 160 m in
fault block, the megasequence is restricted. West the southern depocentre and 100 m in the north.
of the main fault block, the megasequence thick- Ponding of this sequence is further accentuated by
ness is indeterminate. Well results indicated that the the second order features on the basal surface. A
distal part of the megasequence wedge is marine. small antiformal structure appears to separate the
Backstripping indicated that the coastline was in the main ponded area from the section of the sequence
vicinity of the fault dividing zones C and D in Fig. which gradually thins onto the fault block crest. It
12. It appears, therefore, that the megasequence was also forms the approximate boundary between the
deposited in shallow to deep marine environments, undisturbed and disturbed seismic reflector charac-
with the paleowater depth increasing as shown by the ters. The seismic sequence onlaps onto the fault crest
TWT map in Fig. 6. and terminates very close to the pinch-out trace of
the megasequence.
146 K. Gerdes, J. Hurst and R. Jeans

72 30'4

72 20'

Fig. 13. Isopach map of seismic sequence 1 showing seismic facies.

This seismic sequence was divided into two seismic in an even deeper marine environment.
facies. A zone outlined on the map between the main Parallel, continuous seismic facies calibrate with
fault and the small antiformal structure on the base low energy marine mudstones in the well. The
synrift surface enclosed a disturbed seismic char- change in seismic facies in this sequence down-dip
acter which became rumbly on the high resolution to a disturbed character, is used to infer a corre-
data. This seismic facies illustrates strong, high angle sponding change in lithology from mud dominated
downlaps and the events are generally of moderate to sand-prone sediments. Given the paleogeography
amplitude and high frequency with occasional bursts and structural configuration at this time, the only
of high amplitude, low frequency energy. mechanisms which could transport sand-prone sedi-
The seismic character of the remainder of the ments into this deep marine setting would be mass
sequence was generally continuous, with events of flow processes (Nardin, 1983). The manner in which
lower amplitude and higher frequency onlapping the disturbed seismic facies is confined to the mor-
onto the fault block crest and disappearing prior to phological deeps is consistent with the interpretation
the pinch out of the whole megasequence. The onlap of massflowdeposits. The more rumbly seismic char-
in this part of the sequence is of a much lower angle acter within this facies is also located at the point
than in the rest of the sequence. of closest approach to the two main sedimentary
conduits postulated for the sequence, as shown in
Sedimentary interpretation and Fig. 9, which is consistent with the areal distribution
depositional model of sand-prone sediment being controlled by these
The well results indicate that the lowermost units structural features.
of the megasequence consist of glauconitic, biotur- Correlations between similar seismic facies and
bated siltstones and mudstones, indicative of low lithofacies have been made in a number of other
energy marine conditions (Fig. 4). The sequence 1 setting, e.g., Nardin (1983), Geoquest manual
sediments down-dip of this location were deposited (Macurda, 1900) and the Brae field of the UKCS.
Seismostratigraphic study of a synrift megasequence, from the Barents Sea, northern Norway 147

RUMBLY-MOUNDED
MASS-FLOW DEPOSITS

TRANSFER ZONE

SCARP-SLOPE
FAILURE
PARALLEL-DISCONTINUOUS
LOW-ENERGY MARINE
Fig. 14. Depositional model for seismic sequence 1 (seismic character above, sedimentary interpretation below).

In the Magnus area of the UK North Sea, a similar Seismic sequence 2 was correlated with the upper
change in seismic facies occurs in a similar structural part of sequence C (Fig. 4).
setting and the disturbed seismic facies relates to The isopach map of this unit is shown in Fig. 15.
high density sand-prone turbidites (M. Shann, pers. This sequence is the most extensive of all sequences
commun.). mapped, reaching a maximum thickness in both de-
The geometry and internal configuration of seis- pocentres of over 180 m. Prior to the deposition of
mic sequence 1 is not consistent with any of the this sequence, a substantial amount of the topogra-
classic radial or elongate submarine fan depositional phy on the fault block had been infilled by sequence
models (Stow et al., 1981; Mitchum, 1985; Surlyk, 1. Hence, none of the second order features on the
1978). It is most closely comparable with the subma- basal synrift surface appear to affect the distribu-
rine ramp model proposed recently by Surlyk (1987) tion of sequence 2. A distinct, laterally impersistent
for the Upper Jurassic Hareelv Formation of central mound runs subparallel and very close to the fault
East Greenland. In this model axially transported scarp (Fig. 11).
sediment passes down a ramp, along the fault terrace In a zone broadly defined by the main fault and the
with input via the transfer zone and is deposited in 100 m contour the events are of a lower amplitude,
deep water under poorly oxygenated conditions dur- higher frequency and are less continuous than the
ing a period of increasing base level. The sand-prone rest of the sequence.
sediments form anastomosing bodies embedded in The sequence northwest of the 100 m contour con-
a shaly matrix. Lateral connectivity between sand sists of very continuous events of higher amplitude
bodies is extremely limited (see Fig. 14). This forms and lower frequency (Fig. 11). The character in this
the basis for the depositional model developed for area is also affected by the attenuation of the beds as
seismic sequence 1 shown in Fig. 14. Requejo et the sequence thins onto the fault block crest.
al. (1989) have shown that the oil found in seeps
within the sandstones of the Hareelv Formation was Sedimentary interpretation and
generated from the Hareelv Formation interbedded depositional model
shales. Hence, the depositional environment pro- The geometry of this seismic sequence most closely
posed for seismic sequence 1 includes both reservoir resembles that of a "classic" synrift wedge (Mitchum,
and source rocks. 1985). The parallel continuous seismic character of
the distal part of this sequence (Fig. 11) is correlated
Seismic sequence 2 with the low energy marine, probably condensed,
mudstones and siltstones encountered in the well.
The basal unconformity of this sequence (Fig. 11) The disturbed seismic character down-dip of this fa-
was defined as the first onlap surface above the basal cies is inferred to indicate a lithology change to more
unconformity. This event is a moderately continuous, sand-prone turbidites and other massflowunits. This
high frequency, moderate amplitude positive peak. facies also shows poorly developed internal downlap.
148 K. Gerdes, J. Hurst and R. Jeans

72 30'-f

KEY
(in msecs)
0-50
50-100
100-150
150 +
72 20'-f ZONE A *
20 00' SCALE 5km

Fig. 15. Isopach map of seismic sequence 2 showing seismic facies.

A further indication of the possible sand-prone of which parallels the trend of the main fault and
nature of the mound feature is the antiform in which is localised around the footwall cusp (Fig.
the overlying sequences (Fig. 11). Examples of such 15). The location, morphology and presence of inter-
features on seismic data from the Far East have nal downlap within this feature is taken to indicate
been shown to be caused by differential compaction. the possibility of a lateral input of sediment from
Drilled examples from Thailand indicate that the the higher fault terrace via the footwall cusp (Fig.
mounds consist of sand-prone alluvial fans encased 10). Lateral derivation of sediments into a major
and overlain by low energy mudstones (C. Sladen, graben is commonly associated with overspill from
personal communication). Since such an effect is de- small footwall grabens, e.g., Gulf of Corinth, Greece,
pendent upon the physical properties of sand and Gulf of Suez, East African Rift System (Jackson
shale, one may expect to see it in all mixed deposi- et al., 1982; Rosendahl et al., 1986; Leeder and
tional environments. Gawthorpe, 1987; Gawthorpe et al., 1990). A direct
The features of this seismic sequence are taken to consequence of such a model would be that the bulk
indicate a substantial lateral input of sediment via the of the sediments within the footwall cusp correlate
footwall cusp during deposition. The characteristics with the earliest seismic sequence on the main fault
of this seismic sequence are more indicative of fans block. This depositional model is similar to those
radiating out from the footwall, such as borderland proposed for the Brae field (Stow, 1981) and the
fans associated with small transfer zones (Nardin, late Jurassic fans of the Wollaston Foreland of East
1983). Greenland (Surlyk, 1977) (Fig. 16). In these models
An important consideration in the development the sand-prone sediments are ponded up against the
of the depositional model for this sequence is the fault scarp (see Fig. 16). However, the differences in
structural position and external geometry of the scale should be borne in mind: radial fans in seismic
disturbed seismic facies unit. This facies unit in- sequence 2 are no more than 2 km across and be-
corporates a distinct elongated mound, the apex tween 300-400 metres thick which is less than half
Seismostratigraphic study of a synrift megasequence, from the Barents Sea, northern Norway 149

RUMBLY-MOUNDED
MASS-FLOW DEPOSITS

PARALLEL-CONTINUOUS
LOW-ENERGY MARINE (CONDENSED)

Fig. 16. Depositional model for seismic sequence 2 (seismic character above, sedimentary interpretation below).

the size of those in the Brae and Woolaston Foreland stones (Fig. 3). The truncation of reflectors at the
examples. footwall crest, illustrated on Fig. 17, is possibly the
result of low angle slumping in unconsolidated sedi-
Seismic sequence 3 ments. Similar features have been observed on three-
dimensional data from analogous structural settings,
The base of this sequence, see (Fig. 11) was the e.g., Magnus field (Price, personal communication).
most prominent unconformity and strongest event In the Magnus example, sediments of the same age
in the dataset. The event is a very continuous, high and lithologies as in Bj0rn0ya South slide down the
amplitude, moderate frequency positive peak with hanging wall to form cuspate reflector geometries
high frequency events onlapping onto it. The event directly analogous to those observed within seismic
is deformed at the fault block crest with two im- sequence 3.
persistent cuspate features running sub-parallel to The entire sequence consists of one seismic facies
the subcrop trace of the wedge. These features were which was calibrated at the well with deep marine
mapped and are shown in Fig. 17. Seismic sequence 3 mudstones. The external sequence geometry and
was correlated with sequence B from the well results internal characteristics are consistent with deposition
(Fig. 4). on a low energy slope apron as shown in Fig. 18
The isopach map of this sequence is shown in Fig. (Pickering, 1984).
17. This sequence is the thinnest of all the seismic
sequences mapped and shows the least variation in
thickness. It attains a maximum thickness of over 80 Seismic sequence 4
m in the southern depocentre and around 40 m in
the northern area. This "blanket like" morphology The basal unconformity of this sequence (Fig. 11)
indicates that the basal synrift surface morphology was defined as an onlap surface. This unconformity
has been completely masked prior to the deposition is thought to be the base of sequence A from cor-
of seismic sequence 3 and that the sequence was de- relation with the well (Fig. 4). The isopach map of
posited on a semi-continuous inclined plane. The se- this sequence is not given. The sequence thickens
quence extends over the fault block northwestwards regularly into the main fault to a maximum of over
to the pinchout trace of the whole megasequence. 120 m in the south and 160 m in the north. It thins
The higher resolution data to the south show the se- to zero prior to the pinch out trace of the whole
quence to consist of very continuous, high amplitude, wedge along the fault block crest. The thickening of
moderate-high frequency events which onlap at a the sequence into the fault probably represents reac-
low angle onto the basal unconformity. tivation of motion along this feature after the period
of relative quiescence associated with the deposition
Sedimentary interpretation and of seismic sequence 3.
depositional model The sequence consists of parallel continuous, high
This seismic sequence is correlated at the well frequency, low amplitude events which onlap at a
with low energy, deep marine organic rich mud- low angle onto the basal unconformity. The area of
150 K. Gerdes, J. Hurst and R. Jeans

72 30'4

KEY
(in msecs)
0-50
U l i 50-100
100-150
150 +
72 2 0 H ZONE A -
20 00'

Fig. 17. Isopach map of seismic sequence 3 showing seismic facies.

discontinuous events shown on the map may well amplitude and frequency. These events onlap at a
be a function of data quality as both sequences low angle onto the basal reflector (Fig. 11).
4 and 5 were difficult to pick in this area. The
parallel/convergent continuous seismic facies at the Sedimentary interpretation and
fault block crest is correlated with low energy mud- depositional model for sequences 4 and 5
stones in the well. Towards the main graben fault The wedge shaped geometry of both seismic se-
a slightly rumbly seismic facies was seen. This could quences is indicative of continued (albeit reduced)
indicate the possibility of coarser grained, more sand- movement along the main fault during their de-
prone material in this area, possibly laterally derived position. The parallel, continuous seismic facies of
through the footwall. sequences 4 and 5 are correlated at the well with red
mudstones and minor associated sandstones (Fig. 4).
Similar deposits of Valanginian age are seen in north-
Seismic sequence 5 ern East Greenland at the crests of tilted fault blocks
and are interpreted as condensed units deposited
The basal unconformity of this sequence is a very in low energy, deep marine environments (Surlyk,
continuous positive peak of moderate amplitude and 1977, 1978). A similar depositional environment is
frequency and events of similar character onlap onto proposed for seismic sequences 4 and 5.
it. Closer to the main fault, this gross depositional
The sequence has the same general shape as that environment may be modified by the input of later-
of sequence 4, with slightly less thickening into the ally derived material via the footwall. This situation
main fault. Sequence 5 attained a maximum thick- may be broadly analogous to that described by Pick-
ness of over 120 m in the southern part of the ering (1984) from the late Jurassic of Scotland along
fault block and 140 m in the more poorly con- the Great Glen Fault where debris spalled the
strained northern part. The sequence consists of very footwall and is locally concentrated along the scarp.
continuous, parallel and sub-parallel events of high However, the majority of seismic sequences 4 and 5
Seismostratigraphic study of a synrift megasequence, from the Barents Sea, northern Norway 151

RUMBLY-MOUNDED
MASS-FLOW DEPOSITS

PARALLEL-CONTINUOUS
LOW-ENERGY MARINE
Fig. 18. Depositional model for sequences 3, 4 and 5 (seismic character above, sedimentary interpretation below).

are remote from the fault scarp and, hence, the to be high, are concerned with its position, banked
depositional system most probably approaches that up against one of the major faults in the region.
of a deep marine, low energy muddy slope apron Hence, two discrete plays were developed from a syn-
(Fig.18). rift megasequence using the above techniques, with
the overriding assumption that sand-prone sediments
Conclusions would have gained access to the synrift environment.

(1) Using a combination of seismic stratigraphy

Acknowledgements
and the results of an inhouse synrift R&D project, an
attempt was made to refine the risks associated with a This work incorporates the ideas and efforts of a
synrift megasequence play in a frontier area. Seismic number of our colleagues within BP Exploration.
stratigraphy was used to subdivide the megasequence We would like to thank Chris Sladen and Rob
into a number of sequences and seismic facies units: Gawthorpe who formed the core of the Synrift R&D
detailed structural mapping of restored two-way time team. Dave Jackson backstripped a regional line
surfaces were used to generate the approximate across the area. Dave Nash, Steve Edrich and Dave
morphology of the depositional surfaces during the Roberts are thanked for their encouragement during
rifting phase. the various stages of the project as is John Bellamy,
(2) This study highlights the two lowermost seis- whose comments considerably improved the final
mic sequences (1 and 2) as being the most likely text. We would like to thank BP Exploration and
sites for the accumulation of sand-prone sediments the licence partners, i.e., Statoil, Norsk Hydro, Mobil
of sufficient size to be of interest to hydrocarbon and Fina for allowing the publication of this paper.
exploration. The general characteristics of seismic
sequence 3 would be consistent with the develop-
ment of a synrift sequence with source potential. References
Interbedded shales within seismic sequence 1 are
Crossley, R., 1984. Controls of sedimentation in the Malawi Rift
also likely to have source potential. Valley, Central Africa. Sediment. Geol., 40: 33-50.
(3) Two distinct depositional models were gener- Gawthorpe, R.L., Hurst, J.M. and Sladen, C.P., 1990. Evolution
ated for the two most likely sand-prone sequences. of Miocene footwall derived coarse grained Deltas, Gulf of
Suez, Egypt: implications for exploration. Am. Assoc. Pet.
These models imply completely different distribu-
Geol., Bull, 74(7): 1077-1086.
tions of sand-prone sediments, sand body geome- Faleide, J.I., Gudlaugsson, S.T. and Jacquart, G., 1984. Evolution
tries and net: gross estimates and, hence, have com- of the Western Barents Sea. Mar. Pet. Geol., 1: 123.
pletely different model risks associated with them. Jackson, J.A., Gagnepain, J., Housemand, G., King, G.C.P., Pa-
The model risks for seismic sequence 1 focus on padimitriou, P., Soufleris, C. and Virieux, J., 1982. Seismicity,
normal faulting and the geomorphological development of
the lateral inhomogeneity of the anastomosing sand
the Gulf of Corinth (Greece): the Corinth earthquakes of
bodies, inherent low connectivity and axial updip February and March, 1981. Earth Planet. Sei. Lett., 57: 377-
carrier bed. The risks associated with the model for 397.
sequence 2, in which lateral connectivity is likely Leeder, M.R. and Gawthorpe, R.L., 1987. Sedimentary models
152 K. Gerdes, J. Hurst and R. Jeans

for extensional tilt-block/ha lf-gra ben basins. In: M.P. Coward, R0nnevik, H.C., Jacobsen, H.P., 1983. Structural highs and basins
J.F. Dewey and P. Hancock (Editors), Continental Extensional in the western Barents Sea. In: A.M. Spencer et al. (Editors),
Tectonics. Geol. Soc. London, Spec. Publ., pp. 139-152. Petroleum Geology of the North European Margin. Norw.
Macurda, D.B., Jr. (Editor), 1900. Seismic Facies Analysis Pet. Soc, Graham and Trotman, London, pp. 19-32.
Manual. Geoquest International, 355 pp. Rosendahl, B.R., Reynolds, D.J., Lorber, P.M., Burgess, C.F,
Mitchum, R.M., 1985. Seismic stratigraphic expression of sub- McGill, J., Scott, D., Lambiase, J.J. and Derksen, S.T.,
marine fans. In: O.R. Berg and D.G. Woolverton (Editors), 1986. Structural expressions of rifting: lessons from Lake
Seismic Stratigraphy II, Am. Assoc. Pet. Geol., Mem., 39: Tanganyika, Africa. Geol. Soc. London, Spec. Publ., 25: 29-
117-136. 43.
Nardin, T.R., 1983. Late Quaternary depositional systems and Sangree, J.B. and Widmier, J.M., 1977. Seismic stratigraphy and
sea-level change Santa Monica and San Pedro Basins, global changes of sea level, Part 9: Seismic interpretation
California Continental Borderland. Am. Assoc. Pet. Geol., of clastic depositional facies. In: Seismic Stratigraphy
Bull., 67: 104-124. Applications to Hydrocarbon Exploration. Am. Assoc Pet.
Olausson, S., Dalland, A., Gloppen, T.G. and Johannessen, E., Geol., Mem., 26: 165-184.
1984. Depositional environment and diagenesis of Jurassic Stow, D.A.V., Bishop, C D . and Mills, S.J., 1982. Sedimentology
reservoir sandstones in the eastern part of the Troms 1 area. of the Brae Oilfield, North Sea: fan models and controls, J.
In: A M . Spencer et al. (Editors), Petroleum Geology of the Pet. Geol., 5: 129-148.
Northern European Margin. Norw. Pet. Soc, Graham and Surlyk, F , 1977. Stratigraphy, tectonics and paleogeography of
Trotman, London, pp. 61-79. the Jurassic sediments of the areas north of Kong Oscar's
Pickering, K.T., 1984. The Upper Jurassic "Boulder Beds" and Fjord, E. Greenland. Gr0n. Geol. Unders., 123, 56 pp.
related deposits: a fault controlled submarine slope, N.E. Surlyk, F , 1978. Submarine fan sedimentation along fault scarps
Scotland. J. Geol. Soc. London, 141: 357-374. on tilted fault blocks (Jurassic-Cretaceous boundary, E.
Requejo, A.G., Hollywood, J. and Halpern, H.I., 1989. Recog- Greenland, Gr0n. Geol. Unders., 123, 108 pp.
nition and source correlation of migrated hydrocarbons in Surlyk, F , 1987. Slope and deep shelf gully sandstones, Upper
Upper Jurassic Hareelv Formation, Jamesonland, E. Green- Jurassic, E. Greenland, Am. Assoc. Pet. Geol., Bull., 71:
land. Am. Assoc. Pet. Geol., Bull., 73(9): 1065-1088. 464-475.

K. GERDES BP Exploration, Britannic House, Moore Lane, London EC2Y9BU, UK

Present address: BP Norway Limited U.A., PO. Box 197, N-4033 Forus, Norway
J. HURST BP Exploration, Britannic House, Moore Lane, London EC2Y9BU, UK.
R. JEANS BP Exploration, Britannic House, Moore Lane, London EC2Y9BU, UK.
 153

Tertiary Uplift and Erosion in the Barents Sea: Magnitude,

Timing and Consequences

B. Nyland, L.N. Jensen, J. Skagen, O. Skarpnes and T. Vorren

In 1920 Fridtjof Nansen postulated an uplift and erosion in the order of 500 m in the Barents Sea. Hydrocarbon exploration
in the 1980's has since confirmed a significant uplift and erosion in this area.
In 1985 Statoil initiated a multidisciplinary study to evaluate and quantify the amount of uplift and erosion. The methods
used were: Shale compaction curves, Dc-exponent from drilling, vitrinite reflectance trends, diagenesis of clay minerals, apatite
fission track studies, volumetric mass balance calculations of erosion and deposition, and PVT-modelling of known reservoirs.
Studies of porosity and density trends in Tertiary shales from wells showed an uplift and erosion of the order of 500-1000
m in the Hammerfest Basin, increasing to about 1700-1800 m towards the Fingerdjupet Subbasin area. Vitrinite trends from
exploration wells and IKU shallow drillings confirmed these results also showing increasing erosion towards the north (3-4 km
erosion at Bj0rn0ya).
Depending on the paleo-drainage area chosen, volumetric calculation of the Late Tertiary deposition along the western shelf
margin gives a magnitude of erosion of about 1-1.2 km in the southern part of the Barents Sea, increasing to approximately 3
km on Svalbard.
Used together, the methods above give a fairly accurate estimate on the magnitude of the erosion but not on timing. Fission
track studies indicate one or more phases of uplift from Late Oligocene to Pliocene/Pleistocene, while biostratigraphic redatings
of sediments from the wells on the Senja Ridge suggest that a major part of the erosion is related to glacial processes during
the last 2-3 Myr.
Extensive residual oil staining found below the present hydrocarbon/water contacts indicates that reservoirs, such as the
Sn0hvit field in the Hammerfest Basin, once contained larger quantities of oil. PVT-modelling has shown that this oil was lost
probably as a consequence of gas expansion during uplift, and can explain in part the lack of commercial oil finds so far in the
Barents Sea.

Introduction fest Basin is the unusually low sandstone porosities

at present depth. Olaussen et al. (1984) suggested
Based on bathymetry, seabottom topography and 1000-1500 m deeper burial of the sediments to
geological information from surrounding areas, explain these low values. Quartz cementation by
Fridtjof Nansen postulated a Late Tertiary uplift pressure solution also indicates that the Jurassic
and erosion of about 500 m in the south western part sandstones in the Hammerfest Basin have been more
of the Barents Sea (Nansen, 1920). Later, coal petro- deeply buried in the past (Berglund et al., 1986).
graphic studies on Svalbard (Manum and Throndsen, Maturity data also indicate a maximum burial
1978) confirmed a substantial uplift of the area, and depth of at least 1000 m deeper burial, or higher heat
about 1.7 km erosion was suggested for the central flow than observed at present to explain the level of
part of Svalbard. maturation seen for the source rocks penetrated.
The first exploration wells drilled in the Hammer- Three large fan systems are deposited along the
fest Basin (Fig. 1) demonstrated a marked unconfor- western margin of the Barents Shelf; outside the
mity between Plio-Pleistocene and underlying strata. Bj0rn0ya Ttough, outside the Storfjorden Trough
This has later been confirmed by all wells drilled and north of Spitsbergen. These fans are interpreted
in the area, and this major unconformity is clearly to be depocentres for the erosional products from
seen on the seismic data. Where Tertiary sequences the shelf (Vorren et al., 1988, 1990; N0ttvedt et al.,
are present, they consist mainly of Paleocene and 1988).
Eocene sediments, whereas Oligocene and Miocene A multidisciplinary study has been continuing
sediments are absent. since 1985 to quantify the magnitude and timing
Another observation from wells in the Hammer- of the uplift and erosion. The different methods and

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T Larsen and E. Talleraas.
NPF Special Publication 1, pp. 153-162. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
154 B. Nyland, L.N. Jensen, J. Skagen, O. Skarpnes and T. Vorren

9 10
11
r 20
" 21
" 22
" 23
24
25
26
27
28
29
" 30
32 33
34
35

Fig. 1. Barents Sea; main structural elements (after Gabrielsen et al., 1990).

the main results will be summarized in this paper. are confidential, the results have to be presented in a
The possible effects of the uplift and erosion on general way without showing any raw data.
a hydrocarbon filled reservoir with respect to PVT- In addition, regional mapping and evaluation of
modelling, has also been studied in order to test if the Cenozoic sequence in the western part of the
this can explain gas discoveries with only traces of Barents Sea have been performed at the University
residual oil and the presence of residual oil in most of TYoms0.
of the dry wells.

Methods TABLE 1

Different methods have been used in order to Methods used to quantify uplift and erosion
quantify the uplift and erosion (T^ble 1). The
Vitrinite reflectance profiles
database used consists of data available from explo- Shale compaction curves
ration wells, IKU (Continental Shelf and Petroleum Drilling parameters
Technology Research Institute) shallow drillings, Fission track analysis
Diagenesis of clay minerals
seismic data, and results from fieldwork on Sval- Volumetric mass balance
bard and Bj0rn0ya. Since most of the well data used
Tertiary Uplift and Erosion in the Barents Sea: Magnitude, Timing and Consequences 155

Vitrinite reflectance profiles and heat flow history to try to avoid some of the
problems mentioned. Based on these considerations,
Vitrinite reflectance analyses are available from this method gives an error within 200 m in our
all exploration wells and most IKU shallow drillings, experience.
and are used as a measure of the level of maturation.
A typical vitrinite reflectance profile through a well
Shale compaction curves
in the Hammerfest Basin is shown in Fig. 2. The
profile can be used to estimate the thickness of This study is based on the use of published curves
a missing sequence by extrapolating the trendline to for compaction of shales in the central North Sea
VR = 0.2%, which is anticipated to be the reflectivity Basin (Sclater and Christie, 1980). The method uses
in the sediments deposited on seabottom (Dow, the irreversible compaction of shale (also sandstone,
1977). The differences between the Base Quaternary limestone and shaly sandstone) as a function of
unconformity and a VR = 0.2% will represent the depth. Values from density logs define the input into
thickness of the sequence missing, which is about the equation used. Compared with the compaction
1100 m for the well shown in this case. curves, a higher density than expected at the present
The use of this method is highly debatable (Katz depth indicates a deeper burial, and the differences
et al., 1988; Feazel and Aram, 1990) since the vit- in depth represent the amount of erosion. An ad-
rinite reflectance is controlled by parameters as the vantage of the method is that calibration points from
heat flow, burial rate, type of organic material, etc., the entire well can be obtained. The great number of
but also on the quality of the sample materials used data points can then be further analysed statistically
(cores, cuttings, etc.) and that the measurements are and also different lithologies can be used.
performed on in-situ material. The method is also A statistical programme has been applied to den-
sensitive for small changes in tracing the gradient. sity values from log data and the results compared
Due to these sources of errors the database was care- with the published curves in order to construct com-
fully examined and poor samples were excluded. We paction curves for the Barents Sea. Erosion from
also compared wells with the same geological setting shale compaction curves are calculated for all wells
and the results seem to be consistent. A source
of error using this method is that the parameters
0.5 0.7 1.0 VR %
taken from the functions used by Sclater and Christie
(1980) may not be representative for other areas such
as the Barents Sea. This problem can to some extent
be eliminated by using compaction curves from wells
with minimal erosion in the area (as the wells on the
Senja Ridge), and then to compare the other wells in
the area.

Erosion from drilling parameters

An internal Statoil study has shown that different

parameters measured during drilling can be used to
estimate the missing sequence in a well (J. Skagen,
pers. commun., 1989). Tests have shown that the rate
of penetration (ROP) is related to the compaction
trend of shales with depth, given that other drilling
parameters are constant.
The Dc-exponent (Jordan and Shirley, 1966), used
in the prediction of pore pressure, is a function of
the ROP (and other drilling parameters) and gives
an estimate on the compaction trend in a well. The
Dc-exponent is dependent on the depth and the
porosity and will, during drilling, indicate anomalous
compaction trends caused by, for instance, uplift and
erosion if present. The uncertainty of the method is
mainly associated with the picking of representative
Fig. 2. Vitrinite reflectance profile well A, Hammerfest Basin. values from the Dc-curve and a deviation of about
156 B. Nyland, L.N. Jensen, J. Skagen, O. Skarpnes and X Vorren

400 m can be expected. Only the wells drilled in calculated. Effects caused by higher heat flow in tec-
the Hammerfest Basin have been examined by this tonically more active areas, and around salt diapirs
method. have been modelled and taken into consideration
in the evaluation. Based on the results from vitri-
nite reflectance profiles, a regional map showing the
Fission track analysis
amount of erosion has been constructed (Fig. 3).
The previously mentioned methods used in this The least uplift and erosion are seen in the western
study give an estimate on the amount of erosion, but part of the area; on the Senja Ridge and in the
they say nothing about the timing of this event. Ttoms0 Basin (0-500 m). In the Hammerfest Basin
Apatite is present as an accessory mineral in and on the Loppa High the values are between 1000-
sandstones and contains small quantities of ura- 1500 m. On the Finnmark Platform about 1500-2000
nium which decays radioactively by fission through m is absent, decreasing towards the south and east.
time. The fission process creates small defects in the The magnitude of uplift and erosion increase towards
crystal structure of the apatite (fission tracks) and the north and northwest, and more than 2000 m of
the length of these defects can be measured. The uplift and erosion are observed on the Stappen
density of tracks in an apatite grain is a function High and Bj0rn0ya. Data from And0ya also suggest
of the amount of uranium, the temperature, and erosion of the order of 2000 m.
the time spent below the annealing temperature. By The results from the use of compaction curves
measuring the density of tracks and the uranium con- seem to follow the results from the vitrinite re-
tent it is possible to estimate the age of the apatite. flectance profiles. The main difference is that the
Knowing the age of the apatite and the fission track amount of erosion is lower in the Hammerfest Basin
length distribution, it is possible to model the time (800-1000 m), but the differences are within the
and temperature history of the rocks containing the limits of confidence established for this study. As
apatite (Gleadow et al., 1983). with the vitrinite reflectance profiles, the shale com-
In this study the method has been used mainly to paction curves show an increase in magnitudes to-
establish the timing of the uplift. However, together wards north, reaching about 2000 m in the Fingerd-
with well temperature information an estimate of the jupet Subbasin.
magnitude of uplift can be given. The results from drilling parameters are in ac-
cordance with the results achieved using shale com-
paction curves and vitrinite reflectance profiles, with
Diagenes is of clay minerals an order of 800-1500 m erosion in the central part
of the Hammerfest Basin and decreasing erosion
The Crystallinity Index (CI) of authigenic illite, moving westwards in the basin.
calculated from XRD, can be used to evaluate the The mean Crystallinity Indices (CI) calculated for
level of diagenesis reached in sandstone and with wells in the Hammerfest Basin are in the range
that the depth of burial (Foscolos and Kodama, 1974; 12-17. These values correspond to a burial depth
Foscolos et al., 1976). The CI's measured from Lower of about 900-1100 m deeper than present day com-
to Middle Jurassic sandstones on the Troms0flaket pared with the Haltenbanken area. Most uplift and
have been compared with CFs from equivalent for- erosion is seen in the central part of the Hammerfest
mations that have undergone continuous burial in Basin (in the Sn0hvit and Albatross fields), decreas-
the Haltenbanken area. ing westwards towards block 7119/12 (300-400 m).
A source of error possibly influencing the results is Least uplift is observed in the wells drilled on the
the difference in temperature history between these Senja Ridge.
two areas. In addition, pore water composition and Based on a map of the Upper Regional Unconfor-
reaction kinetics are other factors that may influence mity (URU) combined with bathymetric maps (Vor-
on the illite formation (K. Gibbons, Statoil, pers. ren et al., 1989, 1990), the drainage system sourcing
commun., 1988). the three depositional areas in the Bj0rn0ya Trough,
Storfjorden Ttough and north of Svalbard has been
Amount of uplift constructed (Fig. 4). Volumetric calculations of the
fans indicate about 1.2 km of uplift and erosion in the
Vitrinite reflectance profiles and shale compaction area draining the southwestern part of the Barents
curves give the most consistent data coverage for Shelf, while a thickness of about 3.1 km sediments
estimates of uplift and erosion in the Barents Sea. have been removed from the Svalbard drainage area
Profiles, as shown in Fig. 2, have been constructed (Vorren et al., 1990). Calculations also indicate that
for all wells, and the thickness of the missing section the deposition rate increased from Miocene to Pleis-
Tertiary Uplift and Erosion in the Barents Sea: Magnitude, Timing and Consequences 157

10 12 14 16 18 20 22 24 26 28 30 32 34
75 75

74

73

72

71

70

km 200
MAGNITUDE
0-500m
H 500-1000 m
H 1000-1500 m
M 1500-2000 m
> 2000 m
Fig. 3. Postulated uplift and erosion. Based on vitrinite reflectance profiles.

tocene, possibly reaching a maximum in the Late and erosion in the Fingerdjupet Subbasin, and 1000-
Pliocene. 1300 m in the Hammerfest Basin.
Fission track analysis can also be used to some
extent to model the burial history, since the temper- Timing of uplift and erosion
ature at maximum burial is one of the parameters
given. Generally, the measurements show that most The timing of uplift and erosion is an important
of the samples have been exposed to temperatures factor controlling the generation and migration of
of 90-110C. The difference between the measured hydrocarbons. The generation of hydrocarbons will
formation temperatures in the wells and the pale- cease during uplift and the main phase of generation
otemperature given by the fission tracks can be used and migration will be at maximum burial.
to predict the uplift. Depending on the temperature The uplift and erosion history can be related
gradient used, this gives between 1500-2000 m uplift to two important regional events. These are the
158 B. Nyland, L.N. Jensen, J. Skagen, O. Sharpnes and T. Vorren

.Afiillllili''

/ ^ UPLIFT CENTRES j::: DEPOSITIONAL AREAS

WATERSHEDS - - - COASTLINE
^ RIVERS
Fig. 4. Reconstruction of the Neogene fluvial drainage pattern in the Barents Sea. Present bathymetry and geography are used as base
map.

structural development related to the opening of to the opening of the Norwegian-Greenland Sea in
the Norwegian-Greenland Sea (Eidholm et al., Oligocene time.
1984; Faleide et al., 1984; Spencer et al., 1984; The use of fission track analysis indicates two
Myhre and Eldholm, 1988), and the Late Cenozoic distinct periods of uplift and erosion (Fig. 5). The
glaciations (Matishov, 1980; Elverh0i and Solheim, first phase took place about 40-50 Myr ago, which
1983; Vorren et al., 1988, 1990; Eidvin and Riis, might have been in response to the early rifting of
1989). the Norwegian-Greenland Sea. The second phase
The lack of Late Tertiary sediments in all the occurred about 5-10 Myr ago. This phase might
wells except for those drilled on the Senja Ridge reflect isostatic uplift in response to glaciations.
(Spencer et al., 1984), have for a long time been Fission track modelling in wells also indicates that
used to explain the uplift and erosion as a response the magnitude of the first phase of uplift and erosion
Tertiary Uplift and Erosion in the Barents Sea: Magnitude, Timing and Consequences 159

CONCENTRATION OF WELL
EXTRACTED MATTER (ppm) LITHOLOGY mRKB
2000 4 000 6 000 800

:"HOT"SHALE
.

- -- T3 2 : ;
"OFF SCALE
"
J_? 1 "
ST0 FM

: GAS

;:.
2400
PALEO GOC-
Fig. 5. Apatite Fission Track modelling. Tertiary cooling in two GOC
phases of uplift and erosion from the southwestern Barents Sea. OIL
~~."^ r F SCALE
"OF " OWC
NORDMELA
was approximately 500 m greater than the later event | --
(Fig. 5).
However, modelling of fission track data from
Svalbard suggests a continuous uplift and erosion
since Early Oligocene time, and not two phases as
indicated for the south western part of the Barents 2500

Shelf.
Most of the works performed on the Cenozoic PALEO OWC p^^^^
T2 3 \
sequence in the Barents Sea have been based on
biostratigraphic datings of sequence boundaries in
the wedge penetrated by the wells on the Senja
Ridge (Fig. 1) (Spencer et al., 1984; R0nnevik, 1981;
Faleide et al., 1984; Henriksen and Richardsen, 1988;
> I : T 2 2-Z~_
Vorren et al., 1989, 1990). Redatings of sediments in
these wells (Eidvin and Riis, 1989) indicate that
the main part of the wedge (reaching about 4 km
in thickness) consists of Pliocene and younger sedi-
ments. The redatings of the wedge are supported by
Sr-isotope datings giving an age of 2.5-3 Myr (Eidvin
and Riis, 1989). This suggests that the main part of
the erosion took place during the glacial periods. Fig. 6. Paleo oil/water and oil/gas contacts in a Sn0hvit well
estimated from concentrations of extracted organic matter.

Consequences of uplift on hydrocarbon

exploration

Observations indicate that the hydrocarbon poten-

tial in the area has been significantly affected by
uplift and erosion. To date the exploration on the
Barents Shelf has been disappointing. But in almost
all wells drilled, good shows of residual oils have
been found indicating that oil has been generated in
the area. In wells where both gas and oil have been
Pi Pi
tested, high concentrations of extractable hydrocar-
bons are observed far below and above the present 1 OIL
.1.
contacts (Fig. 6). We suggest that gas expansion dur-
WATER
ing uplift caused oil to spill out of the structures as a Pi, Ti : Initial pressure and temperature
P, T : Reservoir pressure and temperature
result of pressure reduction, as outlined in Fig. 7. Pa, Ta : Actual pressure and temperature today

This model has been tested on the Sn0hvit field in Fig. 7. Phase change caused by pressure and temperature
the Hammerfest Basin, assuming that the major drop reduction. P{, T\ = initial pressure and temperature; P, T =
in the Q5+ concentration (Fig. 6) represents the pa- reservoir pressure and temperature; Pa, Ta = actual pressure
leo oil-water contact (PALEO OWC) at maximum and temperature today.
160 B. Nyland, L.N. Jensen, J. Skagen, O. Skarpnes and T. Vorren

200 150

TIME ( Myr )

Fig. 8. Burial graph based on well stratigraphy for the Sn0hvit field at reservoir level. Only one phase of uplift is used in the modelling.

B SN0HVIT AT MIN DEPTH B' SN0HVIT AT MIN DEPTH

(Calculated from max depth situation) (Calculated from present situation)

1750 i 1 1750r

0 200 400 600 800 1000 0 200 400 600 800 1000
HCPV, Mill. Rm3 HCPV, Mill. Rm3
Oil volume: 166.3 Mill.Rm3 Oil volume: 138.9 Mill.Rm3
Gas volume: 787.7 Mill.Rm3 Gas volume: 656.6 Mill.Rm3

A SN0HVIT AT MAX DEPTH SN0HVIT TODAY

0 200 400 600 800 1000 0 200 400 600 800 1000
HCPV, Mill. Rm3 HCPV, Mill. Rm3
Oil volume: 5 3 9 Mill.Rm3 Oil volume: 148 Mill.Rm3
Gas volume: 4 1 5 Mill.Rm3 Gas volume: 5 6 2 Mill.Rm3

Fig. 9. Volume calculations estimating the effect of uplift and erosion on the Sn0hvit field at different burial depths (see text for further
explanations).
Tertiary Uplift and Erosion in the Barents Sea: Magnitude, Timing and Consequences 161

burial. The low concentration of heavier hydrocar- from the Norwegian mainland during the glacial
bons (>Ci5+) in the upper part of the reservoir period (Riis and Fjeldskaar, this volume).
can be an indication of a former paleo gas-oil con- In a recent presentation, Jensen and Schmidt
tact (PALEO GOC) shallower than the contact seen (1990) demonstrated 1000-1500 m of Late Cenozoic
today (GOC). erosion in the near-shore area along the Skagerrak
The burial history at reservoir level in the well used coast and in Northern Jylland (Denmark) using the
(Fig. 8) is the input for the calculations where A is same methods as in this paper.
maximum burial, B represents the time of maximum Uplift may further have caused the tilting of the
uplift and C is the depth of the reservoir today. reservoir contacts seen in the TYoll field (Gray, 1987).
Knowing the hydrocarbon pore volume (HCPV) at Assuming that TYoll was at one time an oil field with
different depths to spillpoint, and the composition of a minor gas cap as suggested from the residual oil
oil and gas in the structure, is it possible to calculate present, the tilting may have provided more space
volumes at different depths. for the gas to migrate within the structure (Riis, pers.
To be able to run this model, some assumptions commun., 1989).
and simplifications have to be done and the restric-
tions in the modelling are: (1) The HCPV of the
Conclusions
reservoir has not changed during uplift, erosion and
renewed burial; (2) the composition of the oil and the The evidence for uplift and erosion with variable
gas is constant, and the oil is saturated at "bubble" magnitude on the Barents Shelf has been presented
point at any depth; (3) the pressure in the aquifer is and quantified by a number of methods. The magni-
constant (no overpressure); and (4) the temperature tude is in the order of 1000-1500 m in the southwest-
gradient has been constant through time. ern part, increasing to more than 2000 m towards the
Figure 9 summarizes the results from the volume Stappen High.
calculations from burial at maximum depth (A) to The timing of the uplift must be post-Paleocene.
minimum depth (B), and the same calculations per- Fission track data suggest two distinct periods with
formed for Sn0hvit today (C) to minimum depth uplift and erosion. The first situation might be corre-
('). All calculations are based on the paleocontacts lated with the opening of the Norwegian Greenland
indicated in Fig. 6. Sea, the second (and major event) with the post
In the first situation (A to B), the modelling shows Miocene glaciations. Both the fission track data and
that about 370 million Rm 3 oil have been expelled evaluation of the sedimentation rate from the wedges
from the structure, while the gas has expanded to indicate that the bulk of the erosion took place in the
almost twice the original volume. Late Neogene (Eidvin and Riis, 1989).
In the next case (C to B') modelling gives slightly Uplifts on the order of magnitude seen on the
lower volumes than the first calculation. However, Barents Shelf will have an effect on hydrocarbon
the difference in volumes between B and B' repre- accumulations, especially where reservoirs have a
sents only 3-4 m difference in depth to spillpoint. gas cap and a light oil at "bubble" point. In such
The decrease in gas volume from B/B' to C is within reservoirs the combined effect of gas expansion and
the expected range for gas diffusion loss during the evaporation from the oil will cause oil to be expelled
last 2 Myr. during uplift.
Our model further indicates that the Sn0hvit field Studies of uplift and erosion have so far been
would have contained gas only if the area had been neglected in other places on the Norwegian Shelf,
uplifted just a further 150 m. but increasing amounts of new data suggest that
there has been a major uplift and erosion, probably
Uplift and erosion on the Norwegian shelf as an isostatic response to the glacial overburden in
Late Cenozoic time. The effect is best seen in near
To date the major studies of uplift and erosion of shore areas, where the sequences are subcropping.
the Norwegian Shelf have focused on the Barents Tilting of the TYoll field during a late phase of uplift
Sea. However, evidence of an uplift in Late Cenozoic may explain the gas column and residual oil contact
time is observed along the entire Norwegian coast, seen today.
but very little has been published about the amount
and consequences of this uplift.
A Late Cenozoic wedge similar to that in the References
Barents Sea has been mapped in the Haltenbanken Bcrglund, L.T., Augustson, J., Faerseth, R., Gjelberg, I. and
area, and volumetric calculations suggest that these Ramberg-Moe, H., 1986. The evolution of the Hammerfest
deposits originated from about 1000 m of erosion Basin. In: A.M. Spencer (Editor), Habitat of Hydrocarbons
162 B. Nyland, L.N. Jensen, J. Skagen, O. Skarpnes and T. Vorren

on the Norwegian Continental Shelf. Norw. Pet. Soc, Graham Spitsbergen Tertiary. Nor. Polarinst. rbok 1977, pp. 159-
and Irotman, London, pp. 319-338. 177.
Dow, W.G., 1977. Kerogen studies and geological interpretations. Matishov, G.G., 1980. Geomorphological signs of the action of
J. Geochem. Explor., 7: 79-99. the Scandinavian, Novaya Zemlya, and Spitsbergen ice sheets
Eidvin, T and Riis, E, 1989. Nye dateringer av de tre vestligste on the floor of the Barents Sea. Oceanography, 20: 440-447.
borehullene i Barentshavet. Resultater og konsekvenser for Myhre, A.M. and Eldholm, O., 1988. The western Svalbard
den tertiasre hevingen. NPD Contrib. (preprint). margin. Mar. Pet. Geol., 5: 134-156.
Eldholm, O., Sundvor, E., Myhre, A.M. and Faleide, J.I., 1984. Nansen, F., 1920. En ferd til Spitsbergen. J. Dybwad, Oslo, 227
Cenozoic evolution of the continental margin off Norway pp.
and western Svalbard. In: A.M. Spencer (Editor), Petroleum N0ttvedt, A., Berglund, T, Rasmussen, E. and Steel, R., 1988.
Geology of the North European Margin. Norw. Pet. Soc, Some aspects of Tertiary tectonics and sediments along the
Graham and Trotman, London, pp. 3-18. western Barents Shelf. In: A.C. Morton and L.M. Parson
Elverh0i, A. and Solheim, A., 1983. The Barents Sea ice sheet, (Editors), Early Tertiary Volcanism and the Opening of the
a sedimentological discussion. Polar Res., N.S., 1: 23-42. NE Atlantic. Geol. Soc. London, Spec. Publ., 39: 421-425.
Faleide, J.I., Gudlaugson, S.T. and Jacquart, G., 1984. Evolution Olaussen, S., Gloppen, TG., Johannessen, E. and Dalland, A ,
of the western Barents Sea. Mar. Pet. Geol., 1: 123-150. 1984. Depositional environment and diagenesis of Jurassic
Feazel, C.T and Aram, R.B., 1990. Interpretation of discontinu- reservoir sandstone in the eastern part of Troms I area.
ous vitrinite reflectance profiles: Discussion. Am. Assoc. Pet. In: A.M. Spencer (Editor), Petroleum Geology of the North
Geol., Bull., 74: 91-93. European Margin. Norw. Pet. Soc, Graham and Trotman,
Foscolos, A.E. and Kodama, H., 1974. Diagenesis of clay min- London, pp. 61-80.
erals from Lower Cretaceous shale of northeastern British Riis, F and Fjeldskaar, W, 1992. On the magnitude of the Late
Columbia. Clays Clay Miner., 22: 319-335. Tertiary and Quaternary erosion and its significance for the
Foscolos, A.E., Powell, T.G. and Gnther, P.R., 1976. The uplift of Scandinavia and the Barents Sea. In: R.M. Larsen, H.
use of clay minerals and inorganic and organic geochemical Brekke, B.T Larsen and E. Talleraas (Editors), Structural and
indicators for evaluating the degree of diagenesis and oil Tectonic Modelling and its Application to Petroleum Geology.
generating potential of shales. Geochim. Cosmochim. Acta, Norwegian Petroleum Society (NPF), Special Publication 1.
40: 953-966. Elsevier, Amsterdam, pp. 163-185 (this volume).
Gabrielsen, R.H., Faerseth, R.B., Jensen, L.N., Kaiheim, J.E. R0nnevik, H.C., 1981. Geology of the Barents Sea. In. L.V.
and Riis, F., 1990. Structural elements of the Norwegian Illing and G.D. Hobson (Editors), Petroleum Geology of
continental shelf. Part I: The Barents Sea Region. NPD Bull., the Continental Shelf of North-West Europe. Institute of
6, 33 pp. Petroleum, London, pp. 395-406.
Gleadow, A.J.W., Duddy, I.R. and Lovering, J.F., 1983. Fission Sclater, J.G. and Christie, P.A.F., 1980. Continental stretching:
track analysis; a new tool for the evaluation of thermal An explanation of the post mid-Cretaceous subsidence of the
histories and hydrocarbon potential. APEA J., 23: 93-102. Central North Sea Basin. J. Geophys. Res., 85: 3711-3739.
Gray, D.I, 1987. Troll. In: A.M. Spencer (Editor), Geology of Spencer, A.M., Home, P.C. and Berglund, L.T., 1984. Tertiary
the Norwegian Oil and Gas Fields. Norw. Pet. Soc, Graham structural development of the western Barents Shelf: Troms
and Trotman, London, pp. 389-401. to Svalbard. In: A.M. Spencer (Editor), Petroleum Geology
Henriksen, E. and Richardsen, G., 1988. Cenozoisk geolo- of the North European Margin. Norw. Pet. Soc, Graham and
gisk utvikling av kontinentalmarginen i det s0rvestlige Bar- Trotman, London, pp. 199-209.
entshavet. Thesis, University of Troms0 (unpublished). Vorren, TO., Lebesbye, E., Henriksen, E., Knutsen, S.-M. and
Jensen, L.N. and Schmidt, B.J., 1990. Late Tertiary uplift and Richardsen, G., 1988. Cenozoic erosion og sedimentasjon
erosion in the Skagerrak area; magnitude and consequences. i det s0rlige Barentshav. XVIII Nord. Geol. Vinterm0te,
Nord. Geol. Vinterm0te, Stavanger, 19: 61 (abstract). K0benhavn, 1988, p. 424 (abstract).
Jordan, J.R. and Shirley, O.J., 1966. Application of drilling Vorren, TO., Lebesbye, E., Andreassen, K. and Larsen, K.-B.,
performance data to overpressure detection. J. Pet. Technol., 1989. Glacigenic sediments on a passive continental margin
18: 1387-1399 as exemplified by the Barents Sea. Mar. Geol., 85: 251-272.
Katz, B.J., Pheifer, R.N. and Schunk, D.J., 1988. Interpretation Vorren, TO., Richardsen, G., Knutsen, S.-M. and Henriksen, E.,
of discontinuous vitrinite reflectance profiles. Am. Assoc. Pet. 1990. The western Barents Sea during the Cenozoic In: U.
Geol., Bull., 72: 926-931. Bleil and J. Thiede (Editors), Geological History of the Polar
Manum, S.B. and Throndsen, T, 1978. Rank of coal and Oceans: Arctic versus Antarctic. NATO ASI Series C 308.
dispersed organic matter and its geological bearing in the Kluwer Academic Publishers, Dordrecht, pp. 95-118

B. NYLAND Norwegian Petroleum Directorate, P.O. Box 600, N-4001 Stavanger, Norway
L.N. JENSEN StatoilA/S, P.O. Box300, N-4001 Stavanger, Norway
J. SKAGEN StatoilA/S, P.O. Box40, N-9401 Harstad
O. SKARPNES StatoilA/S, P.O. Box40, N-9401 Harstad
T. VORREN University ofTroms0, P.O. Box3085, N-9001 Troms0
 163

On the magnitude of the Late Tertiary and Quaternary

erosion and its significance for the uplift of Scandinavia and
the Barents Sea
R Riis and W. Fjeldskaar

Geological evidence related to the Late Tertiary uplift and erosion of the Barents Sea and Scandinavia includes marine seismic
data, well data and field observations on land. The amount of erosion has been studied using geochemical and sedimentological
data from wells and topographical data and data on the block fields onshore. The amount of erosion has been compared to the
amount of deposition in the huge Pliocene-Pleistocene wedges situated offshore Norway and in the Barents Sea.
The data suggest that there is a close relation between the uplift and erosion of Scandinavia and the Barents Sea, and that
the uplift is related to two main events: (1) mountain building in Central South Norway and along the trend Lofoten-Bj0rn0ya-
Svalbard due to Palaeocene to Oligocene tectonic phases; and (2) glaciation in the Late Pliocene and Pleistocene, which was
related to plateau uplift of large areas.
Based on lithospheric flexural rheology parameters from studies of the post-glacial uplift, modelling shows that the
compensating isostatic uplift is close to 70% of the magnitude of erosion where crystalline basement is eroded. The magnitude
of the observed uplift in Scandinavia seems, however, to be greater than the erosion in some areas, and therefore mechanisms
other than pure isostatic compensation must be sought to explain this. One part of the uplift is thought to date back to
Paleogene tectonism, but observations suggest that there is an uplift in excess of these two components. We have investigated
the possibility that migration of phase boundaries in the lithosphere may be the mechanism of this additional uplift.
The response of a phase transition in the lithosphere to pressure relief caused by erosion leads to thickening of the upper
lighter phase and thus greater uplift than expected from isostasy alone. Migration of the phase boundary is a time-dependent
process. Theoretical models show that for load cycles of the order of 1 Ma, the surface deflections associated with the movement
of the phase boundary can be large, up to 50% of the isostasy. Therefore, modelling of the phase migration combined with
isostasy in response to erosion might lead to an uplift larger than the amount of eroded rocks. Thus, theoretically the phase
migration model is of interest, although it is not supported by independent geological evidence.

Introduction A similar wedge is situated on the shelf offshore

Mid-Norway (Fig. 2), although it contains a smaller
The late phase of erosion in the Barents Sea volume of sediments. This wedge has been pene-
significantly affected the petroleum accumulated in trated by several wells on Haltenbanken, and its
traps (e.g., Nyland et al., this volume). Also, in large age has been determined to be Late Pliocene and
parts of the platform areas, TYiassic and Jurassic Pleistocene. Thus, this age correlates with the wedge
source rocks situated in the oil window have been outside Bj0rn0yrenna.
uplifted and cooled. Therefore, much effort has been Results from ODP wells on the V0ring Plateau
put into the task of quantifying the uplift and erosion (Jansen et al., 1988) indicate that the first major
in the Barents Sea. glaciations took place at approximately 2.5 Ma. This
This study was initiated as a consequence of new age correlates well with the age dating of the wedges.
dating results of the Late Cenozoic sedimentary Examination of the coarse fraction of the cutting ma-
wedge penetrated by two wells on the Senja Ridge terial has yielded large amounts of angular fragments
(Fig. 1). Our results indicate the wedge to be of of crystalline basement in the wedges (Eidvin and
the Late Pliocene and Pleistocene age (Eidvin and Riis, 1989).
Riis, 1989). The size of the wedge corresponds to an In both areas, the Miocene and Early Pliocene
erosion of 500-1000 m in the Southern Barents Sea deposits appear to be thin, on average in the order
(Vorren et al., 1988; Vorren et al., 1990; N0ttvedt et of a few hundred metres at most. Thus, geological
al., 1988). evidence indicates that in the Late Pliocene and

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 163-185. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
164 F. Riis and W. Fjeldskaar

34
BATHYMETRY OF WESTERN BARENTS SEA
Contour Interval 100m (20m)
100km
-H
Fig. 1. Location map of Barents Sea, showing key wells and bathymetry. Map compiled by Y. Kristoifersen, B. Beskow, M. Sand and
Y. Ohta, Norsk Polarinstitutt (1989).
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 165

Fig. 2. (a) Time thickness map of Late Pliocene-Pleistocene wedges, Mid-Norway shelf.

Pleistocene period a dramatic increase in deposition glaciation, and whether the interpreted uplift related
rates along the shelf was related to glaciation of to erosion and glaciation can be separated from
the Barents Sea and of Scandinavia. The late phase uplift due to older tectonic phases related to the
of erosion in the Barents Sea must, therefore, be opening of the North Atlantic.
seen in a larger context. The observation that large
amounts of erosion were related to glaciations led Late Pliocene-Pleistocene geology
us to investigate the thermal and isostatic effects of
the rapid loading and unloading resulting from the Eros ion a I and depositional pattern
erosion process. The study is based on simplified
maps showing estimated amounts of erosion and Occurrence and dating of sedimentary wedges
deposition in Scandinavia and in the Barents Sea in Along the Barents Sea margin, large sedimentary
glacial times (the last 2.5 million years). wedges occur at the mouth of bathymetric troughs
The aim of this study is to determine whether (Fig. 1). The wedge fed by Bj0rn0yrenna is more than
the observed uplift pattern can be modelled using 4 km thick (Fig. 3), and the wedge off Storfjordrenna
simple assumptions about the effect of erosion and reaches the same thickness, although its areal extent
166 E Riis and W. Fjeldskaar

66

66

64

Mid Norway data base

62 c Seismic Line
-- Well 162

0 40 80 120 160 Km
R 3 U U=P=Q
Fig. 2 (continued). Maps (b) and (c) show location of wells and seismic lines and bathymetry.

is smaller. These two wedges are associated with depositional age of the sequence is Late Pliocene
large positive Bouguer gravity anomalies. Fjeldskaar and Pleistocene.
and Riis (1988) relate parts of these anomalies to The wedge off Storfjordrenna was penetrated by
a lack of isostatic compensation due to rapid recent one DSDP well in a distal position close to the
deposition. Knipovich Ridge (Fig. 1). The section was dated
Along the northern Barents Sea margin, a similar as Pliocene (Tklwani et al., 1976). The uppermost,
pattern appears. Sparse seismic data collected by mainly Quaternary part of the wedge can be corre-
NPD and by Baturin (1988) and regional bathymetry lated seismically to the wedge off Bj0rn0yrenna.
(Perry and Fleming, 1985) indicate that sedimentary In the Mid-Norwegian shelf, there are two main
wedges occur in a similar pattern, although probably wedges which are also linked to major bathymetric
with smaller sediment volumes. troughs the wedge at the mouth of Norskerenna
The wedge off Bj0rn0yrenna (Fig. I) was sampled and the "Mid-Norway wedge" at the mouth of
by the two wells, 7117/9-1 and 2, which penetrate ap- troughs extending from Vestfjorden and Ranafjorden
prox. 800 m Upper Pliocene and Pleistocene (Eidvin (Figs. 2 and 4). These two wedges have a maximum
and Riis, 1989). The dating of this sequence has been thickness in the order of 2000 m. Both have been
much debated due to resedimentation (e.g., Spencer penetrated by a number of wells (Fig. 2), and they
et al., 1984; N0ttvedt et al., 1988). However, the may be assigned a Late Pliocene-Pleistocene age
micropalaeontology, Sr isotope analysis and study of (Rokoengen and R0nningsland, 1983; Dalland et al.,
the basement fragments all strongly indicate that the 1988) (Naust Formation).
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 167

7 9
Fig. 2 (continued).

Estimates of erosion known, and for South Norway-South Sweden, the

In the Barents Sea and in Scandinavia, the amount area of deposition is not easily mappable. Further-
of erosion has been estimated by different methods. more, there is no major fault activity along the coast
For the Southern Barents Sea and for Mid-Norway, of Nordland post Triassic times. This means that the
it has been assumed that all erosion products which geometry of the uplift should be simple. However,
were transported to the west have been deposited Lofoten was an active structural element in the Late
in the respective offshore wedges. This mass balance Cretaceous and Paleogene.
gives an independent control on the erosion map,
although the drainage divides cannot be established (1) Barents Sea. The magnitude of erosion in
with certainty. We have concentrated on applying the Barents Sea was estimated by geochemical data
the method of mass balance on Mid-Norway-North (vitrinite reflectance and Tmax) from the deep wells
Sweden, because the uncertainty of the size of the (Fig. 5). In the western parts, the map is based
drainage area is not too great. For the southern Bar- on the occurrence of a seismic reflection which
ents Sea and Finnmark, the drainage area is less well cuts across the seismic stratification in the Late
168 F. Riis and W. Fjeldskaar

BJ0RN0YA

THICKNESS OF
PLIO-PLEISTOCENE WEDGE
BELOW UPPER UNCONFORMITY

Contour interval 1km

Poor data coverage

Limit of continental crust

14 16
Fig. 3. Thickness map of Late Pliocene and Pleistocene sediments below the Upper Regional Unconformity, southwestern Barents Sea.

Paleocene and Early Eocene layers. This reflection In the eastern and northern parts of the Bar-
shows a regional dip to the west (Fig. 6), and is ents Sea, there are very few data points (Antonsen
correlated to the transition from Opal A to micro- et al., 1991). Seismic stacking velocities (Norwegian
crystalline quartz (opal CT) observed in well 7117/ Petroleum Directorate, unpublished work) and geo-
9-1 (Ramberg Moe et al., 1988). This transition takes logical data from Svalbard (Manum and Throndsen,
place at approx. 50C, and the dip of the reflector 1978) suggest a general decrease in erosion from
indicating the diagenetic change may be explained by west to east in this area.
tilting and erosion after its formation. The deepest It is well known that Spitsbergen was uplifted
level observed for the reflector was assumed to be and eroded as a consequence of Eocene-Oligocene
the level of no erosion, and the erosion in the rest of tectonism (e.g., Steel and Worsley, 1984). Also, the
the area was estimated by the difference in elevation, Bj0rn0ya-Stappen High area was uplifted along the
assuming no lateral change in temperature gradient. Kn0legga Fault in different tectonic phases in the
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 169

62

62

60
60

58c
58

Contour interval 100m

0 40 80 120 160 Km
ty y y y y^y^y
Fig. 4. Bathymetry of the North Sea. Based on data from Statens Kartverk.

Eocene and Oligocene (Fig. 7). This happened also year (L0v0 et al., in preparation), and at 0.5-1 mm/
to the Senja Ridge and Lofoten islands farther south year for active glaciers (Elverh0i et al., 1980, 1983).
(Brekke and Riis, 1987). As a consequence, the 1000 m of erosion is equivalent to 0.4 mm/year,
large amount of erosion observed in the western and suggesting that the Barents Sea was not submerged
northern parts of the Barents Sea can be related for much of the glaciation period.
partly to tectonic uplift in the Paleogene and partly
to the glacial effects we are concerned with in this (2) Mid Norway. The calculated erosion is based
study. on the assumption that a pre-glacial surface can
Farther east and south in the Barents Sea, the be mapped onshore in Scandinavia. If the present
Early Tfertiary tectonic effects can be expected to be height and topography of the pre-glacial surface
more subdued. Thus, an estimate of 800-1000 m of can be established, the amount of erosion can be
Late Pliocene-Pleistocene erosion based on the sed- calculated by subtracting the present mean heights
iment volume in the wedge off Bj0rn0yrenna seems from the height of the pre-glacial level.
reasonable for the central parts of the Southern Block fields are observed on many of the high
Barents Sea. mountains in South Norway. Nesje et al. (1988) sug-
It should be pointed out that these values are gest that the block fields were formed by weathering
surprisingly high compared to studies of present day processes prior to glaciation. It is assumed that the
erosion rates at Svalbard, estimated at 0.2-0.5 mm/ lowest limit of the block fields approximates a pre-
170 F. Riis and W. Fjeldskaar

SOUTHERN BARENTS SEA

Thickness of rocks eroded in
the Late Tertiary-Quaternary

Deep well

Fig. 5. Map of southern Barents Sea showing interpreted amount of Tertiary erosion (in metres). This interpretation is based on
vitrinite and pyrolysis data from deep wells as well as on Figs. 6 and 12, profiles EE1 and FF'.

glacial surface. In M0re, these pre-glacial surfaces calculated by extrapolating the eroded part of the
may be traced almost to sea level. sedimentary sequence to the shore. The pre-glacial
Fig. 8 shows a simplified map of mean heights pinch out line of sedimentary rocks on the crystalline
contoured from an 8 x 8 km grid prepared by the basement was inferred by using the profiles AA' to
Geographical Survey of Norway and the Geographi- FF' (Fig. 8), and is shown in Fig. 11. The pinch out
cal Survey of Sweden (1963). The summit level map line was drawn so that the block field regions were
(Fig. 9) is based on smoothed heights of the block not given a pre-glacial sediment cover.
fields where they exist, combined with summit levels A comparison was made between the deposited
in areas where there are few data on block fields or volume of sediments in the wedge off Mid-Norway
where block fields do not occur. This map may be and the eroded volume in Mid-Norway suggested in
regarded as an approximation of a pre-glacial sur- Fig. 11, using tentative drainage divides as shown on
face, but as it does not take into account topography Fig. 9. The result is given in Table 1. The eroded
in this surface, it could be regarded as a pre-glacial thicknesses suggested in Fig. 11 are in the order of
surface of maximum probable elevation. 1.5-2 times too large to fit into the wedge. This sup-
A. Nesje (University of Bergen, unpublished work)
has compiled the paleic surface in Norway (Fig. 10).
The interpretation is based on geomorphological TABLE 1

field work in Southern and Mid-Norway, and mainly Erosion and accumulation Mid-Norway wedge
on map studies in Northern Norway. This map may
Area Volume
be interpreted as an approximation of the present (km 2 ) (km 3 )
day level of the pre-glacial topography, and it must
Glacial wedge (Fig. 2) 139000 100000
be regarded as a better approximation than the Eroded sedimentary rocks (Fig. 11) 88600 64900
summit level map. Eroded basement (Fig. 11) 108000 89300
Onshore erosion (Fig. 11) was calculated by sub- Corrected basement (2.7/2.2) 108000 109000
tracting the mean height map (Fig. 8) from the Total erosion: 173900 km 3
summit level map (Fig. 9), and therefore it must Total deposition: 100000 km 3
be regarded as a maximum case of eroded rock in The assumptions are seismic velocity of 2200 m/s and density of
glacial times. In the offshore areas, erosion has been basement/density of sediment = 2.7/2.2.
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 171

ports the idea that the paleic surface map (Fig. 10) correlates fairly well with the erosion interpreted in
may be a better approximation of the pre-glacial the Barents Sea, given that there was not a marked
surface than the summit level. If this is the case, the topographic break between these areas prior to the
topography before glaciation may have been quite onset of major erosion.
rough even in Nordland, with local relief in the order The summit level map (Fig. 9) and the cross-
of 500-700 m. In this case, a "summit level" surface sections AA' to F F , combined with a regional
must be interpreted as older than Neogene. Alterna- overview (Fig. 13), show that there are some com-
tive ways of explaining the volume difference would mon structural features related to the uplifted Scan-
be to change the drainage divides and to reduce the dinavia and Barents Sea.
amount of eroded sedimentary rocks. - The sequence of sedimentary rocks surrounds
Our present knowledge of Tertiary tectonic events the coast from Skagerrak to the White Sea with
and thickness of the Oligocene to Early Pliocene a fairly uniform dip away from the mainland. The
section on the shelf, suggests fairly low deposition boundary between sediments and crystalline base-
rates and, thus, a fairly low relief at the onset of ment is at a fairly regular distance from the coast line
glaciation. It seems that more detailed studies are and commonly coincides with a bathymetric low.
needed to resolve this problem. - The base of Late Pliocene (glacial) sediments
In conclusion, Fig. 11 is a maximum case for the combined with the surface of the summit level to-
amount of material eroded in the glacial period, and gether form an asymmetric dome structure with the
the volume could be reduced by 30-40%. Even so, steep side facing offshore. The structuring of the
the amount of erosion in the time of glaciation is base of the glacial sediments shows that the dome
higher than was expected. formation is recent. The dome has two maxima in
North Sweden and in South Norway and in map
Structural pattern view it turns around the coast (Figs. 9 and 13). The
boundaries of the dome do not seem to be faulted
Six profiles were constructed across Scandinavia in to a large extent, although they coincide with the
different structural positions (Fig. 12). Profiles AA' boundary faults along some segments of the coast.
and DD' describe the situation with a fairly large - The maximum amount of erosion seems to occur
uplift onshore, no obvious tectonic activity along in a zone along the coast. The extremal value is com-
the margin and, for profile DD', a thick wedge of monly located close to the boundary of sedimentary
Plio-Pleistocene sediments deposited on the shelf. cover and crystalline basement.
Note that for those profiles the summit level can The pattern of a structural dome in Scandinavia
be projected into the sedimentary sequence, sug- may be regarded as part of a geomorphological/
gesting a pinch out of sediments in a position structural pattern of concentric highs and lows, as
presently onshore before erosion and uplift took pointed out by White (1972). The following observa-
place. tions may be of importance.
Profile BB' shows a large uplift onshore, tectonism - The exposed crystalline shield is geographically
offshore, and a small Plio-Pleistocene wedge. On- linked to the areas of glaciation.
shore, the paleic surface can be traced almost to sea - A central shallow depression is found in Bot-
level, fitting the position of the base of the wedge tenvika (Fig. 13), (even when allowing for a residual
offshore. This supports a pre-glacial, but Tertiary age glacial rebound of approx. 100 m, as proposed, e.g.,
of the paleic surface. by Balling, 1980).
Profile CC traverses the anomalously low TYond- - Marginal highs occur concentrically, including
heim area, where Jurassic sediments are known in a Scandinavia, Kola and the hills at the Finland-Soviet
fault-bounded basin in Beitstadfjorden (B0e, 1989). boundary. The outcrop of the Cretaceous in Den-
Here the summit level projects into the basement- mark and Scania indicates that even in the south
sediment boundary offshore. Accordingly, the sed- there has been a large amount of erosion and uplift.
imentary cover in this profile has been extended - In the periphery outside the marginal highs
farther onshore, to give the onshore dome a shape we find shallow marginal morphological depressions.
similar to the other sections. The sediments re- Offshore, they include the bathymetric lows along
moved were included in the calculation of erosion the coast of Norway and Kola, the White Sea and
(Fig. 11). the Gulf of Finland. Onshore, they include the great
Profile EE" and profile FF' have been constructed lakes of Vnern, Vttern, Ladoga and Onega (Fig.
by adding a sedimentary cover in the Barents Sea 13). This zone of marginal lows is connected to
according to Fig. 5. The profiles suggest that the the boundary between crystalline basement and sedi-
summit level inferred from TYoms and Finnmark mentary cover.
 E Riis and W. Fjeldskaar
172

i i
/ I TWT MAP
Opal A to CT transformation
Western Barents Sea
Countour Interval 100ms
/ \ Salt
/ I __ Highs, Jurassic level
f 7117 9-2 J
^ Subcrob plio-pleistocene wedge
1
s ^ o
; /7117 9-1 07119/7 1 7119/9-1

/ /
Fig. 6. (a) Map of two-way travel time to seismic anomaly representing the Opal A to CT transition in the Barents Sea.

- The sedimentary cover surrounding the glaciated Sea ice sheets were developed on thick deposits of
area in general dips away from the exposed core of sedimentary rocks and the present topography differs
basement rocks. from the other areas by its low elevation 100-400 m
White (1972) noted such a concentric geomorphic below sea level. However, in this context, Svalbard,
pattern in Fennoscandia and compared it with a Bj0rn0ya, Franz Josef Land and Novaya Zemlya may
similar pattern in North America. He connected this be regarded as marginal highs.
pattern to glaciation, but his explanation involved It is concluded that the development of the con-
unrealistically high values of erosion (Sugden, 1976). centric pattern of highs and lows is related to glacia-
It may briefly be commented that a similar pattern tion both in time and space, and hence we suggest
of a central depression, marginal highs and huge sed- that its development should be explained in terms of
imentary wedges occurs even in connection with the effects related to glaciation itself.
Greenland ice sheet (Weidick, 1976). The Barents
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 173

i /
v TERTIARY EROSION
Western Barents Sea

-h Contour Interval 50m

Salt
t i ~ Highs, Jurassic level

S
7117 9-2 J
~"~" Subcrob plio-pleistocene wedge
o
0 5 01 Km 7119/7-1 7119/9-1
/ 7117 9-1
y o

Fig 6. (continued), (b) A conversion of the map to show the eroded thickness of the sediment.

Modelling of crustal response tribute to the uplift. This is the migration of phase
boundaries in the lithosphere in response to pressure
The removal of load in eroded areas and ad- relief caused by erosion.
ditional loading in the areas of deposition causes In addition, rifting in the North Atlantic causing
isostatic response of the Earth's lithosphere. The iso- uplift in Scandinavia (Sales, 1989) is regarded as a
static response is the most obvious mechanism for part of the story. The modelling presented in this
the uplift/subsidence in the study area. The magni- paper is an attempt to calculate the magnitude of
tude of the observed uplift in some areas, however, the effects expected from glacial erosion, so that
seems to be greater than the erosion. The isostatic these fundamentally different tectonic events can be
process alone cannot explain an uplift of this magni- quantified and separated.
tude. An additional process is needed. Here we have
looked at one possible candidate which may con-
 Vestbakken Volcanic Province STAPPEN HIGH

NW KN0LEGGA FAULT SE

20KM
_J

Fig. 7. Cross-section across the Kn0legga Fault southwest of Bj0rn0ya indicating motion in the Eocene and post-Eocene. Line BV-12-86. BQ = base of Quaternary; BUTp = base of Upper
Pliocene; ITej and 2 = intra Eocene; Pj = Permian; BP = base of Permian. From Gabrielsen et al. (1990).
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 175

V y " 4 *5 * I - 7 * I V *10 I*I 12

* 15
* 14
15
* ' 17
* 18
* 1' 2' 0 21
* 22
* 23
* 24
* 25
* 26
* 2' 7 2B
" IV *JO J*l 32
* 33
* 34
" 35
* 36
* 37
* 3' 8 39
*

10 02
< " i -7 ' 1 /
is= 7- 'I ' 10
' 12* ' 13 - 14 ' 15 1
/l 17
'i" 18 -'-I
" U
f\ * 22H* 23 1' 2 4\
' 20 2 ' 25 * 26 " 27 * 28 2 " 30 * 31

Fig. 8. Mean height map of Scandinavia, based on contouring of 8 x 8 km grid values.

Isostasy sphere and partly by the buoyant forces of the

asthenosphere. Loads of short wavelengths will be
The isostatic process is simulated by a non- balanced by the lithosphere, so that the lithosphere
spherical three-dimensional model, in which the acts as a low-pass filter. The characteristics of this
lithosphere is a uniformly thick elastic layer overlying filter depend on the flexural rigidity. In this study we
a viscous asthenosphere. With this model, we are used a value for the flexural rigidity of D = 1023 Nm.
able to treat the isostatic problem analytically using A study based on data on the present rate of uplift
the Fourier transform technique. The mathematical and postglacial sea level changes has revealed a value
model is described in Appendix A. of the flexural rigidity of Scandinavian mainland of
A (positive or negative) load applied to the Earth's less than 1024 Nm (Fjeldskaar and Cathles, 1991).
surface will be partly balanced by the elastic litho- The results of the isostatic response reported here
176 F. Riis and W. Fjeldskaar

3 2 I 0 1 2 3 4 5 6 8 97 01 I 21 31 41 51 6 71 61 91 02 12 22 32 42 52 28
* 27
* 28
" 29
" 30
* 3 32
* 33
" 34
' 35
" 36
* 37
" 38
" 39
*

! UU^\i\\VY\'1

Fig. 9. Summit level map of Scandinavia. Constructed from height of block fields (Nesje et al., 1988) and by drawing isolines through
the highest summits.

(Fig. 14) can, therefore, be considered as maximum 1989) are not taken into account.
values. The resulting isostatic adjustments of the
The relaxation time for the isostatic flexural com- Earth's surface due to Pleistocene/Pliocene erosion/
pensation estimated from the postglacial uplift is deposition are shown in Fig. 14a. The modelling was
only a few thousand years (Fjeldskaar and Cathles, based on the erosion model presented in Fig. 13 and
1991). Thus it is reasonable to assume that the flex- the parameter values are given in Tkble 2.
ural equilibrium was achieved during the erosional Calculation based on the amounts of erosion in-
process. As the time period considered here is less ferred show that the isostatic uplift of Late Pliocene/
than 5 million years, the viscous properties of the Pleistocene age amounts to 900-1400 m in the west-
lithosphere (as described by Fjeldskaar and Pallesen, ern Barents Sea, and the subsidence of the sediment
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 177

Contour interval 100 (500)m ^WJ / ^

Fig. 10. Map of altitude of paleic surface in Norway. Compiled by Atle Nesje (University of Bergen, unpublished work). 500, 1000 and
1500 m contours are reinforced.

wedge is 700-1800 m. Onshore, the most severe ero- Phase boundary migration
sion took place in the Nordland area, which caused
an uplift of 200-800 m, depending on the lithosphere It has been proposed (e.g., Kennedy, 1959) that
rigidity of the area. phase transitions in the Earth's upper layers provide
178 F. Riis and W. Fjeldskaar

. ,t . , - j . 7 - - 5 - 4 - j - 2' - 0* *2 3* 4* 5* *7 10* 12* U* 14* 15* 1 17* U* IT 2 22* 23" 24' 25" 2' 27' 21* 1 3d' 3 32* 33" 34" 35' 36* 37* IV 39*
*

Fig. 11. Smoothed map showing difference between summit levels and mean height.

TABLE 2

Parameter values

Young's modulus () 10.3 x 10 10 N m - 2

Poisson's ratio () 0.25
Lame's parameter (,) 3.34 x 10 10 Nm
Asthenosphere density 00 3.3 x 103 kg/m3
Inverse slope of Clausius Clapeyron curve W 75 K/kbar
Geothermal gradient () 12K/km
Density of upper phase (Pi) 2.8 x 10 3 kg/m3
Density of lower phase in) 3.2 x 103 kg/m3
Bulk modulus (K) 5.56 x 10 10 Nm
Partly based on O'Connell and Wasserburg (1967).
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 179

/''
.7^\. - 1 0 0m0 ^ /
- -"" \
^ - ICOO
m

/ "--- ^__ B

BUF>
/
_V7
I 1
s / /$H 12
o 1 - 1 0m0 0

*// eo? - 2 0m0 0

, 8 0 4, ,,. BCJ
- 3 0m
00

- 4 0m
00

- 5 0m0 0

--<^ <
~~^ c

8
o
// '*
B C/
1

LEGEND

- 1 000m
/^^ -~ ~ ^ ^SS-^-^ O F F S HEO R O N S HEO R

> / o
5 ^^^^. D Top of e r odd es e d i m
t en
Sea Floo
r * v ' Y
'
Q Q u a t ye r n a r \
m
- 1 0 0m BU
P Bea sU prp eP l i oec e n
BU
P ^
- BC Bea s C r e t a sc e o u
2 0m
00

7" " 18 ;' 19 " 20

Fig. 12. Structural cross-sections showing summit level (solid line), mean height (dashed line), basement outcrop and Pliocene-
Pleistocene deposits. For location of profiles, see Fig. 8.

a mechanism of uplift and subsidence of the Earth's and numerical approximations of the movement of
surface. This is based on the assumption that phase the Earth's surface by phase boundary migration
transitions at depth respond to pressure changes at have been established.
the surface. Subsidence at the surface occurs when The phase transition considered here is the transi-
an increase in pressure (by sedimentation) causes tion from gabbro to eclogite at the base of the crust
the upward migration of the phase boundary. Uplift at a depth of 30-50 km. Migration of this phase
would follow a decrease in pressure (by erosion) boundary has been suggested as an explanation for
causing downward motion of the phase boundary. the formation of sedimentary basins (e.g., Mareschal
The mechanism has been studied by several authors and Lee, 1983).
(e.g., O'Connell and Wasserburg, 1967, 1972; Gjevik, We have adopted an equilibrium approximation,
1972; O'Connell, 1976; Mareschal and Gangi, 1977; since it is easily shown, using the analytical approxi-
Mareschal and Lee, 1983). In these works, analytical mations of O'Connell (1976), that the phase bound-
180 E Riis and W. Fjeldskaar

Erosion
Base map

Contour interval 1km x l g / c m 3

20 30 40 50
Fig. 13. Contoured grid showing values of erosion in the area covered by the Scandinavian and Barents Sea ice sheets in the glacial
period. Erosion is given in terms of thickness (km) multiplied by density. The map is based on computer contouring of grid values.
Contour interval: 1 km x 10 kg/m3. Areas of sedimentation are given with negative values.

ary migration will be 90% compensated in 3 million Discussion of modelling results

years, using the parameter values given in Tkble 2
and a phase transition temperature of 1000 K. The The calculation of crustal response was based on a
results shown in Fig. 14b are based on an assumption grid representing the thickness of material eroded in
of 50% phase transformation in the area. In spite the glacial period multiplied by its density (Fig. 13),
of this, we consider the resulting surface adjustments as well as the amount of sediment accumulations in
to be maximum values. The mathematical model is the offshore wedges (negative numbers).
formulated in Appendix B. For the modelling, it was necessary to input infor-
We are aware that field studies of eclogites indicate mation for the total area of glaciation and erosion
that granulites may be metastable for a long time, and, therefore, very tentative values have been given
and eclogitization may take place only along shear in the eastern areas on which we have no informa-
zones where fluid has moved (Jamtveit et al., 1990). tion. However, errors in the eastern parts will not
The modelling of surface adjustments caused by seriously affect the results in Norway and the Barents
possible changes in phase boundaries shows that this Sea.
effect may be significant, of the order of 500 m The contoured values in the western Barents Sea
of uplift in the western Barents Sea, and similar do not coincide with data in Fig. 5, as we have
amounts of subsidence in the depositional wedge. tentatively tried to separate the last (glacial) phase
Approximately 300 m of the uplift in the Nordland from the Paleogene phase of uplift and erosion.
area may be explained by phase changes in the crust/ The three-dimensional modelling results indicate
mantle boundary. that the uplift of the Scandinavian mountains should
The resulting adjustments of the Earth's surface be comparable with the amount of erosion. The
due to movements of the phase transition are shown areas of sediment accumulation subside substantially,
in Fig. 14b. The calculation was based on the erosion/ allowing space for continued rapid deposition.
deposition model in Fig. 13, and the parameter In the Barents Sea and other areas where sed-
values given in Tkble 2. iments are removed, the net effect of the erosion
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 181

Modelled isostatic uplift

Contour Interval 200m

50

Modelled uplift due

to phase transition

Contour Interval 50m

50

Fig. 14. Three-dimensional modelling of crustal response in the Scandinavia-Barents Sea areas, (a) Uplift and subsidence due to
isostatic adjustment, using a flexural rigidity of 1 x 10 23 Nm. The contour interval is 200 m. (b) Uplift and subsidence as a result of
phase boundary migration, using a crustal thickness of 35 km and a density contrast of 0.4 x 103 kg/m 3 . The contour interval is 50 m.

process is lowering of the surface. In the southern the elevation must have been higher. Similar results
Barents Sea, the model predicts that removing 1200- are obtained from the southern part of Sweden and
1400 m of sediments (assuming a density of 2.2), Denmark, where mainly sediments were removed.
results in an uplift of 600-800 m (compare Figs. 13 The three-dimensional modelling gives results
and 14). Hence, a pre-glacial elevation of 300-400 m which simulate quite well the observed structural
above sea level is needed to explain the present el- pattern with the marginal highs and peripheral lows.
evation. In the western and northern Barents Sea, The marginal highs are formed where dense base-
182 F. Riis and W. Fjeldskaar

2000m
BEFORE EROSION

h 1000m

Crystalline Basement

Marker bed
-1000m

L -2000m

2000m
AFTER EROSION AND DEPOSITION OF WEDGE

1000m

-1000m

-2000m
Fig. 15. Two-dimensional modelling of crustal response along profile DD'. The shape of a simplified pre-glacial topography is shown
after erosion and deposition. The modelled results are compared with the observed profile DD'. For location, see Fig. 8.

ment rocks are eroded, and they are slightly dis- observed, leaving a difference in the order of 500 m.
placed towards the central parts of the ice sheet Offshore, the subsidence was calculated by adding
from the area of maximum erosion. The periph- the isostatic and phase change effects to the effect of
eral lows seem to form where the erosion is at a compaction of the underlying sediments. For the cal-
maximum, while the subsidence/uplift is small due culation of compaction, we used a general porosity
to the hinge line position at the boundary between trend based on well data from Haltenbanken. Fig-
crystalline basement and sedimentary cover. Where ure 15 demonstrates that the subsidence modelled is
sediments only are eroded, the model predicts sur- slightly smaller than that observed. However, the fit
face lowering, as is observed. In the central depres- is better than for the onshore part.
sion of Bottenvika, subsidence is predicted, and ob- There are many possible explanations for the dis-
served. crepancy observed. One way out of the problem
The low elevation of the Trondheim area can be could be to fit the onshore part by assuming that
explained by the assumption that it was covered the pre-erosional elevation was much higher than
by sedimentary rocks considerably lighter than the 250 m. This explanation is supported by the paleic
basement before onset of erosion. surface map, Fig. 10. However, the low sedimenta-
However, the total theoretical adjustment is not tion rates on the shelf from the Oligocene through
in accordance with observation everywhere. Adjust- Early Pliocene suggest that the relief was not high at
ment due to isostasy and phase change was calculated this time. This points in the opposite direction.
on profile DD' crossing Vega, Nordland, extending It is thought that some anomalies can be explained
from the shelf edge to Bottenvika (Fig. 12). This by assuming a pre-erosional high relief. The best
profile is regarded as a cross-section in an area example is South Norway. Figure 10 indicates that
which was not affected strongly by Paleogene tec- the paleic surface is elevated to more than 1500 m
tonic events, and the uplifted dome as well as the between Sognefjorden and the high peaks of Jotun-
depositional wedge are well developed. heimen. These high peaks represent a topography
The results of calculation (Fig. 15) show that the where the summit level is elevated several hundreds
isostatic and phase change effects may comprise al- of metres above the paleic surface. In the North
most the amount of erosion in the onshore area. Sea and M0re Basin, thick Paleogene sedimentary
The calculated uplift is considerably smaller than wedges are observed with a progradational direction
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 183

roughly from the east. Thus, there is independent U N (F k )

evidence that the uplift of the highest mountains pga(k)
in South Norway took place largely prior to the
where g is gravity, p is density of the mantle, and a(k)
Pliocene.
is a parameter characterizing the flexure response of
Hence, separation of the Paleogene uplift from
the lithosphere. The lithosphere flexure depends on
the Plio-Pleistocene is problematic. Our calculations
the elastic strength of the lithosphere and a(k) is
indicate that an important part of the concentric
given by the equation (Cathles, 1975):
morphology can be explained by rapid erosion and
deposition of the material in the glacial periods. 2k
However, if the calculated uplift is removed from [S2 - (kH)2] + (CS + IcH)]
the present topography, we will still be left with a a(k) _ PS
S+kHC
concentric morphology, although more subdued. We
consider it unlikely that plate tectonic effects related where S is sinh kHy C is cosh kH, and is Lame's
to the opening of the Atlantic in the Paleogene parameter.
should create such a topography. Also, the defor- The mechanical thickness of the lithosphere, H, is
mation of the base of the Upper Pliocene and even given by the equation:
Quaternary layers (Fugelli and Riis, in press) into 1/3
the regional dome structures is such that it cannot be \2{\-v2)D
H =
explained totally using the present model.
where D is flexural rigidity, E is Young's modulus,
Conclusions and v is Poisson's ratio.

- The uplift of Scandinavia and the Barents Sea

Appendix B
is related to Paleogene thermal effects due to the
North Atlantic rifting and opening, and to isostatic Theoretical formulation of phase boundary mi-
adjustments due to Pliocene/Pleistocene glaciation gration
and erosion. Let us first, for the purpose of illustration, consider
- The models presented do not explain satisfacto- a one-dimensional case assuming a phase boundary
rily the total magnitude of the observed uplift of the at depth in the mantle, separating two phases of
Scandinavian mountains, although they simulate well the same component. Suppose the temperature at
the observed structural pattern. which the phase change takes place (Tc) is related
- Observation and modelling suggest that the ero- to the pressure by the following equation (Clausius-
sion effects are very important in the study of recent Clapeyron equation):
structures in glaciated areas worldwide.
Tc = T0 + 7PC = T0 + jgpiz
Acknowledgements
To is the transition temperature at the Earth's sur-
face, p\ is the density of the upper phase, z is the
The authors are grateful to the Norwegian
depth of the phase boundary and 7 is the inverse
Petroleum Directorate for supporting this study. We
slope of the Clausius-Clapeyron curve.
also thank our colleagues in the NPD who have
Consider a uniform erosion of a half-space which
taken part in the discussion on the uplift problem.
undergoes a phase change at depth z p . The decrease
Further, we acknowledge the important contribution
in hydrostatic pressure due to the erosion is AP. If
to this paper by Dr. Atle Nesje, Bergen University.
the geothermal gradient is initially zero, the phase
boundary will migrate downward until the pressure
Appendix A decrease is cancelled. The equilibrium migration of
the phase boundary is:
Isostatic model
The lithosphere is modelled as a thin elastic plate AP
=
overlying an incompressible fluid substratum, ex- PS
pressed in the Fourier domain with the initial topo-
graphic load F(k) and the resulting flexural defor- This migration will cause an uplift of the Earth's
mation ho(k) written as a function of wavenumber k. surface of:
The equilibrium isostatic compensation is achieved AP
= p
by subsidence (Cathles, 1975): Ap + p
184 F. Riis and W. Fjeldskaar

where /> is the density difference between the Eidvin, T. and Riis, F , 1989. Nye dateringer av de tre vestligste
upper and lower phases. borehullene i Barentshavet. Resultater og konsekvenser for
den tertiaere hevingen. NPD Contrib., 27, 43 pp.
In this study we use an analytical approximation
Elverh0i, A., Antonsen, P., Flood, S.B., Solheim, A. and Vullstad,
for the equilibrium position of a phase boundary A.A., 1988. The physical environment Western Barents Sea
under the horizontally varying surface loads given by 1:1500000. Shallow bedrock geology. Nor. Polarinst. Skr. 179
Mareschal and Gangi (1977). Under the assumptions D.
that the Earth behaves as an elastic solid above the Elverh0i, A., Liest0l, O. and Nagy, J., 1980. Glacial erosion, sed-
phase boundary, that the equilibrium position does imentation and microfauna in the inner part of Kongsfjorden,
Spitsbergen. Nor. Polarinst. Skr., 172: 33-61.
not depend on deviatoric stress and that the two Elverh0i, A., L0nne, 0 . and Seland, R., 1983. Glaciomarine
phases have the same thermal properties, they found sedimentation in a modern fjord environment, Spitsbergen.
that the vertical displacement of the phase boundary Polar Res., 1: 127-149.
due to the applied load P(k,a) is in the Fourier Fjeldskaar, W. and Cathles, L.M., 1991. Rheology of mantle and
domain: lithosphere inferred from postglacial uplift in Fennoscandia.
In: R. Sabadini et al. (Editors), Glacial Isostasy, Sea Level and
Mantle Rheology. Kluwer Academic Publishers, Dordrecht,
S(k) = Y(k)-P{k)a) pp. 1-19.
Fjeldskaar, W. and Riis, F , 1988. Modellering av tertiaerbasseng
where: p den vestlige marginen av Barentshavet. In: K. Binzer
(Editor), Abstracts, 18. Nordi. Geol. Vinterm0de, K0benhavn,
1988. Danmarks Geologiske Unders0gelse.
exp(-|fc|) Fjeldskaar, W. and Pallesen, S., 1989. The application of a
(jgpi ~ )
Y(k) = viscoelastic lithosphere model to isostatic subsidence in back-
OLK ka[sinh(2ka) - 2ka] stripping. In: J. Collinson (Editor), Correlation in Hydro-
1 + carbon Exploration. Graham and Trotman, London, pp. 5 3 -
(A + /i). a D(ka)
59.
Fugelli, E.M.G. and Riis, F in press. Neotectonism in the
and:
Jasren area, southwest Norway. In: L.N. Jensen and F Riis
i2 (Editors), Post-Cretaceous Uplift and Sedimentation Along
+ 3
+ cosh2y +y2 the Fennoscandian Shield. Nor. Geol. Tidsskr.
[X + \ A + /i Gabrielsen, R.H., Faerseth, R.B., Jensen, L.N., Kaiheim, J.E.
and Riis, F , 1990. Structural elements of the Norwegian
_ continental shelf. Part I: The Barents Sea Region. NPD Bull.,
uK _ p2 - pi \ a 6, 33 pp.
a Geographical Survey Office of Sweden, 1963. Mean height map
Pi JgPi ~ of Sweden. Rikets Allmnna Kartverk, Meddelande, Ser. A,
No. 30.
7 is the inverse slope of Clausius-Clapeyron curve,
Gjessing, J., 1967. Norway's paleic surface. Nor. Geol. Tidsskr.,
is the geothermal gradient, is the density of the 21: 69-132.
upper phase, p2 is the density of the lower phase, Gjevik, B., 1972. Surface adjustment due to a subcrustal phase
a is the depth of the phase change, K is the bulk transition. Phys. Earth Planet. Inter., 5: 403-408.
modulus, and A and are Lame's parameters. Jamtveit, B., Bcher, K. and Austrheim, H., 1990. Fluid con-
trolled eclogitization of granulites in deep-crustal shear zones,
Bergen Arcs, Western Norway. Contrib. Mineral. Petrol., 104:
References 184-193.
Jansen, E., Bleil, U., Henrich, R., Kringstad, L. and Slettemark,
Antonsen, P., Elverh0i, A., Dypvik, H. and Solheim, A., 1991. B., 1988. Paleoenvironmental changes in the Norwegian Sea
Shallow bedrock geology in the Olga Basin area, northwestern and the Northeast Atlantic during the last 2.8 Ma: DSDP/
Barents Sea. Am. Assoc. Pet. Geol., Bull., 75(7): 1178-1194. ODP Sites 610, 642, 643 and 644. Paleoceanography, 3: 563-
Balling, N., 1980. The land uplift in Fennoscandia. In: N.A. 581.
Mrner (Editor), The Earth Rheology, Isostasy and Eustasy. Kennedy, G.C., 1959. The origin of continents, mountain ranges
J. Wiley and Sons, New York, N.Y., pp. 297-321. and ocean basins. Am. Sei., 47: 491-504.
Baturin, G.N., 1988. Dokl. Akad. Nauk SSSR, 299, No. 2. L0V0 V., Elverh0i, A., Antonsen, .., Solheim, A., Butenko, G.,
Brekke, H. and Riis, E, 1987. Tectonics and basin evolution Gregersen, O. and Liest0l, O., 1990. Submarine permafrost
of the Norwegian shelf between 62N and 72N. Nor. Geol. and gas hydrates in the Northern Barents Sea. Nor. Polarinst.,
Tidsskr., 67: 295-321. Rapportser., 56, 171 pp.
B0e, R., 1989. The structural development of the Mesozoic Manum, S. and Throndsen, T, 1978. Rank of coal and dispersed
rocks in Beitstadfjorden, Ed0yfjorden and Frohavet, Central organic matter and its geological bearing in the Spitsbergen
Norway. Terra Abstracts (Eur. Union Geosci.), 1: 22-23. Tertiary. Nor. Polarinst., rbok 1977, pp. 159-177.
Cathles, L.M., 1975. The Viscosity of the Earth's Mantle. Mareschal, J.C. and Gangi, A.F, 1977. Equilibrium position of
Princeton University Press, Princeton, N.J, 386 pp. a phase boundary under horizontally varying surface loads.
Dalland, A., Worsley D. and Ofstad, K., 1988. A lithostrati- Geophys. J.R. Astron. Soc, 49: 757-772.
graphic scheme for the Mesozoic and Cenozoic succession Mareschal, J.C. and Lee, C.K., 1983. Initiation of subsidence in
offshore mid- and northern Norway. Norw. Pet. Directorate, a sedimentary basin underlain by a phase change. Geophys.
NPD Bull., 4, 65 pp. J.R. Astron. Soc, 74: 689-712.
On the magnitude of the Late Tertiary and Quaternary erosion and its significance for the uplift of Scandinavia 185

Nesje, A., Dahl, S.O., Anda, E. and Rye, N., 1988. Block fields Norwegian Petroleum Society Workshop, 18-20 October 1989,
in southern Norway: significance for the Late Weichselian ice Stavanger.
sheet. Nor. Geol. Tidsskr., 68: 149-169. Spencer, A.M., Home, P.C. and Berglund, L.T., 1984. Tertiary
Nyland, B., Jensen, L.N., Skagen, J.I., Skarpnes, O. and Vorren, structural development of the western Barents Shelf: Troms
T., 1992. Tertiary uplift and erosion in the Barents Sea; to Svalbard. In: A.M. Spencer et al. (Editors), Petroleum
magnitude, timing and consequences. In: R.M. Larsen, H. Geology of the North Atlantic Margin. Graham and Trotman,
Brekke, B.T. Larsen and E. Talleraas (Editors), Structural and London, pp. 99-109.
Tectonic Modelling and its Application to Petroleum Geology. Steel, R.J. and Worsley, D., 1984. Svalbard's post-Caledonian
Norwegian Petroleum Society (NPF), Special Publication 1. strata an atlas of sedimentational and palaeogeographical
Elsevier, Amsterdam, pp. 153-162 (this volume). evolution. In: A.M. Spencer (Editor), Petroleum Geology of
N0ttvedt, A., Berglund, T, Rasmussen, E. and Steel, R., 1988. the North European Margin. Graham and Trotman, London,
Some aspects of Tertiary tectonics and sediments along the pp. 98-107.
western Barents Shelf. In: A.C. Morton and L.M. Parson, Str0m, K.M., 1948. The geomorphology of Norway. Geogr. J.,
(Editors), Early Tertiary Volcanism and the Opening of the 62: 19-27.
NE Atlantic. Geol. Soc. London, Spec. Publ., 39: 421-425. Sugden, D.E., 1976. A case against deep erosion of shields by
O'Connell, R.J., 1976. The effects of mantle phase changes on ice sheets. Geology, 4: 580-582.
postglacial rebound. J. Geophys. Res., 81: 971-974. Talwani, M., Udintsev, G. et al., 1976. Initial reports of the
O'Connell, R.J. and Wasserburg, G.J., 1967. Dynamics of the Deep Sea Drilling Project, Vol. 38. U.S. Government Printing
motion of a phase change boundary to changes in pressure. Office, Washington, D . C , 1256 pp.
Rev. Geophys. Space Phys., 3: 329-410. Vorren, TO., Lebesbye, E., Henriksen, E., Knutsen, S.M. and
O'Connell, R.J. and Wasserburg, G.J., 1972. Dynamics of sub- Richardsen, G., 1988. Cenozoisk erosjon og sedimentasjon i
mergence and uplift of a sedimentary basin underlain by det s0rlige Barentshav. Abstr., XVIII Nord. Geol. Vinterm0te,
a phase change boundary. Rev. Geophys. Space Phys., 10: K0benhavn, 1988, 424 pp.
335-368. Vorren, TO., Lebesbye, E., Andereassen, K. and Larsen, K.-B.,
Perry, R.K. and Fleming, H.S., 1985. Bathymetry of the Arc- 1989. Glacigenic sediments on a passive continental margin
tic Ocean. Map sheet, scale 1:4704073. Naval Research as exemplified by the Barents Sea. Mar. Geol., 85: 251-
Laboratories, Washington D.C. 272.
Ramberg Moe, H., Roaldset, E. and Gjelsvik N., 1988. Siliceous Vorren, TO., Richardsen, G., Knutsen, S.-M. and Henriksen, E.,
sediments and seismic response a correlation tool in ex- 1990. The western Barents Sea during the Cenozoic. In: U.
ploration on Norwegian shelf. Norwegian Petroleum Society, Bleil and J. Thiede (Editors), Geologic History of the Polar
Correlation in Hydrocarbon Exploration, 3-5 October 1988 Oceans: Arctic versus Antarctic. NATO AS I Series. Kluwer
(abstract). Academic Publishers, Dordrecht, pp. 95-118.
Rokoengen, K. and R0nningsland, T.M., 1983. Shallow bedrock Weidick, A., 1976. Glaciation and the Quaternary of Greenland.
geology and Quaternary thickness in the Norwegian sector In: A. Escher and W Stuart Watt (Editors), Geology of
of the North Sea between 6030'N and 62N. Nor. Geol. Greenland. Geological Survey of Greenland, Copenhagen,
Tidsskr., 63: 83-102. pp. 430-458.
Sales, J., 1989. Late Tertiary uplift West Barents shelf. White, W.A., 1972. Deep erosion by continental ice sheets. Geol.
Character, mechanisms and timing. Paper presented at the Soc. Am. Bull., 83: 1037-1056.

F. RIIS Norwegian Petroleum Directorate, P.O. Box 600, N-4001 Stavanger, Norway
W. FJELDSKAAR Rogaland Research Institute, P.O. Box 2503, Ullandhaug, N-4004 Stavanger, Norway
 187

Pleistocene glacial isostasy implications for petroleum

geology

A. Kjemperud and W. Fjeldskaar

The theory of glacial isostasy evolved in the second half of the 19th century and by the beginning of the 20th century it was
fully accepted. The onshore data in Fennoscandia were overwhelming with traces of marine influence 200-300 m above the
present shoreline. Very few studies have, however, been performed in offshore areas. Only recently, Quaternary research has
reported phenomena interpreted as a result of glacial isostatic movements. Among petroleum geologists the effect has been
almost neglected, and no public reports on how glacial isostasy may effect hydrocarbon bearing rocks (source or reservoir) exist.
Data from the Quaternary and latest Tertiary research have been compiled to establish a stratigraphic framework for the last
2-3 million years. Realistic ice models have been made which incorporate the observations made on three different sizes of
ice sheets in northern Europe. These ice sheet models form the input for the modelling of isostatic deformation of the crust.
The earth is modelled by a non-spherical viscous fluid, overlain by a uniform thick elastic lithosphere. The isostatic problem is
treated using the Fourier transform technique.
The main product of the modelling is isobase maps describing the equilibrium deformation of the crust during the different
glacial events. The model is calibrated using Fennoscandian sea-level curves after the last glaciation. The maps are therefore
supposed to be valid also in the offshore parts of northern Europe including shelf areas with considerable oil and gas
accumulations.
At least once and perhaps several times during the last 2-3 million years the central North Sea has been downwarped and
subsequently uplifted by up to 400 m, on the Haltenbanken up to 250 m and in the Barents Sea as much as 650 m. The vertical
movements have been differential in the sense that the areas overlain by thick ice domes have been downwarped more than
peripheral areas. The modelled gradients are more than 1.0 ra/km in graben areas of the North Sea and parts of Haltenbanken
and about 1.3 m/km in the western parts of the Barents Sea.
The differential movements will have affected the entire crust including the sediments with their fluid content. The geometric
change during the maximum downwarping may have important implications for the secondary migration of hydrocarbons.
Hydrocarbons in traps filled to the spill point will start to spill during the isostatic movements and today the structure may be
only partly filled. This effect can be severe in flat and in areally large structures. Our calculations show that as much as 30% of
the total hydrocarbon volume can be lost during maximum glacial isostatic downwarping.

Introduction The theory of glacial isostasy as an explana-

tion for the marine deposits found high above the
It is generally accepted that crustal layers have a present shoreline in Scandinavia was generally ac-
tendency to approach a position of hydrostatic equi- cepted in the late 19th century when De Geer (1888,
librium. Dutton (1889) proposed the term isostasy 1890) published his famous work on the uplift of
for this process. However, there are different mech- Fennoscandia. A full understanding of the interplay
anisms by which this may be achieved. Pratt (1855) between glacial isostasy and eustasy was, however,
postulated that mountains are composed of rocks not reached before the work of Nansen (1922, 1927)
which are less dense than average, while Airy (1855) and Ramsey (1924). Nansen (1922) was also the first
suggested that mountains have roots, i.e., material of to point out the possibility of a peripheral forebulge
uniform density which extends deeper into the un- around an ice sheet
derlying denser material beneath the mountain than Large isostatic adjustments have occurred in
beneath the lowlands. Fennoscandia and adjacent areas in the last 2-3
Realizing that the Earth's mantle is not rigid, million years. The cause of these adjustments is
Jamieson (1865) postulated that the load of an ice extensive glaciations with ice caps up to several kilo-
cap will deform the surface of the earth, giving rise metres thick in some areas. A number of glaciations
to glacial isostasy. have taken place during the Quaternary, but the

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 187-195. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
188 A. Kjemperud and W. Fjeldskaar

isostatic movements will in principle be the same

for all, although the magnitude and duration will be
different.
Since the mid-sixties an extensive exploration for
oil and gas has taken place in the North Sea, and
somewhat later in the Haltenbanken area and in the
Barents Sea. It has long been clear that these areas
have been completely or partly covered by glaciers
at least once and perhaps several times during the
Quaternary. However, very little attention has been
paid to the effects of glaciations on hydrocarbon
generation, migration and trapping. The objective of
this study is to examine the implications of glacial
isostasy for petroleum geology.

Pleistocene glaciations in Northern Europe

Timing of glaciations
The initiation of northern hemisphere glacia-
tions has been a subject of debate. Based on early
DSDP results several authors (Laughton et al., 1970;
Berggren, 1972; Berggren and Van Couvering, 1974)
defined the onset of glaciations to approximately 3
Ma. Dating of glacial marine deposits from the Ama-
rasian basin in the central Arctic ocean (Clark et
al., 1980) gave approximately 5.3 Ma for the first
ice-rafted material. The chronostratigraphy of this
report has, however, been questioned (Grantz et a l ,
1982).
Based on results from Pacific cores (Shackleton
and Opdyke, 1976; 1977) suggest initiation of north-
Fig. 1. Isotope records and deduced glacial events and eustatic
ern hemisphere glaciation at 3.2 Ma, with a severe curve (isotope curve from Schackleton and Opdyke, 1977)
deterioration at 2.5 Ma. A correlation of the oxygen
isotope stage zonation of four detailed Pleistocene
oxygen isotope records from the Pacific (V28-239, on material from DSDP and ODP cores from the
DSDP 504), Caribbean (DSDP 502B) and the North North Atlantic and the Norwegian Sea. This is
Atlantic (DSDP 552) (Williams 1988) show a good based on oxygen isotope analyses, magnetostratig-
correlation of nine cold events in the last 0.8 million raphy and observation of ice-rafted debris. Several
years and in the time interval between 2.2 Ma and severe glaciations took place in the period 2.6-2.0
0.8 Ma as much as 17 cold events are interpreted Ma. Between 2.0 and 1.2 Ma the glaciations in Scan-
in one or all three areas. We have used the most dinavia were small, while a transition towards larger
complete oxygen isotope curve from the Pacific (Fig. glaciations took place during the period 1.2-0.6 Ma.
1), which is shown to correlate with the North At- In the last 0.6 million years the climate has been
lantic curves (Williams, 1988), to construct a eustatic more variable with short warm interglacial periods
curve (actually a curve showing the ocean water vol- between period with extensive glaciations seen as
ume change given in metres of global change, cf., pulses of ice-rafted material in the cored material.
Fjeldskaar, 1989) which may represent the magni- The oxygen benthic isotope record in the Norwegian
tude and importance of the glacial events in the last Sea (borehole ODP 644 A) shows relative changes
2 million years A calibration of 0.11%o 6uO per 10 of up to \%o (occasionally 2 units), which represent
m of sea-level change has been used (Fairbanks and approximately 100 m (occasionally 200 m) of eustatic
Matthews, 1978). change, using the same method as for the Pacific
A recent investigation (Jansen et al., 1988) con- core.
cluded that a major advance of the Scandinavian
ice sheet took place at approximately 2.6 Ma based
Pleistocene glacial isostasy implications for petroleum geology 189

The extent and thickness of ice sheets An ice distribution like this has been proposed for
The regional extent and the thickness of ice sheets the last ice age by Grosswald (1980) and Andersen
are the two controlling factors for glacial isostasy. (1981, maximum alternative). There is, however, an
From the stratigraphical record it seems to be certain ongoing discussion as to whether the North Sea and
that the North European area has been influenced the Barents sea were glaciated during the last ice age
by both small and large ice sheets during the last 2-3 (see discussion in Boulton, 1979; Grosswald, 1980;
million years Andersen, 1981; Solheim and Kristoffersen, 1984;
Ice sheets can be reconstructed on the basis of Sejrup et al., 1987; and Vorren et al. 1989).
analogy with observational data from recent sheets It can, however, be concluded that both areas
and traces left behind by ancient sheets. The ice were glaciated once or most probably several times
sheets used in this study have been reconstructed by during the Pleistocene. Sejrup et al. (1987) conclude
the method of Hughes (1981). The most important that in the North Sea the central area (Fladen
input for determining the size of the ice-covered area Ground) was overridden by ice several times in the
has been the reported observations of dated or age time interval 1-0.7 Ma. The area was also glaciated
estimated end moraines, while ice sheet gradients sometime between 0.2 and 0.13 Ma (the Saalian).
from recent sheets are the major input for estimating In the Barents Sea glacigenic processes have been
ice thickness and volume. active at least four times during the Pleistocene and
As a basis for our calculations of isostatic effects, as far west as the shelf break according to Elverh0i
three different ice sheets have been reconstructed. and Solheim (1987) and Vorren et al. (1989).
They are all based on different interpretations of In the Haltenbanken area marginal moraines sit-
observations from the last glaciation. The three ice uated at the shelf edge have been mapped by An-
models thus represent realistic geometries which dersen (1979). Between 64N and 66N the marginal
most probably closely resemble the real situation in moraines are found approximately west of 6, which
one or several of the previous glacial events. means that the entire shelf area most probably was
covered by ice during the last glacial event. Since
Ice model I (Fig. 2). The ice sheet is characterized the Haltenbanken is situated relatively close to the
by three large ice domes. One over Novaya Zemlya mountainous area and central Fennoscandia, even
in the East Barents Sea reaching 3200 m, one in minor glacial events could have covered the shelf.
central Fennoscandia of the same thickness and one
in northern Great Britain reaching 2000 m. These Ice model II (Fig. 3). A large ice dome in central
three domes represent the growth centre of the ice Fennoscandia (thickness 3200 m) is isolated from
sheets which has coalesced in the saddle areas in the an ice dome in northern Great Britain (thickness
south east Barents Sea and in the central North Sea.

Fig. 3. Ice model II. Based on the minimum model of Andersen

Fig. 2. Ice model I. Based on the maximum model of Andersen (1981) for the Barents Sea and on Sejrup et al. (1987) for the
(1981). Isopacks are each 400 m. Outer line 400 m. North Sea. Isopacks are each 400 m. Outer line 400 m.
190 A. Kjemperud and W. Fjeldskaar

1600 m) and one in Svalbard (thickness 1200 m). is achieved 20000 years after loading/unloading) is
Most of the North Sea and the entire Barents Sea modelled. The isostatic model is then rather simple,
are ice free. The ice cover in the Haltenbanken area as the important parameter is only the lithospheric
is the same as in ice model I. The model is based rigidity.
on Andersen's (1981) minimum alternative for the
Barents Sea, and that of Sejrup et al. (1987) for the Lithosphere modelling
North Sea. If an ice load is applied to a fluid, the surface of
This model can represent the last glaciation (We- the fluid will deform until the weight of the fluid
ichselian) in the North Sea. Data from the western displaced from the equilibrium level balances the ap-
part of the North Sea and from Scotland (Suther- plied load (local isostatic compensation). If an elastic
land, 1984) indicate that the British ice sheet did lithosphere covers the fluid the applied load will be
not reach much further east than today's land area. supported partly by the lithosphere and partly by the
This supports the conclusions made by Sejrup et al. buoyant forces of the fluid beneath acting through
(1987) that the British and the Scandinavian ice did the lithosphere (regional isostatic compensation). If
not coalesce in the Fladen Ground area (Sleipner the size of the load is small compared to the litho-
Field). spheric thickness, the base of the lithosphere is not
deformed at all and the entire load is balanced by
Ice model III. This model is similar to model II the lithosphere. If the load is large, the lithosphere
except for a reduced ice thickness in the mountain supports none of the load, and the deformation at
areas in the eastern part of Norway according to the the base of the lithosphere is the full isostatic defor-
theory of Nesje et al. (1988). This isostatic modelling, mation.
however, gives results very similar to those obtained The lithosphere thus acts as a lowpass filter. The
when using model II (Tkble 1). characteristics of this filter depend on the elastic
A detailed history of outbuilding and downmelting strength of the lithosphere. A measure of the elastic
of the ice sheets has not been applied in any of strength of the lithosphere is a parameter called the
the models. Both processes are modelled as being flexural rigidity. The elastic strength of the litho-
instantaneous. sphere is a function of mechanical thickness and is
determined by the following equation:
Methods of glacial isostatic modelling
EH3
Flexural rigidity
B y D = ^
The Earth is modelled by a non-spherical vis- 12(1 - i/2)
coelastic fluid in which the viscosity may vary with
depth, overlain by a uniform thick elastic lithosphere. where H is elastic thickness, E is Young's modulus,
With this flat earth model, we are able to treat the v is Poisson's ratio.
isostatic problem analytically using the Fourier trans- The regional isostatic equilibrium compensation
form technique. The method used is developed by due to a harmonic load F(x) is achieved by subsi-
Cathles (1975). dence (Cathles, 1975):
The Fourier transform allows a function such as
h F(k)
the ice sheet configuration to be expressed as the sum
pga(k)
of harmonic components of different wavelengths.
The asthenosphere is treated as a viscous fluid For a local compensation model the subsidence is:
in which the rate of displacement varies with the
wavelength of the harmonic component of the load.
The elastic lithosphere is treated as a low-pass filter, pg
because loads of small size tends to be balanced by
the lithosphere itself and not by buoyancy. where F(k) is Fourier transformed ice load, p is
density of the upper mantle, and g is gravity.
Mantle viscosity The "lithosphere filter" is:
Data on the present rate of uplift and sea-level
changes after the last glaciation show that the mantle ^(S2 - k2H2) + (CS + kH)
is of low viscosity (Fjeldskaar and Cathles, 1991a, b).
In calculations of the time-dependent isostatic
deflections, a detailed glacial history must be taken where S = sinh kH; C = cosh kH, H is mechanical
into account. This is not done in the present study. thickness of the lithosphere, is Lame's parameter,
Only the situation at isostatic equilibrium (which and k is wavelength.
Pleistocene glacial isostasy implications for petroleum geology 191

In this study we use a mechanical thickness of INTERGLACIAL

the lithosphere equal to 70 km (flexural rigidity of
50 x 123 Nm). This is based on a study by Fjeldskaar DEPOSITION
and Cathles (1991a) concerning data on the present
rate of uplift and sea-level changes after the last
glaciation in Scandinavia (Kjemperud 1986).
The model presumes a uniformly thin lithosphere.
This is of course a simplification. The spatial varia- Flow
+ ASTHENOSPHERE
tion of the lithosphere thickness is, however, difficult =1.3 1(rupoise
to estimate. Moreover, the effects introduced by this "?t -J?-
presumption are probably of second-order only.
MESOSPHERE
The flexural rigidity for the shelf areas is not n=1.0x10 22poise
known, the study mentioned above is based on data
from the Scandinavian mainland. It is, however,
CORE
reasonable to assume the same rigidity as for Scandi-
navia, because the shelf areas also have continental Fig. 4. Geophysical model of the earth as used in this study.
crust. The large sediment thickness on the shelfs Geological and geophysical processes during interglacial periods
does not influence the flexural rigidity; the elastic (left) and glacial periods (right) are shown.
thickness of the lithosphere is almost 70 km, while
the sediment package is only 5-10 km thick. 600 m are reached by applying ice model I. In the
The geophysical model of the Earth used in this North Sea the maximum values exceed 400 m and on
study is shown in Fig. 4. The geological and geo- Haltenbanken it reaches 250 m.
physical processes associated with periods of glacia- The largest gradients, 1.3 m/km east-west, are
tions/deglaciations are also shown. These processes seen in the western Barents Sea. However, the other
are isostasy, eustasy, flow in asthenosphere, glacial areas also show values above 1.0 m/km when applying
erosion and deposition. ice model I.
Because of the lack of a deglaciation history,
Results from isostatic modelling the time-dependent forebulge is not represented
in the modelling results (Table 1). A transient
The results from the isostatic modelling are given model has, however, been applied in the North
in the isobase maps (Figs. 5 and 6) and in Table 1. Sea and in the Haltenbanken; indicating that the
The largest downwarping of the lithosphere seen is forebulge reaches a maximum just after the deglacia-
in the eastern Barents Sea where values of more than tion and is somewhat higher than the values shown

TABLE 1

Isostatic effects in key areas of the Norwegian shelf

Ice models Central North Viking Halten Barents Sea Barents Sea
graben graben banken west of 2 2 east
Ice model I
Ice cover (m) 1200 to 1600 1200 to 1600 400 to 800 400 to 1200 1200 to 2400
Isostasy (max) (m) 400 450 230 400 650
Forebulge no no no no no
Gradients N-S -1.1 1.2 0.6 0.4 0.3
Gradients E - W -0.5 0.4 1.0 1.3 0.8
Ice model II
Ice cover (m) 0 0 400 to 800 0 0
Isostasy (max) (m) - 2 5 to +70 110 250 - 1 0 to +160 - 3 0 to +160
Forebulge yes no no yes yes
Gradients N-S -0.4 0.2 0.8 0.9 0.8
Gradients E - W -0.3 0.8 1.0 0.5 0.3
Ice model III
Ice cover (m) 0 0 0 to 800 0 0
Isostasy (max) (m) - 5 to +40 50 230 - 1 0 to +160 - 3 0 to +160
Forebulge insignificant no no yes yes
Gradients N-S -0.3 0.1 0.7 0.9 0.8
Gradients E - W -0.3 0.5 1.1 0.5 0.3
192 A. Kjemperud and W. Fjeldskaar

Fig. 5. Modelled isostatic effect of ice model I. Isobases each Fig. 6. Modelled isostatic effect of ice model II. Isobases each
100 m. 100 m.

Table 1. The height of the forebulge does not seem in the gas-filled part, limited by the uppermost fossil
to exceed 60-70 m. HCWC.
The effect of the glacially induced hydrocarbon
migration is determined by several factors:
Implications for hydrocarbon migration and - The gradient of the deformation. The gradient
trapping is dependent on the geometry of the ice sheet. The
result observed today is the effect of the ice sheet
The theory of glacially isostatic induced hydrocar- creating the highest gradients through time.
bon migration combines the glacial isostatic effects - The number of isostatic events. Tilting will be the
with the fact that liquid interfaces are equipotential result of each isostatic event. A number of isostatic
surfaces. In normal pressure areas they are parallel events will create overlapping and chaotic patterns
to the geoid. Figure 7 shows how a simple structure of fossil contacts and residual oil which will be very
behaves during one glacial cycle. It is assumed that difficult to interpret.
the structure is filled with hydrocarbons to the spill - The orientation of the long axis of the hydro-
point at the onset of vertical movements. carbon-filled structure. The direction of the length
When the lithosphere and the structure are down- axis of the field will greatly influence the volume
warped differentially due to glaciation, hydrocarbons changes during the isostatic events. The differential
start to spill out of the structure. This goes on until downwarping will be highest when the longest axis
the maximum tilt situation is reached (Fig. 7b). The is parallel to the main gradient. A 30 x 5 km field
closure volume is now reduced proportional to the directed along a 1 m/km gradient give a differential
differential movements. tilt value of 30 m, while the perpendicular orientation
When the lithosphere and the structure are up- would give a differential value of only 5 m.
lifted to the starting position due to deglaciation, a - The initial fill of the trap. An incompletely filled
new hydrocarbon-water contact (HCWC) is estab- trap could possess enough freeboard to accommo-
lished (Fig. 7c). The freeboard which is the vertical date for the differential changes. The freeboard height
difference between the HCWC and the spill point is the difference between the spill points of the trap
will be defined by the magnitude of the differential at maximum uplift and maximum downwarping.
movement. - The spill point orientation. If the spill point is
Fossil HCWCs can theoretically be found at the situated in the area of least relative downwarping the
spill point level for the two extreme positions. Resid- entrapped volume will be reduced and vice versa. In
ual oil can be found in the water zone down to the light of the isobase maps (Figs. 5 and 6) fields in the
original spill point level. In a gas-filled reservoir with Barents Sea with spill points in the west and north,
an oil leg originally, residual oil can also be found and fields on Haltenbanken with spill points towards
Pleistocene glacial isostasy implications for petroleum geology 193

WELL1 WELL 2 WELL 3

UNDER
MAXIMUM
TILT
b

INTERVAL OF POSSIBLE
RESIDUAL HYDROCARBONS

Fig. 7. Geometric changes of schematic reservoir during one cycle of glacial isostatic deformation.

the north/west will get a reduced closure volume A hypothetical example from the Barents Sea
during glaciation. In the northern North Sea a spill By using a realistic geological model for the last
point in the north and east will give reduced closure 0.3 million years we have calculated the changes in
volume, while in the southern North Sea a spill point closure volume in a hypothetical field in the western
in the east and south will give the same effect. Barents Sea. Three isostatic events are seen in this
- The geometry of the receiving area. The geometry period of time, i.e., the Elsterian (event 6 in Fig. 1)
of the receiving area can give several scenarios for the Saalian (event 4) and the Weichselian (event 2).
the fluids migrating out of the trap. If a saddle The field has the following characteristics:
area is opened, the fluids can migrate into a higher
lying trap or be lost to the surface. Alternatively, a Dimensions: length, 30 km; width, 5 km
surrounding flat area can become a part of the trap Orientation: east-west
during the downwarping and fluids will migrate back Spill point: west
into the trap under uplift. Initial freeboard: 0m
194 A. Kjemperud and W. Fjeldskaar

TABLE 2

Glacial isostatic induced migration in a hypothetic oil/gas field in the western Barents Sea

Glacial event Vertical movement Diff. vertical movement HC-column Freeboard Relative
(ice model) (maximum) (m) of spillpoint (m) (m) (m) volume
Initial 0 _ 100 0 1.0
During 6 (II) -160 -15 - 0 1.0-0.9
After 6 +160 +15 92 7 0.9
During 4 (I) -300 -40 - 0 0.9-0.7
After 4 +300 +40 80 20 0.7
During 2 (I) -300 -40 - 0 0.7
Present -250 30 85 15 0.7
The calculations are based on isostatic compensation for the two first events, and a remaining uplift of 50 m after the last.

Fluid: gas and oil Summary and conclusions

Initial volume: 1
Initial HC column: 100 m Northern Europe including the continental shelf
Field geometry: bar triangular in section areas has been covered by extensive ice sheets at
least once and perhaps several times during the last
The results of the calculations based on the isobase 2-3 million years
maps of ice model I (for event 2 and 4) and ice model Three ice sheet models have been established
II (for event 6) are given in Table 2. based on stratigraphic and geomorphologic data
from the North European mainland and shelf ar-
During event 6 (Elsterian) the field is tilted to the eas.
south and east with gradients of 0.9 and 0.5 m/km, The most extensive ice sheet model covers both
respectively (Tkble 1). This gives a differential ver- the entire North Sea and the Barents Sea with
tical movement of 17 m and a loss in volume of ice thicknesses in excess of 1500 and 2000 m, respec-
approximately 10%. tively. The least extensive ice sheet model covers only
During event 4 (Saalian) the change is more dras- parts of the shelfs, except for in the Haltenbanken
tic. The gradients are 1.3 m/km towards the east area where the shelf edge is the ice limit. All models
and 0.4 m/km towards the south. The differential fulfil the onland data of glacial uplift after the last
vertical movement will be 40 m and the total loss glaciation.
of volume will now add up to nearly 30% compared The isostatic response to the ice load is modelled
to the initial value. The field will reach a maximum by a flat earth model and the Fourier transform
downwarping of 300 m during this event. technique. Isobase maps for the ice sheet models
During event 2 (Weichselian) there will not be have been calculated. In the shelf areas the largest
any dramatic effect because a high freeboard is isostatic effect, more than 600 m, is found in the
established after event 4, and although the isostatic eastern Barents Sea when the most extensive ice
effect is of similar magnitude during event 2 the model is used. The highest deflection gradients are
implications for the volume are insignificant. seen in the western Barents Sea reaching 1.3 m/km.
The present situation is not in isostatic equilibrium A theory of glacial isostatic induced hydrocarbon
and a remaining uplift of some 50 m will give 5 m migration is established and a hypothetical hydrocar-
remaining differential uplift at the spill point. bon accumulation in the western Barents Sea is being
The effects of pressure differences and fluid equi- modelled using a realistic geological history for the
librium are not taken into account, but are believed last 0.3 million years The results show that as much
to be of minor importance. as 30% of the closure volume can be lost during the
This exercise, although hypothetical, is realistic most extensive glacial events.
and shows that up to 30% of the initial bulk volume
can be affected during the glacial isostatic down- Acknowledgements
warping. In a more or less homogeneous reservoir
this would mean that nearly 30% of the hydrocarbon This study was supported by funding from the
volume would be lost. In fields where there is an oil Norwegian Council for Technology (NTNF).
leg this lost volume would mainly be oil. The lost
volume is controlled by field geometry to a large References
extent and the 30% is only related to the example Airy, B.G., 1855. On the computation of the effect of the attrac-
given. tion of mountain masses disturbing the apparent astronomical
Pleistocene glacial isostasy implications for petroleum geology 195

latitude of stations in geodetic surveys. Philos. Trans. R. Soc. Sheets. John Wiley, New York, N.Y., pp. 222-274.
London, 145: 101-104. Jamieson, T.F., 1865. On the history of the last geological
Andersen, B.G., 1979. The deglaciation of Norway 15,000-10,000 changes in Scotland. Q. Geol. Soc. London, 1, 21: 161-
B.P. Boreas, 8: 79-87. 203.
Andersen, B.G., 1981. Late Weichselian ice sheets in Eurasia Jansen, E., Bleil, U., Henrich, R., Kringstad, L. and Slettemark,
and Greenland. In: G.H. Denton, and T. Hughes (Editors), B., 1988. Paleoenvironmental changes in the Norwegian Sea
The Last Great Ice Sheets. John Wiley and Sons, New York, and the Northeast Atlantic during the last 2.8 m.y. Deep Sea
N.Y., pp. 1-65. Drilling Project/Ocean Drilling Program sites 610, 642, 643
Berggren, W.A. 1972. Late Pliocene-Pleistocene glaciation. In: and 644. Paleoceanography, 3: 563-581.
A.S. Laughton, W.A. Berggren et al. (Editors), Initial Reports Kjemperud, A., 1986. Shore line displacement in Tr0ndelag,
of the Deep Sea Drilling Project, Vol. 12. U.S. Government Central Norway. Boreas, 15: 61-82.
Printing Office, Washington, D.C., pp. 953-963. Laughton, A.S., Berggren, WA. et al. 1970. Deep Sea Drilling
Berggren, W.A. and Van Couvering, J.A., 1974. The Late Neo- Project, Leg 12. Geotimes, 15(9): 10-14.
gene: development in paleontology and stratigraphy 2. Else- Nansen, F., 1922. The strandflat and isostasy. Nor. Vidensk.
vier, Amsterdam, 216 pp. Akad. Oslo, Mat. Nat., 1927, 11: 1-350.
Boulton, G.S., 1979. Glacial history of the Spitsbergen Nansen, F. 1927. The earth's crust, its surface forms and isostatic
archipelago and the problem of a Barents shelf ice sheet. adjustments. Nor. Vidensk. Akad. Oslo Mat. Nat., 1928, 12:
Boreas, 8: 31-57. 1-121.
Cathles, L.M., 1975. The Viscosity of the Earth's Mantle. Nesje, A., Dahl, S.O., Anda, E. and Rye, N. 1988. Block fields
Princeton University Press, Princeton, N.J., 386 pp. in southern Norway: significance for the Late Weichselian ice
Clark, D.L., Whitman, R.R., Morgan, K.A. and Scudder, D.M. sheet. Nor. Geol. Tidsskr., 68: 149-169.
1980. Stratigraphy and glacial-marine sediments of the Ama- Pratt, J.H., 1855: On the attraction of the Himalayan mountains
rasian Basin, central Arctic Ocean. Geol. Soc. Am. Spec. and of elevated regions upon the plumb line in India. Philos.
Paper, 181: 1-57. Trans. R. Soc. London, 145: 53-100.
De Geer, G., 1888. Om Skandinaviens nivforandringer un- Ramsey, W, 1924. On relation between crustal movements
der Quartarperioden. Geol. Foren. Stockh. Forh., 10: 366- and variations of sea-level during the late Quaternary time,
379. especially in Fennoscandia, Fennia, 44: 1-39.
De Geer, G., 1890. Om Skandinaviens nivforandringer under Sejrup, H.P., Aarseth, I., Ellingsen, K.L., Reither, E. and Jansen,
Quartarperioden. Geol. Foren. Stockh. Forh., 12: 61-110. E., 1987. Quaternary stratigraphy of the Raden area, central
Dutton, C.E., 1889. On some of the greater problems of physical North Sea: a multidisciplinary study. J. Quaternary Sei., 2:
geology. Proc. Am. Philos. Soc. Washington, 11: 51-64; 536- 35-38.
537. Shackleton, N.J. and Opdyke, N.D., 1976. Oxygen isotope
Elverh0i, A. and Solheim, A., 1984. The physical environment, and paleomagnetic stratigraphy of Pacific core V28-239, late
Western Barents Sea 1:1 500 000, sheet B; surface sediment Pliocene to latest Pleistocene. Geol. Soc. Am. Mem., 145:
distribution. Nor. Polarinst. Skr., 179B: 1-23. 449_464.
Elverh0i, A. and Solheim, A., 1987. Shallow geology and geo- Shackleton, N.J. and Opdyke, N.D., 1977. Oxygen isotope and
physics of the Barents Sea. Nor. Polarinst. Rapportser., 37: paleomagnetic evidence for early northern hemisphere glacia-
1-71. tion. Nature, 270: 216-219.
Fairbanks, R.G. and Matthews, R.K., 1978. The marine oxygen Solheim, A. and Kristoffersen, Y., 1984. The physical environ-
isotopicrecord in Pleistocene coral, Barbados, West Indies. ment, Western Barents Sea, 1:1500000, sheet B; Sediments
Quaternary Res., 10: 181-196. above the upper regional unconformity: thickness, seismic
Fjeldskaar, W. and Cathles, L.M., 1991a. Rheology of mantle and stratigraphy and outline of the glacial history. Nor. Polarinst.
lithosphere inferred from post-glacial uplift in Fennoscandia. Skr., 179 B: 1-26.
In: R. Sabadini et al. (Editors), Glacial Isostasy, Sea-Level Sutherland, D. 1984. The Quaternary deposits and landforms of
and Mantle Rheology. Kluwer Academic Publishers, Dor- Scotland and the neighbouring shelves: a review. Quaternary
drecht, pp. 1-19. Sei. Rev., 3: 157-254.
Fjeldskaar, W and Cathles, L.M., 1991b. The present rate of Vorren, T and Kristoffersen, Y. 1986. Late Quaternary glaciation
uplift of Fennoscandian implies a low-riscosity asthenosphere. in the south-western Barents Sea. Boreas, 15: 51-59.
Terra Nova, 3: 393-400. Vorren, T Lebesbye, E., Andreassen, K. and Larsen, K.-B.,
Grantz, A., Johnson, G.L. and Sweeney, J.F 1982. The Arctic 1989. Glacigenic sediments on a passive continental margin
Region. In: A.R. Palmer (Editor), Perspectives in Regional as exemplified by the Barents Sea. Mar. Geol., 85: 2 5 1 -
Geological Synthesis, Planning for the Geology of North 272.
America. D-NAG Spec. Publ., 1: 105-115. Williams, D.F, 1988. Evidence for and against sea-level changes
Grosswald, M.G., 1980. Late Weichselian Ice Sheet of Northern from the stable isotopic record of the cenozoic. In: A.
Eurasia. Quaternary Res., 13: 1-32. Wilgus et al. (Editors), Sea-Level Changes An Integrated
Hughes, T.J., 1981. Numerical reconstruction of Paleo-Ice Sheets. Approach. Soc. Econ. Paleontol. Mineral., Spec. Publ., 42:
In: G.H. Denton and TJ. Hughes (Editors), The Great Ice 31-36.

W FJELDSKAAR Rogaland Research Institute, P.O. Box 2503, Ullandhaug, N-4004 Stavanger, Norway
A. KJEMPERUD READ Production Geology Services A/S, P.O. Box 145, N-1300 Sandvika, Norway
Present address: PETEC, P.O. Box63, N-3478Ncernes, Norway
 197

Structural style and the habitat of hydrocarbons in the

North Sea
I.J. Stewart, R.R Rattey and I.R. Vann

The structural development of the North Sea Basin is described. While economic interest has largely focussed on structuring
at Jurassic and younger levels, the structural framework was inherited from earlier times. The northern North Sea is separated
from the central North Sea by a postulated extension of the Tornquist Line, south of which Early Permian Variscan relaxation
with associated subsidence accommodated the Late Permian salt basin. The northern North Sea is typified by relaxed Caledonian
structure, extended in Devonian and Triassic times. The two structural provinces exhibit significantly different structural styles at
Jurassic levels.
In the northern North Sea the earliest significant Jurassic extension is of Bathonian to Early Callovian age and is succeeded
by a major phase in the Oxfordian. Patterns of deposition and the amount of footwall erosion can be related to the fault
pattern which in general has a broad structural wavelength. A further major faulting episode of latest Jurassic and earliest
Cretaceous age is related to Atlantic rifting and is only observed on a significant scale in the extreme north where it has played
a fundamental role in trap formation in the Magnus area. In the central North Sea, where the major period of Jurassic extension
is Oxfordian to Late Kimmeridgian, the Permian evaporites have had a considerable influence in the control of cover structure.
Extension in the basement largely utilised planar Early Permian fault surfaces, cut up-section, and was accommodated in the
cover by shallow detaching faults and short wavelength structures. The Early Cretaceous extension of the northernmost North
Sea is locally apparent in this area and has accentuated the earlier geometries.
Both areas have suffered post-rift structural modification through episodic Cretaceous and Tertiary inversion, associated salt
diapirism, and gravitational sliding. The significant effects of inversion and diapirism are restricted to the central North Sea
where they have resulted in trapping geometries in both the rift-fill and the post-rift sections.

Introduction erence to specific examples, while the final section

covers post-Jurassic structural modification.
The North Sea is a major hydrocarbon province Figure 1 is a simple regional structure map of the
that is now entering a mature phase of exploration. North Sea and adjacent areas. The area described
Continued technical, and more particularly commer- here is restricted to the central and northern North
cial, success will increasingly depend on our ability to Seas, that is the area to the north of the Mid North
correctly interpret structural geometries and also to Sea High, and to the south of the M0re Basin. The
recognise the structural controls on sediment entry present-day geometry of the basin is shown on three
and dispersal. With this in mind, we have constructed simple regional sections in Fig. 2 and the location of
this paper partly as a review and partly as a series these traverses is given in Fig. 3. The sections are
of constrained, largely regional observations, and in necessarily squashed for display purposes.
keeping with this conference volume we have given
the paper a slight Norwegian bias. Individual hydro-
Structural grain
carbon traps are not described in detail, but rather
placed in a more regional structural context. It is becoming increasingly clear that many ex-
The paper is divided into four parts, centering on tensional sedimentary basins form on relatively low-
the Jurassic extensional history. The first section out- angle inherited crustal surfaces (e.g. McGeary et
lines the pre-Jurassic structural grain of the central al, 1987), and that the structures within the basin
and northern North Sea basins. Secondly, the influ- form part of a linked system (Gibbs, 1984). How-
ence of this grain on Jurassic structuring is described ever, most basins cannot be described this simply;
by means of a few simple regional backstripped sec- various models exist for fault plane geometry and
tions and maps. Thirdly, the structural controls on the mechanism and depth of detachment (see for
sediment entry and dispersal are discussed with ref- instance Coward et al., 1987; Brun et al., 1989;

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 197-220. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
 198 LI Stewart, R.R Rattey andl.R. Vann

50 100 miles
Lambert

Fig. 1. Simple tectonic elements map of the North Sea area. This paper concentrates on the area north of the Mid North Sea High and
south of the M0re Basin.

Marsden et al., 1989; Newall and Kusznir, 1989; development (e.g., Gibbs, 1987a).
and Yielding and Roberts, this volume), and more In North Sea hydrocarbon exploration, the seismic
focussed interpretations are showing the interaction data used in structural interpretation are commonly
of both dip-slip and strike-slip processes in basin restricted to a 6 or 7 second two-way time window,
Structural style and the habitat of hydrocarbons in the North Sea 199

NORTH LINE 1
E.SHETLAND BRENT VIKING TROLL
PLATFORM PROVINCE GRABEN TERRA CES

-TSB
7
?^ : V-'^; : ,SvvJ
BT-

HORIZONTAL SCALE
I ^^YU 0 20 0 60 80 100 km
VM\ \ ' v Via
7/ ,

HALIBUT WITCH GROUND FLADEN GROUND S. VIKING UTSIRA STORD LINE 2

HORST GRABEN SPUR GRABEN HIGH BASIN

: . . * - :. : - ,.>. . :. .7-' . . ; . . . ";. 7. . ^: ? - -SB ..".' " ^ - ; ' : ';;'.': " : ' " . " ; " r . T ^ T rv ' - ^ ^ ^ ^ ^ x ~ ^ ~ 7 j
: : .:.'' :
': '!:':''.'.".';'..'V-V"'.'. : ' ' . " : ' ' J i ^ ^ j i ^ ^
x x
^^^^ '*" J - ' " ' \ / /
^/^^Sat^Si. " ^ i s s a a ^ 1 ' '^'^^^Tt^^ -J2^C*A N/
M-
>> , ^ y ^ v # A ^^^^^^^\
Y>lAp^^J
W t d l i LSayXNj/^^ti fl N V
#f f / V y \ r
YCl
r 1 1 Vxl
Q
7500
^ r

/
1JlV
/ V
K x
Nf
v\i
1
LINE 3
AUK CENTRAL JAEREN NORWEGIAN- DANISH
SHELF GRABEN HIGH BASIN

Fig. 2. Present day simple structural cross sections. Locations shown on Fig. 3. The sections are squashed for display purposes; vertical
exaggeration xlO. Dashed areas show Early Permian and older rock. BZ = Base Zechstein; TS = top salt; BTr = Base Triassic; BJr =
Base Jurassic; BK = Base Cretaceous; BT = Base Tertiary. Unlabelled lines are other sequence boundaries.

and the fundamental inherited crustal control on the Bergen Arc, lateral margins of Hornelen and related
basin location and its geometry are rarely observed. Devonian basins north of Bergen) in Norway, and
Identification of the control then has to be based the Great Glen Highland Boundary Fault in
on the few deep seismic reflection lines over the Britain (Fig. 1).
North Sea and adjacent areas, extrapolation from The eastward verging Caledonian Front of Nor-
the shallower, seismically defined, geometries and way is terminated to the south by the Fjerritslev
the regional history of the basin and its exposed Fault Zone, an interpreted extension of the Torn-
margins. quist Zone (Pegrum, 1984). Observations of cover
The basic structural grain of the North Sea was geometry clearly support the extension of the Torn-
established through the Caledonian orogen. In the quist zone as a crustal feature beyond the Oslo
Scandinavian Caledonides the thrusts verge to the Graben (see also S0rensen et al., this volume) to
south-east, whilst the Scottish Caledonides verge to the southern part of the South Viking Graben, and
the north-west (Fig. 1). The reversal is interpreted at least through to an intersection with the offshore
to occur beneath the Midland Valley (Needham and extension of the Highland Fault (Dor6 and Gage,
Knipe, 1986), although the position offshore in the 1987).
North Sea is unclear. While it is beyond the scope of In the northern North Sea, deep seismic profil-
this paper to describe in detail the Caledonian grain ing has been interpreted to indicate that the crustal
(see for instance Gee and Sturt, 1985; Johnson and faults dip eastwards and flatten towards Norway on a
Dingwall, 1981), major oblique-slip motion along a deep crustal or top Moho detachment (Beach, 1985;
north-east strike is indicated by lateral offsets of both Beach, 1986; Gibbs, 1987b) or were accommodated
the Caledonian Front and internal terranes (e.g., by pure shear (White, 1989), although the interpre-
200 LI Stewart, R.R Rattey andl.R. Vann

Fig. 3. North Sea structural elements. G - Gullfaks Field; A = Andrew Field; Sn = Snorre Field.

tation of Beach (1985) of similar easterly dipping tions elsewhere of Brewer and Smythe (1984) in
geometries beneath the Central Graben is more the MOIST data, Cheadle et al. (1987) and Snyder
speculative. The argument that the Viking Graben (1990) in the BIRPS data (see also the review by
crustal faults are relaxed Caledonian contractional McGeary et al., 1987) and the West Orkney interpre-
structures, however, is consistent with the observa- tations of Enfield and Coward (1987); the primary
Structural style and the habitat of hydrocarbons in the North Sea 201

NORTH SEA T E C T O N I C S
32 BASIN SYSTEM
LOCAL NORTH SEA REGIONAL
Q- O

N.ATLANTIC SPREADING

ALPINE INVERSION
POST RIFT
THERMAL \
SUBSIDENCE

N.ATLANTIC SPREADING
ALPINE INVERSION
UPLIFT UK LAND

ALPINE INVERSION
THERMAL N.ATLANTIC SPREADING
POST RIFT <Hh>
SUBSIDENCE

AUSTRIAN INVERSION

THERMAL N.ATLANTIC/BISCAY
POST RIFT I SUCCESSFUL RIFTING
SUBSIDENCE
/ (Includes Magnus Province)

NORTH SEA
SYN RIFT
FAILED RIFTING N.SEA FAILED RIFTING

PRE-RIFT
PRE RIFT
UPLIFT CNS NNS

ONSET OF TETHYAN
SUCCESSFUL RIFTING
THERMAL
POST RIFT I
SUBSIDENCE

EARLY NORTH SEA

SYN RIFT I /
FAILED RIFTING REGIONAL BASE BUNTER U/C

POST RIFT I THERMAL SUBSIDENCE

INFILL OF COLLAPSED CENTRAL GRABEN

SYN RIFT I INVERSION OSLO GRABEN RIFTING

CULMINATION OF
FORELAND BASIN REGIONAL
VARISCAN OROGENY

"INVERSION-
INVERSION
\
THERMAL
POST RIFT
SUBSIDENCE

DINANTIAN
SYN RIFT
RIFTING

BACK ARC RIFTING

_POS_T_RFT_ TH ERM AL _S U BS IDE NC E_
RHENO-HERCYNIAN
COLLAPSE AND

j
EXTENSION OF POST CALEDONIAN
SYN RIFT
CALEDONIAN EXTENSION
STRUCTURE

Fig. 4. North Sea. Principal post-Caledonian basin forming processes.

control of the extensional basins to the north and trending oblique-slip faults; Jurassic fault blocks seen
west of Scotland is the relaxation of Caledonian on offshore seismic data, however, imply the major
faults. The Middle Devonian age of extension identi- relaxed Caledonian structures continue to dip east
fied by these authors may then be anticipated to be and the facing reversal of Caledonian structure in
represented in the deep fill of North Viking Graben this area apparently runs close to the present-day
half-grabens (Fig. 4). Norwegian coastline. The pattern is complicated by
On the Norwegian side, Steel et al. (1985) have several changes in younger offshore half-graben po-
summarised the structural models for the Hornelen larity across the north-east striking faults.
area basins north of Bergen (Fig. 3) that accommo- Early post-Caledonian extension, however, is less
date Lower to Middle Devonian sediment. The ge- well defined in the North Sea area south of the
ometries here combine both shallow westerly dipping eastward projection of the Tornquist Zone. There
detachments inherited from a high level in the Cale- are deep fault blocks seen locally on seismic data
donian pile (e.g., Hossack, 1984) and north-easterly in the Ling Graben, Egersund sub-Basin, Norwegian
202 Ll Stewart, R.R Rattey andl.R. Vann

Danish Basin and the Mid North Sea High area, and east striking oblique-slip transfer faults, and in the
Devonian and Carboniferous extension is recognised Viking Graben to the east on the easterly dipping re-
onshore Britain (Fig. 1) from the Midland Valley laxed Caledonian trend; the East Shetland Platform
southwards on relaxed Caledonian trends (Frser et remained fixed. Accommodation between these ar-
al., 1990). It should thus be more widely expected in eas may have been achieved by transtensional motion
the central North Sea area. along the major Caledonian oblique-slip trends (such
As well as a major Caledonian lineament, the as the Ling Graben and the southern Stord Basin)
Tbrnquist Zone also played a major role in post- and the Tornquist Zone as described by Pegrum
Caledonian extension and in the subsequent Variscan (1984).
phase. Here it acted as a dextral strike-slip sys- The North Sea as a whole can thus be viewed
tem, and in the North Sea area as a limit to the as having a long extensional history, interrupted by
Variscan foreland (Pegrum, 1984). The North Sea Variscan compression, from Devonian until Early
basin lies north of the "Variscan Front" which lies Jurassic times (Fig. 4). The structural grain was es-
beneath southern England, although Late Carbonif- tablished from Caledonian relaxation in the northern
erous inversion (Fig. 4) can be traced through the North Sea, and by Caledonian and Variscan relax-
onshore U.K. basins (Bowland Basin, Northumber- ation in the central North Sea. It is clear that the
land TVough) and into the Midland Valley. overall pattern of the subsequent Jurassic extensional
Early Permian extension is well represented in the system was in place by Late Tfriassic times.
central North Sea and the Moray Firth, and may be
represented in the southern Stord Basin, although Jurassic extension
for the major part, the extension is limited to the
south of the projection of the Tbrnquist Zone and The Jurassic extension of the North Sea was the
its intersection with the Highland Boundary Fault primary structural control on the majority of known
Zone (Dore and Gage, 1987). It may be that the hydrocarbon traps. It has been studied extensively,
extension was limited to the relaxation of Variscan but in detail it is still a subject of much debate.
faults, probably on deep through-crustal structures, The first problem is the scale and extent of the
but nonetheless it has fundamental consequences for Bajocian (pre-Brent) uplift. A central North Sea do-
the subsequent structural history. The Late Permian mal uplift of this age, centering at the junction of the
(Zechstein) marine evaporites of the north Permian Central TYough and the South Viking Graben, has
sag basin (Ziegler, 1982) are effectively restricted to long been advocated by Ziegler (e.g., 1975, 1982) to
the central North Sea area south of the Highland explain the poor preservation of Liassic strata in the
Boundary Fault and the Tornquist Zone intersection central North Sea, the supposed northward progra-
(Figs. 1 and 3), and only locally extend northwards dation of the Brent system, and the progressive
through the Ling Graben into the Egersund sub- onlap of Callovian and younger Jurassic strata onto
Basin and the southern South Viking Graben. The the "mid-Cimmerian" (rift-onset) unconformity. The
Forth Approaches Basin (Fig. 1) at Permian levels is onlap would be expected under most current rift
a simple, more or less saucer-shaped depression with models and in itself is not evidence of thermal up-
few faults, although in both the Central TYough area lift. On the first point though, stratigraphic evidence
and the Norwegian-Danish Basin, extensional faults of a major Pliensbachian to Early Toarcian sedi-
struck west north-west to east south-east and were mentary break in the drilled Liassic sections of the
accommodated by north north-west to south south- central North Sea suggests a major base-level fall
east strike-slip surfaces (e.g., Vejbaek and Andersen, corresponding to the period of Cook Formation de-
1987; Brown et al., this volume). position on the northern shelf, and that much, if not
Triassic extension is recognised in both the north- all, of the missing section may have been removed by
ern and central North Sea areas. The Inner Moray sea withdrawal in Late Liassic times. By the Bajocian
Firth area extended and subsided in an arc east- (Figs. 5 and 6) coastal plain sedimentation was estab-
wards and southwards through the Witch Ground lished over much of the central North Sea and, thus,
Graben, accommodated by dextral strike-slip on the a persistent local uplift seems unlikely. Secondly, the
Great Glen Fault. This extension can also be in- Brent Group represents a 300 kmfluvial-deltaicsys-
terpreted to be linked to relaxed Variscan structure tem (Fig. 5) prograding roughly northwards over the
with increased subsidence and localised faulting in site of the Viking Graben; it is difficult to see a single
the Forth Approaches/Central TVough/Norwegian- uplift as the principal sediment source and recent
Danish Basin. This is in contrast to both easterly work (e.g., Richards et al., 1988) supports the idea
and westerly directed extension in the Stord Basin that sediment entered the rift basin transversely from
(Fig. 3) on detachments that flipped across north- regionally elevated margins, and was worked axially.
Structural style and the habitat of hydrocarbons in the North Sea 203

2E 3E 4E

r^| VOLCANICS /
U l i UPPER SHOREFACE COASTAL PLAIN L-^J VOLCANICLASTICS
LOWER SHOREFACE PELAGIC SHELF

Fig. 5. Bajocian palaeogeography.

 E.SHETLAND TROLL
LINE 1
PL A TFORM TERRACES

0 (Sea Level)
HORIZONTAL SCALE
0 20 0 60 80 100 km
2500H 2500

5000 5000

HALIBUT WITCH GROUND FLADEN GROUND S. VIKING UTSIRA STORD LINE 2

HORST GRABEN SPUR GRABEN HIGH BASIN

0 (Sea Level)

2500H 2500

5000 h5000

7500 7500

AUK CENTRAL JAEREN NORWEGIAN- DANISH

LINE 3
SHELF GRABEN HIGH BASIN

L
0 (Sea Le

2500

5000 5000

7500 7500

Fig. 6. Simple cross sections of Fig. 2 restored and decompacted to Late Oxfordian. Vertical exaggeration x 10. Labelling as in Fig. 2. Unlabelled horizons are sequence boundaries.
Structural style and the habitat of hydrocarbons in the North Sea 205

A second problem is the major change in structural interpretations made by Beach et al. (1987), and the
style from the northern to central North Sea and the interpretation of Beach (1985) that the North Viking
nature of the so-called triple junction. In both areas, Graben is a major transtensional basin. The oblique-
the Jurassic extension utilised the existing structural slip is accommodated along the South Viking Graben
grain, although clearly there are fundamental dif- boundary fault zone and the western Utsira Hight by
ferences in the structural style. The northern North a series of transfer faults or fault zones with the
Sea, particularly the north Viking Graben, is largely inherited north-east to south-west strike, and one of
characterised by long wavelength (ca. 5-20 km +) these separates the Raden Ground Spur from the
tilted fault blocks on easterly dipping, often gently East Shetland Platforms and forms the Bruce-Beryl
concave-up synthetic faults (the relaxed Caledonian) embayment.
which have a broad north-south strike (Fig. 3). Onto This structural style has to be offset against
these link an antithetic westerly dipping set over a the Central TYough-Witch Ground Graben-Outer
large part of the Norwegian sector (Beach, 1986; Moray Firth area, where extension was primarily
Gibbs, 1987b). These fault blocks typically contain dip-slip in an arc extending eastwards and south-
the pre-rift (pre-Trbert) Brent Group and older eastwards from the Great Glen fault (Fig. 1). Trans-
strata and provide a prolific and well described hy- fer faults and zones are again prominent with a
drocarbon play (e.g., Campbell and Ormaasen, 1987; north-east to south-west trend and the polarity of
Badley et al., 1984). the basins is switched locally across these. Within the
In the central North Sea, apart from the Moray Moray Firth, deep seismic data suggest that exten-
Firth and Witch Ground graben areas, the struc- sion took place within a system above a fault that dips
tural pattern is complicated by the presence of the southwards under north-east Scotland. As the basin
Late Permian salt. Jurassic extension in the Central is traced south-eastwards towards the Witch Ground
Though area utilised the Early Permian extensional Graben the polarity apparently switches (e.g., Beach,
system (the relaxed Variscan); these apparently pla- 1984). The Central Ttough can be similarly inter-
nar faults seen on conventional seismic data, cut-up preted as part of a linked extensional system (e.g.,
section and were locally controlled by the salt wall Gibbs, 1984), above a detachment that continues be-
and Ttiassic sediment pod geometries. The result neath the Forth Approaches and the Central (Mid
is a series of short wavelength commonly complex North Sea) High (Fig. 3).
structures governed by irregular fault plane and salt The critical separation between the Viking Graben
wall geometry, superimposed on a longer wavelength and the Central TYough is marked by a zone of
structural pattern formed by the underlying pre- complex structure that extends along the southern
Zechstein fault blocks. However, whilst the presence margin of the Utsira High (Pegrum, 1984) and be-
of salt and the associated complexity accounts for neath the Andrew Field in the southern part of
the generally smaller Jurassic structural hydrocar- UKCS Quadrant 16 (Fig. 3). This area represents
bon trap size compared to the Viking Graben, it the extension of the Tornquist Zone (Pegrum, 1984),
does not on its own explain the more fundamental which together with the Ling Graben faults and
observations. the oblique-slip trends of the southern Stord Basin,
The structural elements map (Fig. 3) shows two apparently accommodated the necessary rotation of
clear trends for the north Viking Graben area (see Norway relative to Scotland.
also Halstead, 1975; and Johnson and Dingwall, The simple cross-sections are shown restored to
1981). The sense of dip-slip extension is more or Late Oxfordian times in Fig. 7, and a palaeogeo-
less east-west, although the extensional system is graphic map of this timeslice is given in Fig. 8. One
repeatedly offset by inherited north-east to south- observation is paramount; the Late Oxfordian of the
west transfer or strike-slip faults. The interaction of northern North Sea was deposited in a complex,
these two trends has created a mosaic of rhomb- and in fact starved rift basin with great submarine
shaped elements (a classic example is the Gullfaks relief, whereas in the central North Sea sedimenta-
Field, Erichsen et al., 1987) which together are tion kept up more or less with sea level. The ages
bound to the north-west by the Great Glen and of extension in the Northern North Sea were ini-
related fault systems, and to the south-east by a tially Bathonian to Early Callovian, which, coupled
series of transfer faults related and parallel to the with a rise in sea level, saw the drowning of the
Hitra and Verran fault zone (M0re margin faults) of Brent delta system ("T^rbert transgression"), and
Mid-Norway, and the major Caledonian oblique-slip Middle Oxfordian. The latter was a period of great
trends to the north of the TVoll Field area (Badley et change with extremely rapid, probably near instan-
al., 1988) and onshore Hornelen. We believe that the taneous, hanging-wall subsidence and footwall uplift;
map observations strongly support the deep seismic the mid-Oxfordian to Kimmeridgian strata show little
LINE 1 NORTH
E.SHETLAND BRENT TROLL
VIKING
PL A TFORM PRO VINCE TERRACES
GRABEN

HORIZONTAL SCALE
0 20 0 60 80 100 km

5000

7500

LINE 2 HALIBUT WITCH GROUND FLADEN GROUND S. VIKING UTSIRA STORD

HORST GRABEN SPUR GRABEN HIGH BASIN

5000

7500

AUK CENTRAL JAEREN NORWEGIAN- DANISH

SHELF GRABEN HIGH BASIN
LINE 3

2500 H2500

h5000
iT 5000

7500

Fig. 7. Simple cross section of Fig. 2 restored and decompacted to Late Oxfordian. Vertical exaggeration xlO. Labelling as in Fig. 2.
Structural style and the habitat of hydrocarbons in the North Sea 207

| I NEAR SHORE SHELF (PROGRADATIONAL/DROWNED)

BASIN H SLOPE (PROGRADATIONAL/DROWNED)

Fig. 8. Late Oxfordian Palaeogeography.

208 /./ Stewart, R.E Rattey and LR. Vann

rotational growth into the faults and much of the ob- Footwall uplift characterises the basin margin and
served hanging-wall geometry seen on seismic data is the majority of the fault blocks of the Viking Graben
a consequence of subsequent differential compaction (see for instance Marsden et al., 1989). The crests of
which is removed when sections are restored and the blocks are typically eroded with varying amounts
decompacted. of pre-rift Jurassic and THassic section removed, and
A Middle Oxfordian age of extension is similarly there appears to be a simple relationship between
recognised in the Inner Moray Firth (Andrews and the width of the individual fault blocks, the throw
Brown, 1987), where a complex, relatively deep- on the bounding faults and the amount of erosion.
water basin was established against the Great Glen Where extension is taken up on a few faults, the
Fault system, although eastwards into the Witch fault throw and structural wavelength is increased,
Ground Graben and Central TYough areas the basin as is the footwall uplift. This is particularly apparent
shallowed and was dominated by shallow-marine in the Snorre area (Fig. 3) of the Norwegian sector
elastics (commonly fault accommodated, e.g., John- where THassic rock in the footwall subcrops rocks of
son et al., 1986). Minor deeper marine troughs devel- Early Cretaceous age.
oped over the eastern and western areas of the Cen- The structural dip and the consequent drainage
tral TYough (Fig. 8), although more regional deeper ensured that the majority of the erosion products
marine conditions as a consequence of extension could not enter the hanging-wall in front of the indi-
were not established in the central North Sea area vidual blocks, and these were transported down the
until Late Kimmeridgian to Early Portlandian times. dip slope to accumulate as narrow shelf systems nor-
This observation and the map highlight the major mal to the dip. A cartoon through the Brent Province
stratigraphic difference in the Jurassic hydrocarbon showing the geometry and the chronostratigraphy
play types between the northern and central sectors; (Fig. 10) indicates that this process occurred sub-
the pre-rift marginal to shallow-marine Jurassic se- sequent to both the Bathonian to Early Callovian
quences are the main target in the north, and the extension and the younger Middle Oxfordian exten-
rift-fill shallow-marine Jurassic sequences provide sion, and a series of stacked retrogradational shelf
the major play in central areas. systems can be anticipated, becoming narrower as the
A further major fault episode of the North Sea is dip slope was rotated in the extension phases. The
related to the attempted opening of the Atlantic in pattern would depend on the relative dip and the
earliest Cretaceous times. The major effects of this elevation/erosion level and the rate of subsidence/
are noted in the extreme north of the northern North sediment supply in each of the blocks. The Statfjord
Sea where east north-east striking faults displace the Nord accumulation (Fig. 10) in the Munin sand is
Oxfordian faults and their depositional systems (Fig. of this type (Gradijan and Wiik, 1987), where the
9). The scale of faulting is much greater than in the lithofacies are interpreted as shelfal or storm-beach
North Sea failed rift and resulted in the develop- sands.
ment of the M0re Basin and the Shetland TYough, Deeper marine sandstones may also occupy a com-
but again is believed to have developed almost in- parable structural position and are related to local
stantaneously. The episode was accommodated to (individual fault block) or more regional lowstand
variable extents throughout the North Sea basin by events, where the shelf system was eroded or by-
dip-slip and oblique-slip motion on the existing fault passed and the erosional products were deposited on
surfaces. the dip slope or in the immediate hanging-wall area
of the controlling fault. The Magnus Field (Fig. 9) is
contained within submarine-fan sandstone (De'Ath
Structural controls on reservoir distribution and Schuyleman, 1981) and lies in the footwall to
a cross-cutting Atlantic margin fault, yet on the dip
The Jurassic extension of the North Sea pro- slope of an Oxfordian North Sea fault (Fig. 11;
vided a complex framework in which both deep- Young, this volume). The field area itself has been
water and shallow-water rift-fill Jurassic sequences elevated above sea level and partly eroded through
accumulated, and a wide array of play types in both footwall uplift related to the younger extensional
the northern and central sectors is recognised. For event.
the purposes of this paper only a few representa- In the hanging-wall position, the pattern of foot-
tive examples are described. In simple terms the wall uplift and fault geometry governs the entry of
plays can be separated by their structural position in coarse elastics. Only a small amount of footwall ero-
the footwalls or hanging walls of extensional faults, sional products will enter the hanging-wall across the
or as in the Magnus example (Fig. 9), in areas of faults scarp, and these are typically poorly ordered
cross-cutting faults. apron-fans derived from the mass wastage of the
 S
024' 0 36' 048' 1 1 12' 120'

ft-

DROWNED SLOPE

Fig. 9. Northern North Sea: Magnus area, the timeslice shown is Middle Kimmeridgian. The earliest Cretaceous Atlantic Margin faults displace the earlier extensional structure and the associated
depositional systems.
o
CD
210 I.J. Stewart, R.R Rattey andl.R. Vann

EAST

CONTINENTAL /
SLOPE BASIN
COASTAL PLAIN CONDENSED
CONTINENTAL / SECTION
SHELF
SHOREFACE

Fig. 10. Summary chronostratigraphy Brent ProvinceATampen Spur. The section is a cartoon summarising the Jurassic and Early
Cretaceous depositional sequences and their geometry. The Hthofacies are shown on the chronostratigraphic diagram.
Structural style and the habitat of hydrocarbons in the North Sea 211

MAGNUS
PRESENT DAY NORTH SHETLAND TROUGH PENGUIN

END NEOCOMIAN

? PORTLANDIAN TO ? HAUTERIVIAN (ACCOMODATION SPACE)

MAGNUS
LATE KIMMERIDGE (MAGNUS DEPOSITION)

=q-Om

v\
r\ / \ /\,
\/\
y
\ ' ' N
^ A v \V
\// \ V \ ^ \ \ / \

LINE LENGTH = 100Kms

Fig. 11. Magnus Province decompacted and restored section based on Young (this volume). The location of the section is shown as
Line 4 on Fig. 3 and on Fig. 9. Labelling of horizons as in Fig. 2. Unlabelled horizons are sequence boundaries.
 212 LJ. Stewart, R.R Rattey and I.R. Vann

5850

5830

5820

5810

| DROWNED SHELF [ j PELAGIC BASIN | g | SUBMARINE

[ I DROWNED SLOPE | | NEAR SHORE

Fig. 12. The Brae Trend of the South Viking Graben. The timeslice shown is Middle Ryazanian. The organised fans enter the grab(
from the Fladen Ground Spur through transfer systems or across tip points of the extensional faults.
Structural style and the habitat of hydrocarbons in the North Sea 213

fault scarp itself during uplift, or slumped lithifled the U.K. sector, although here more extensive flat
blocks. The more ordered systems, both in shallow detachments at the base of the salt link into the
and deeper marine settings, require a specific entry originally Permian faults.
point and invariably this is provided by cross-cutting Throughout the central and northern North Sea,
or transfer faults, flexures (transfer-zones), or at the the post-rift thermal subsidence led to drowning of
tip points of extensional faults. A large-scale exam- the majority of the footwall highs and the basin
ple of this is the progradation of the Sognefjord and flanks, and the deposition of the Ryazanian hot
Fensfjord sequences that provide the reservoir at the shale as a pelagic drape (Bertram and Milton, 1989).
Ttoll Field, from the major oblique-slip system that The infill of the rift topography ceased, and a sub-
defines the northern edge of the Horda Platform stantial marine topography was established that was
(Fig. 3). progressively onlapped by deep marine Cretaceous
The South Viking Graben, with dip-slip and trans- sediment.
fer segments along its bounding fault related to
Jurassic and earlier extension, received large vol-
umes of submarine elastics in the Late Jurassic. Post-rift structural modification
Initially these were apron-fan deposits (e.g., Central
Brae, South Brae), although by Ryazanian times, the The structural pattern of the North Sea, the cen-
depositional system had evolved with the erosional tral area in particular, has been modified through
products of the footwall (Fladen Ground Spur) being post-rift tectonic processes. The most striking is
worked down the dip slope and following structural structural inversion which was accompanied in the
topography to enter the graben through the trans- area of the salt basin by diapirism. The inversion is
fer system (Fig. 12). These provide the reservoir widespread (Fig. 14) and can be traced from the Cen-
for the Miller Field which is closed by subsequent tral TYough in Denmark (Gowers and Saeb0e, 1985)
compaction-drape over the fan body. and through into the U.K. Sector, the Norwegian-
The structural style of the Central North Sea, par- Danish Basin and into the Egersund sub-basin,
ticularly the Central Ttough area, has been argued the southern part of the Stord Basin (Biddle and
to be significantly different to that of the Viking Rudolph, 1988) and the south-eastern Witch Ground
Graben. The obvious difference is the presence of Graben. The northern limit of marked inversion cor-
salt, although the major structural difference is the responds in the south-east to the Tornquist Zone
re-utilisation of Permian dip-slip and strike-slip faults (Pegrum, 1984), while the structuring in the southern
during Jurassic extension. Fig. 13 includes a simple Stord Basin appears related to the intersection of
geoseismic section across the eastern margin of the the north-east striking transfer zones and the basin
Central Ttough, through the Gyda Field, and shows margin faults. There is thus a remarkable consis-
the Late Permian salt wall and TYiassic pod geometry tency between the surfaces that accommodated the
to have stabilised by at least Late TYiassic times. The different extensional patterns of the Viking Graben
Jurassic depositional basin margin is controlled by and the central North Sea during Jurassic and earlier
a dip-slip westerly dipping fault, which in Jurassic extension, and those utilised in younger inversion.
extension cut-up section through the Triassic and salt The inversion was thus governed by the existing
wall stratigraphy to involve the cover in a complex structural grain and new faults were only formed at
hanging-wall structure. The steeper Permian faults shallow levels where the existing extensional surfaces
to the east of the dip-slip fault on the section map were unable to accommodate the contraction. Fig-
out as near north-south trending strike-slip surfaces, ure 16 shows a section over the Lindesnes Ridge
which on a regional scale accommodated the dip-slip where the inversion enforced the development of
extension as well as creating cover pull-aparts (e.g., short-cut faults through the footwall, and backthrusts
the Ula Field; Figs. 13 and 15; Brown et al., this vol- in the original hanging wall, the latter utilising orig-
ume) and enforcing the inversion of Triassic pods up inally antithetic faults. The critical aspect of the
salt wall flank faults. Thick Jurassic shallow-marine inversion to hydrocarbon exploration was the devel-
sequences occupy the hanging-wall to the marginal opment of structural traps at both Jurassic levels and
fault system, the local geometry of which governs the in the post-rift section. In the Ula Trend, the sense
position of the shelf and the shelf-break (e.g.,fig.7 of of motion of the originally extensional and sinistral
Bergan et al., 1989). Detailed unpublished mapping strike-slip Early Permian fault surfaces was reversed,
has shown that sediment entry into the hanging-wall such that a series of small wavelength dip-slip related
occurred through north-east striking cover transfer closures was formed over the extensional faults, while
faults. The pattern of the Norwegian sector is repli- the Ula "pull-apart" (Fig. 15) popped-up (Brown et
cated in, for example, the Fulmar-Clyde area of al., this volume).
 SHORELINE OUTER SHELF / SLOPE
[ffi:[ SHELF HIGH WITH CONDENSED SEQUENCE
NON-DEPOSITIONAL SHELF BASINAL

\ :
, V.- / / ;>
. . " ' '*

'

; \c'>''x'' \ V - \ . \ , - * .- r^5Y.t)A \.'."-"';

^ ; ,
- M O : ^ V^lJ^i- BT-^-T
v
.-"' ' - V'V , y - j -' * - 7 ^^7^" . _ J I I J - 1 - 1 _ ^ 1 J - ^ S ^ T " " ' "'

BjC-^

^ * r
*Vk+^|F - v ^

WESTERN
/

FLANK OF NORWEC
/ Y
ffiBk J 0 5km
Structural style and the habitat of hydrocarbons in the North Sea 215

Fig. 14. Inversion structuring in the North Sea. Structural elements map showing areas of Cretaceous and Tertiary structural inversion.
Cross sections are shown in Figs. 15 and 16.

Fig. 13. The Ula Trend of the Norwegian central North Sea. The timeslice shown is Middle Kimmeridgian. The underlying section is
not depth converted. MO = Middle Oligocene; BT = Base Tertiary; K = Middle Albian; BK = Base Cretaceous; BJ = Base Jurassic;
BZ = Base Zechstein.
216 I.J. Stewart, R.R Rattey andl.R. Vann

Ula : Present Day Depth

WEST EAST

M+n o u G p C E w e _ n

0 1 2 3 4 5km
I I I I I I

Ula : D e c o m p a c t e d and R e s t o r e d to Mid. Albian

WEST EAST

o< eA
ot
BASE CRETACEOUS

5km
_l

Fig. 15. Ula Field, central North Sea. Present day depth section, and restored and decompacted to Middle Albian. The location of the
section is shown in Fig. 14.

Similar dextral strike-slip motion along north- are all related to the inversion geometries.
south faults is argued to have governed inversion The principal ages of inversion are Early Creta-
geometries in the Danish Central TYough (Vejbaek ceous (Middle Albian), Late Cretaceous (Middle to
and Andersen, 1987). In the post-rift section, the Late Campanian) and Tertiary (Ypresian and Middle
closures that were formed have a much longer struc- Oligocene; Fig. 4). Onlap unconformities of these
tural wavelength, and can be demonstrated to be ages have been dated over inversion structures in
related to the fault geometry at Jurassic levels or a large number of exploration wells. The Late Cre-
deeper ramp geometries of the fault surfaces. The taceous and younger inversion episodes have been
chalk fields of the Norwegian sector (e.g., Fig. 16) related to Alpine compression and associated strike-
Structural style and the habitat of hydrocarbons in the North Sea 217

WEST EAST

HOD VLHACL-

MID NORTH SEA

HIGH

LINDESNES RIDGE

0-(-1 "EDO-A EKOffSk

Fig. 16. Inversion geometries in the central North Sea. The upper section covers the Lindesnes Ridge, over which sit the Valhall and
Hod fields. The lower section runs through the Ekofisk area. The locations are shown in Fig. 14, upper section is the southern line.
MJ = top Middle Jurassic; other horizons labelled as Fig. 13.

slip motion along major crustal fault zones (e.g., In addition to the structural inversion, the central
Pegrum, 1984; Beach, 1987) of which the Tornquist North Sea contains the Zechstein salt which was
Zone, the offshore extension of the Highland Bound- mobilised in the Cretaceous and Tertiary to form di-
ary Fault and the oblique-slip faults of the southern apiric structures that locally extend up to the sea bed.
Stord Basin acted in the North Sea area as the north- Over the shallow, relatively unfaulted, basin margins
ern buttress. There are only minor Cretaceous and (e.g., Forth Approaches) local pillowing and gravity
Tertiary inversion structures in the northern North sliding can be shown to be related to basin tilting
Sea, although major Oligocene inversion geometries and Middle Tertiary and older sediment loading. In
occur throughout the Cretaceous Atlantic Rift basins the Central Ttough area though, the broad salt wall
of Norway from the More Basin to as far north and THassic sediment pod geometry had stabilised by
as the western Barents Shelf and may in part be at least Late TYiassic times with the major salt walls
related to ridge-push compression. The Early Creta- developing preferentially from the sites of Early Per-
ceous (approximately Middle Albian) contraction is mian hanging-walls. With the later Jurassic extension
more difficult to account for. It is recognised over re-utilising the Permian system, the salt walls also oc-
much of the Norwegian sector of the Central Ttough cupied the hanging-wall areas of the younger faults,
(e.g., Ula; Fig. 15), in the Sleipner area (Pegrum although in the Norwegian sector at least, the faults
and Ljones, 1984) and in the Inner Moray Firth, commonly cut-up section through the salt (Figs. 13,
and in each area appears to be associated with Early 16) leaving the Tiiassic sediment pods unfaulted. The
Permian and older strike-slip trends. The inversion dating of the main ages of diapiric growth, from the
is apparently slightly younger than the Austrian Tec- onlap relationships on the diapir flanks, shows that in
tonic phase (Ziegler, 1982), although it agrees with many examples this was episodic, but rapid, and the
the initial age of inversion of the Sole Pit TVough ages concur with the inversion unconformities seen
(Walker and Cooper, 1987). in structures which lack salt. It would appear that
218 LI Stewart, R.R Rattey andl.R. Vann

Fig. 17. Section across margin of western Central Trough in the UKCS restored and decompacted to Top Early Cretaceous to reveal
palaeo-structural dip. The structure is a gravity slide detached at Base Zechstein and emplaced over the trough margin. The location
of the figure is shown in Fig. 14.

the most significant diapiric growth allowing the salt is the compaction effect as a result of sediment load.
to reach shallow stratigraphic levels occurred when As well as creating simple drape closures (e.g., Mon-
the basin was in contraction. The diapir flank play trose Field) over deeper structures or rock, which
contains few major discoveries in the central North are less amenable to compaction, the recognition
Sea and success to date has relied on the presence of of compaction structure in the post-rift section is
Early Palaeogene sands and locally fractured chalk, commonly critical to structural interpretation. Inver-
although the former are absent throughout the major sion structures can be distinguished from compaction
part of the Norwegian Central TYough. structures by the vergence of the tip-line fold in the
In addition to younger growth, the salt in the post-rift strata, although deeper compaction struc-
central North Sea played an important role in the tures, particularly on the flanks of incompactible
Jurassic extension, providing a major decollement at salt walls, or against incompactible basement in the
the base of the salt section. There was also extension hanging-walls of extensional faults, are frequently
utilising the base salt detachment in the post-rift misinterpreted. This has to be incorporated into
subsidence phase, but not balanced by deeper faults. backstripping routines, which are becoming increas-
The most striking example of this is over the Auk ingly important in the interpretation of structural
Shelf area (e.g., Section 3 on Fig. 2), where the en- geometries (e.g., Young, this volume) and are pro-
tire section above the base of the salt has detached. viding a means to a better understand not only of the
The extension was initiated by flexural tilting associ- structural, but also the stratigraphic habitat of North
ated with the earliest Cretaceous thermal subsidence Sea hydrocarbons.
and is balanced regionally by salt-cored contractional
folds, or by the emplacement of a toe structure over Conclusions
the edge of the western Central Trough (Fig. 17)
which has been proven to contain a few small hydro- The central and northern North Sea cover a se-
carbon pools. ries of structurally complex sedimentary basins that
Probably the final, but extremely important, aspect together form a prolific hydrocarbon province. The
of the post-rift modification of the North Sea Basin observations presented in this paper support the
Structural style and the habitat of hydrocarbons in the North Sea 219

long-held view (e.g., Johnson and Dingwall, 1981) Oseberg feature, Block 30/6, offshore Norway. J. Geol. Soc.
that the basement grain, which was inherited from London, 141: 639-649.
Badley, M.E., Price, J.D., Rambech Dahl, C. and Agdestein, T,
at least Caledonian times, has played a fundamental
1988. The structural evolution of the North Viking Graben
role in the younger structuring. The Viking Graben and its bearing upon extensional modes of basin formation. J.
area can be separated from the central North Sea by Geol. Soc. London, 145: 455-478.
a projection of the Tornquist Zone as advocated by Beach, A., 1984. Structural evolution of the Witch Ground
Pegrum (1984), which together with a series of re- Graben. J. Geol. Soc. London, 141: 621-628.
Beach, A., 1985. Some comments on sedimentary basin devel-
activated north-east striking Caledonian fault zones
opment in the northern North Sea. Scott. J. Geol., 21(4):
accommodated different extensional and transten- 493-512.
sional patterns between the central and northern Beach, A., 1986. A deep seismic reflection profile across the
North Sea areas. Variscan relaxation in the central northern North Sea. Nature, 323: 53-55.
North Sea accommodated the Late Permian age salt Beach, A., 1987. A regional model for linked tectonics in
north-west Europe. In: J. Brooks and K. Glennie (Editors),
basin, which has had a major effect on the sub-
Petroleum Geology of North West Europe. Graham and
sequent structural style and wavelength at Jurassic Trotman, London, pp. 43-48.
levels, while the northern North Sea is dominated Beach, A., Bird, T. and Gibbs, A.D., 1987. Extensional tectonics
by longer wavelength relaxed Caledonian trends that and crustal structure: deep seismic reflection data from the
provide the classic tilted fault block play. The Mag- northern North Sea Viking Graben. In: M.P. Coward, J.F.
nus area of the northern North Sea provides evidence Dewey and P.L. Hancock (Editors), Continental Extensional
Tectonics. Geol. Soc. London, Spec. Publ., 28: pp. 467-476.
of younger extension associated with the Cretaceous Bergan, M., Torudbakken, J. and Wandaas, B., 1989. Lithos-
Atlantic basins. The observations question the ap- tratigraphic correlation of Upper Jurassic sandstones within
plication of simple kinematic models such as that the Norwegian Central Graben: sedimentological and tec-
proposed by Roberts et al. (1989). tonic implications. In: J.D. Collinson (Editor), Correlation
in Hydrocarbon Exploration. Norwegian Petroleum Society,
Despite the different geometries in the two areas, Graham and Trotman, London, pp. 243-251.
the processes of Jurassic rift-fill sedimentation are Bertram, G.T and Milton, N.J., 1989. Reconstructing basin
very similar. The rift-fill sequences provide major hy- evolution from sedimentary thickness; the importance of
drocarbon plays throughout the North Sea, although palaeobathymetric control, with reference to the North Sea.
they are particularly productive in the central North Basin Res., 1: 247-257.
Biddle, K.T. and Rudolph, K.W, 1988. Early Tertiary structural
Sea (e.g., the Ula-Gyda and Fulmar-Clyde trends)
inversion in the Stord Basin, Norwegian North Sea. J. Geol.
where fault controlled subsidence occurred later on Soc. London, 145: 603-612.
a regional scale. The structures of the central North Brewer, J.A. and Smythe, D.K., 1984. MOIST and continuity of
Sea area in particular have also suffered several crustal reflector geometry along the Caledonian-Appalachian
phases of structural inversion; this, combined with orogen. J. Geol. Soc. London, 141: 105-120.
Brown, A., Mitchell, A.W, Nssen, I.R., Stewart, I.J. and Svela,
associated salt diapirism, has complicated the struc-
P.T., 1992. Ula Field: relationship between structure and hy-
tural pattern at Jurassic levels, but has created longer drocarbon distribution? In: R.M. Larsen, H. Brekke, B.T.
wavelength structural closures in the post-rift section, Larsen and E. Talleraas (Editors), Structural and Tectonic
notably containing the chalk fields of the Norwegian Modelling and its Application to Petroleum Geology. Norwe-
and Danish sectors. gian Petroleum Society (NPF), Special Publication 1. Elsevier,
Amsterdam, pp. 409-420 (this volume).
Brun, J.P., Allemand, P. and Ballard, J.F., 1989. Some mechanical
aspects of extensional basin development in nature and
Acknowledgements
experiments. Paper presented at Norwegian Petroleum Society
Workshop, 18-20 October 1989, Stavanger.
The authors are indebted to many colleagues in Campbell, C.J. and Ormaasen, E., 1987. The discovery of oil
Norway and in Britain for useful discussion and and gas in Norway: a historical synopsis. In: A.M. Spencer et
their contributions to this paper, in particular Neil al. (Editors), Geology of the Norwegian Oil and Gas Fields.
Graham and Trotman, London, pp. 1-37.
Parkinson, Nick Johnson and Tony Hayward. This
Cheadle, M.J., McGeary, S. Warner, M.R. and Matthews, D.H.,
paper is published with the permission of the Board 1987. Extensional structures on the western UK continental
of BP Exploration. shelf: a review of evidence from deep seismic profiling.
In: M.P. Coward, J.F. Dewey and P.L. Hancock (Editors),
Continental Extensional Tectonics. Geol. Soc. London, Spec.
References Publ., 28: pp. 445-465.
Coward, M.P, Dewey, J.F. and Hancock, P.L., 1987. Continental
Andrews, J.R. and Brown, S., 1987. Stratigraphic evolution Extensional Tectonics. Geol. Soc. London, Spec. Publ., 28:
of the Jurassic, Moray Firth. In: J. Brooks and K. Glennie 637 pp.
(Editors), Petroleum Geology of North West Europe. Graham De'Ath, N.G. and Schuyleman, S.F., 1981. The geology of
and Trotman, London, pp. 785-795. the Magnus oilfield. In: L.V. Illing and G.D. Hobson (Edi-
Badley, M.E., Egeberg, T. and Nipen, O., 1984. Development tors), Petroleum Geology of North-West Europe. Institute of
of rift basins illustrated by the structural evolution of the Petroleum, London, pp. 342-351.
220 IJ. Stewart, R.P. Rattey and I.R. Vann

Dor6, A.G. and Gage, M.S., 1987. Crustal alignments and wedge, southwestern Scotland. Geology, 14: 303-306.
sedimentary domains in the evolution of the North Sea, Newall, M.J. and Kusznir, N.J., 1989. A coupled quantitative
North-East Atlantic Margin and Barents Shelf. In: J. Brooks, tectonic-sedimentary model of extensional sedimentary basin
and K. Glennie (Editors), Petroleum Geology of North West formation. Paper presented at the Norwegian Petroleum
Europe. Graham and Trotman, London, pp. 1131-1148. Society Workshop, 18-20 October, 1989, Stavanger.
Enfield, M.A. and Coward, M.P., 1987. The structure of the West Pegrum, R.M. 1984 The extension of the Tornquist Zone in the
Orkney Basin, northern Scotland. J. Geol. Soc. London, 144: Norwegian North Sea. Nor. Geol. Tidsskr., 64: 39-68.
871-884. Pegrum, R.M. and Ljones, T.E., 1984. 15/9 Gamma Gas Field
Erichsen, T, Helle, M., Henden, J. and Rognebakke, A., 1987. offshore Norway, new trap type for North Sea Basin with
Gullfaks. In: A.M. Spencer et al. (Editors), Geology of regional structural implications. Bull. Am. Assoc. Pet. Geol.,
the Norwegian Oil and Gas Fields. Graham and Trotman, 68: 874-902.
London, pp. 273-286. Richards, P.C., Brown, S., Dean, J.M. and Anderton, R., 1988. A
Frser, A.J., Nash, D.F., Steele, R.P. and Ebdon, C.C., 1990. new palaeogeographic reconstruction for the Middle Jurassic
A regional assessment of the intra-Carboniferous play of of the northern North Sea. J. Geol. Soc. London, 45: 883-886.
Northern England. In: J. Brooks (Editor), Classic Petroleum Roberts, A., Yielding, G. and Badley, M., 1989. A kinematic
Provinces. J. Geol. Soc. London, Spec. Publ. model for the opening of the North Sea/Norwegian Rift,
Gee, D.G. and Sturt, B.A. (Editors), 1985. The Caledonide with emphasis on the dip-slip nature of the main graben
Orogen Scandinavia and Related Areas. John Wiley and systems. Paper presented at the Norwegian Petroleum Society
Sons, New York, N.Y., 1266 pp. Workshop, 18-20 October, 1989, Stavanger.
Gibbs, A.D., 1984. Structural evolution of extensional basin Snyder, D.B., 1990. The Moine Thrust in the BIRPS data set. J.
margins. J. Geol. Soc. London, 141: 609-620. Geol. Soc. London, 147: 81-86.
Gibbs, A.D., 1987a. Development of extension and mixed-mode Steel, R., Siedlecka, A. and Roberts, D., 1985. The Old Red
sedimentary basins. In: M.P. Coward, J.F Dewey and RL. Sandstone basins of Norway and their deformation: a review.
Hancock (Editors), Continental Extensional Tectonics. Geol. In: D.G. Gee and B.A. Sturt (Editors), The Caledonide
Soc. London, Spec. Publ., 28: pp. 445-465. Orogen Scandinavia and related areas. John Wiley and
Gibbs, A.D., 1987b. Deep seismic profiles in the northern North Sons, New York, N.Y., pp. 293-316.
Sea. In: J. Brooks and K. Glennie (Editors), Petroleum Ge- S0rensen, S., Morizot, H. and Skottheim, S., 1992. A tectonos-
ology of North West Europe. Graham and Trotman, London, tratigraphic analysis of the southeast Norwegian North Sea
pp. 1025-1028. Basin. In: R.M. Larsen, H. Brekke, B.T Larsen and E.
Gowers, M.B. and Saeb0e, A., 1985. On the structural evolution Talleraas (Editors), Structural and Tectonic Modelling and
of the Central Trough in the Norwegian and Danish sectors its Application to Petroleum Geology. Norwegian Petroleum
of the North Sea. Mar. Pet. Geol., 2(4): 298-318. Society (NPF), Special Publication 1. Elsevier, Amsterdam,
Gradijan, S.J. and Wiik, M., 1987. Statfjord Nord. In: A.M. pp. 19-42 (this volume).
Spencer et al. (Editors), Geology of the Norwegian Oil and Vejbask, O.V. and Andersen, C , 1987. Cretaceous-Early Tertiary
Gas Fields. Graham and Trotman, London, pp. 341-350. inversion tectonism in the Danish Central Trough. Tectono-
Halstead, PH., 1975. Northern North Sea faulting. In: K.G. Fin- physics, 137: 221-238.
stad and R.C. Selley (Coordinators), NPF Jurassic Northern Walker, J.M. and Cooper, W.G., 1987. The structural and strati-
North Sea Symposium Volume, Stavanger, JNNSS/10. graphic evolution of the northeast margin of the Sole Pit
Hossack, J.R., 1984. The geometry of listric growth faults in Basin. In: J. Brooks and K. Glennie (Editors), Petroleum Ge-
the Devonian basins of Sunnfjord, W Norway. J. Geol. Soc. ology of North West Europe. Graham and Trotman, London,
London, 141: 629-638. pp. 263-275.
Johnson, H.D., MacKay, T.A. and Stewart, D.J., 1986. The White, N., 1989. Nature of lithospheric extension in the North
Fulmar Oilfield (Central North Sea): geological aspects of Sea. Geology, 17(2): 111-114.
its discovery, appraisal and development. Mar. Pet. Geol., 3: Yielding, G. and Roberts, A., 1992. Footwall uplift during
100-125. normal faulting implications for structural geometries in
Johnson, R.J. and Dingwall, R.G., 1981. The Caledonides: their the North Sea. In: R.M. Larsen, H. Brekke, B.T. Larsen and
influence on the stratigraphy of the Northwest European E. Talleraas (Editors), Structural and Tectonic Modelling and
Continental Shelf. In: L.V. Illing and G.D. Hobson (Editors), its Application to Petroleum Geology. Norwegian Petroleum
Petroleum Geology of the Continental Shelf of North-West Society (NPF), Special Publication 1. Elsevier, Amsterdam,
Europe. Institute of Petroleum, London, pp. 85-97. pp. 289-304 (this volume).
Marsden, G., Yielding, G., Roberts, A. and Kusznir, N., 1989. Young, R., 1992. Restoration of a regional profile across the
Modelling of extensional basin formation: the northern Viking Magnus Field in the northern North Sea. In: R.M. Larsen, H.
Graben. Paper presented at the Norwegian Petroleum Society Brekke, B.T. Larsen and E. Talleraas (Editors), Structural and
Workshop, 18-20 October 1989, Stavanger. Tectonic Modelling and its Application to Petroleum Geology.
McGeary, S., Cheadle, M.J., Warner, M.R. and Blundell, DJ., Norwegian Petroleum Society (NPF), Special Publication 1.
1987. Crustal structure of the continental shelf around Britain Elsevier, Amsterdam, pp. 221-229 (this volume).
derived from BIRPS deep seismic profiling. In: J. Brooks Ziegler, P.A., 1975. Geologic evolution of the North Sea and its
and K. Glennie (Editors), Petroleum Geology of North West tectonic framework. Bull. Am. Assoc. Pet. Geol., 59: 1073-
Europe. Graham and Trotman, London, pp. 33-41. 1097.
Needham, D.T. and Knipe, R.J. 1986. Accretion- and collision- Ziegler, P.A., 1982. Geologic Atlas of Western and Central
related deformation in the Southern Uplands accretionary Europe. Elsevier, Amsterdam, 130 pp.

I J . STEWART BP Norway Limited U.A., P.O. Box 197, N-4033 Forus, Norway
Present address: BP Exploration Inc., P.O. Box 4587, Houston, TX 77210, U.S.A.
R.P. RATTEY BP Exploration, Glasgow, U.K.
I.R. VANN BP Exploration, London, U.K.
 221

Restoration of a regional profile across the Magnus Field in

the northern North Sea
R. Young

Confidence in a geological model can be increased by checking it against quantitative reconstructions. The main extensional
basin forming mechanisms of faulting, thermal subsidence, loading and compaction are described by simple models and then
applied to a regional cross-section over the Magnus Field in the northern North Sea.
The four processes can be approximated by simple numerical models. Upper crustal extension can be modelled using simple
geometries. Thermal subsidence can be modelled using a cooling slab model in which the rate of subsidence decreases with
increasing time after rifting. Bending of the lithosphere can be modelled using an elastic plate whereby the deflection created
by the load is spread over an area larger than the load itself. Compaction can be modelled as an exponential function where
porosity varies with depth of burial.
All these models have been used to restore the present day depth section sequentially back through time. The effects of
compaction on the underlying sediments has been accounted for, allowing them to expand to their pre-burial thicknesses. The
thermal subsidence has been estimated and added back into the section. The load of the overlying units has been removed
resulting in lithospheric rebound.
Restoration to Early Cretaceous times indicates that the Magnus Sandstone Member was uplifted above sea-level and could
be sub-aerially eroded. This is consistent with both seismic stratigraphic and geochemical evidence.
The regional development of a rifted margin is illustrated by the restorations and shows the relative influence of the basinal
processes modelled here.

Introduction The thermal anomaly gradually decays as the litho-

sphere cools down towards its pre-rifted, equilibrium
Application of numerical basin modelling to temperature. Figure 1 illustrates the combination of
petroleum exploration aids in reducing the risk of upper crustal and lithospheric responses to stretching
finding commercial quantities of hydrocarbons. The that will be modelled.
confidence in a geological model can be increased A basin formed by stretching can accumulate sed-
if the interpreted sequence of events can be fitted iments which will represent an external load on
into a quantitative regional setting. This study uses the lithosphere. The lithosphere will deform due to
a regional geological cross-section which crosses the this load, thus creating further space for sediments
Magnus Field in the northern North Sea to show (Watts et al., 1982). As sediments are buried they
that simple models for the main basin forming mech-
undergo a loss in porosity due to mechanical com-
anisms can be combined to produce quantitative
paction (Sclater and Christie, 1980). Loading and
reconstructions.
compaction can drive basin subsidence.
The main basinal processes that combine to form The basinal processes outlined above can be de-
the stratigraphy of an extensional sedimentary basin scribed using simple numerical models. The models
such as the North Sea are structure, thermal sub- to be used to reconstruct the Magnus regional line
sidence, sediment loading and compaction. The up- are discussed below.
per crust responds to stretching by brittle failure
which is expressed as faulting (Hellinger and Sclater,
Structure
1983; Leader and Gawthorpe, 1987; and McKenzie,
1978a). The lower crust and lithosphere respond to A simple model for the response of the upper crust
extension by necking and ductile flow that intro- to stretching is illustrated in Fig. 2. The dimensions
duces hotter less dense material which results in of the model are 100 km long by 10 km deep. The
thermal uplift (McKenzie, 1978a; Royden and Keen, initial unstretched block is shown in Fig. 2a with a
1980; Sawyer, 1986; and White and McKenzie, 1988). listric fault soling out at 12 km depth. The depth

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 221-229. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
222 R. Young

T
0 100 400
all values in kilometres
Fig. 1. A simple sketch diagram showing the modelled response of the lithosphere to extension. The upper crust responds by brittle
failure resulting the formation of faulted basins. The lower crust and lithosphere responds by ductile flow and necking which results in
regional uplift as hot asthenospheric material is brought into the lithosphere.

to detachment used here is within the range of 10- asthenosphere material and the denser mantle ma-
15 km suggested by McKenzie (1978b). Figure 2b terial causes a regional uplift across the stretched
shows the result of moving the hanging wall block 4 basin.
km. The gap opened up between the hanging wall The hot asthenosphere gradually cools, losing heat
and footwall can be filled by the vertical shear of by lateral and vertical conduction and the uplift de-
the hanging wall (Gibbs, 1983). Figure 2c shows cays as the material becomes denser. This thermal
the result of filling the gap from Fig. 2b and the subsidence can be modelled as a cooling slab result-
formation of a half-graben structure. The area in ing in an exponential decay with time (Sleep, 1971;
Fig. 2c coloured pink represents a load that has been Parsons and Sclater, 1977).
removed from the lithosphere. The effect of regional The response of the lithosphere to stretching can
compensation of the unloading can be calculated and be summarised as a combination of structural and
is shown in Fig. 2d. The model used for calculating thermal effects (Fig. 1). There will be an initial
the deformation is discussed below and its main fault controlled subsidence, which occurs relatively
effect is to spread the deflection over an area larger rapidly, followed by a gradual longer term subsi-
than the dimensions of the load. This simple model dence as the lithosphere cools back towards its equi-
can be used to quantify footwall uplift effects. The librium thermal profile (McKenzie, 1978a; Royden
magnitude of the footwall uplift in Fig. 2d has a and Keen, 1980; and Sawyer, 1986).
maximum of 700 m and extends 35 km away from the
fault.
Loading
The deformation modelled in Fig. 2 assumes that
the upper crustal extension event occurs relatively The response of the lithosphere to external loads,
quickly and that no sediments fill in the newly such as sediments filling a basin, can be modelled as
formed hole. As the graben fills with sediments an elastic plate overlying a fluid substratum (McKen-
the lithosphere will deform under the load and the zie and Bowin, 1976; Watts et al., 1980; and McNutt,
footwall will be deflected downwards. 1984) and as illustrated in Fig. 3. The load is sup-
ported by a combination of the plate rigidity and
Thermal uplift the buoyancy effect of the density contrasts between
the load and the displaced fluid (the mantle in this
The lower crust and lithosphere are assumed to case). A major effect of including plate rigidity is that
respond to extension by necking rather than faulting deflection is spread over a large area. The rigidity
(Royden and Keen, 1980; and Sawyer, 1986). Stretch- of the lithosphere can be usefully described by a pa-
ing raises the lithosphere/asthenosphere boundary rameter called the effective elastic thickness (EET).
above its equilibrium depth and replaces dense man- The EET is the thickness an elastic plate would have
tle material by hotter less dense asthenospheric to have in order to respond to loading in the same
material. The density contrast between this hot manner as the lithosphere.
Restoration of a regional profile across the Magnus Field in the northern North Sea 223

FORWARD MODEL: initial block model

0 20 40 60 80 100

FORWARD MODEL: fill gap with simple vertical shear

0 20 40 60 80 100

FORWARD MODEL: remove load from lithosphere

0 20 40 60 80 100

FORWARD MODEL: lithosphere flexurally rebounds (EET = 5 km)

0 20 40 60 80 100

vertical exaggeration values in kilometres

Fig. 2. A simple forward model of upper crustal extension with associated footwall uplift, (a) The initial unstretched block with a listric
fault trace; (b) extension causes a gap to open up between the hanging wall and the footwall; (c) this gap can be filled by simple
vertical shear with the area of the basin indicated by the pink polygon; (d) deforming the lithosphere by assuming it behaves like an
elastic plate the rigidity of which can be expressed as an effective elastic thickness of 5 km. Both the footwall and, to a lesser extent,
the hanging wall are deformed. Footwall uplift has a maximum of 700 m.

Sediments infilling a basin represent a load on Compaction

the lithosphere that will deform the basin and the
basin margins. The amount of subsidence due to The reduction in porosity of sediments as they are
the load and the distance that the deformation will buried can be modelled as a mechanical compaction
extend depends on the density of the sediment fill process which depends on the initial or surface
and the rigidity of the lithosphere. Flexural rigidity porosity and the rate of decrease of porosity with
has been proposed as a simple mechanism to pro- depth. Sclater and Christie (1980) showed that the
duce a "steer's head" geometry (Dewey, 1982) of a observed data from wells can be fitted by an expo-
sedimentary basin. nential decay curve. Different lithologies will have
224 R. Young

external load Porosity (percentage)

o
() 20
I
40
1
60
1
80
1

1-

Effective Elastic Thickness (EET) used to describe RIGIDITY of plate

EET range typically 5 - 3 0 km Deformation extends 10's to

2-
ft
100's of kilometres from load
Fig. 3. The elastic plate model for loading/unloading the litho-
sphere. Loads are supported by a combination of the density
3- r
contrast between the load and the "fluid" mantle and the rigidity t
of the plate.
4-

different initial porosities and rates of compaction.

*?
Figure 4 shows porosity plotted against depth for
shales from four wells across the regional line to be
5- w -Porosity from logged
velocity values.

used in this study. The porosities have been derived -Porosity from logged
density values.
from both logged sonic and density values. 6-

Fig. 4. A plot of porosity, calculated from both logged density

Application of models to the Magnus and sonic values, against depth for four wells from the vicinity
regional line of the Magnus regional line. Values are for shale only and can
be reasonably approximated by an exponential decay curve from
The simple models of basinal processes outlined a surface porosity of 65%.
above can be applied to a regional geological cross-
section. The example used here is based on a 100
km long composite seismic line which crosses the and the magnitude of the sub-aerial exposure of the
Magnus Field from south south east to north north Magnus Fault block will be quantified.
west and extends across the Shetland Trough to the Well data were used to constrain the compaction
Nordfjord High in the northern North Sea (Fig. 5). constants for the decompaction calculations. Porosity
Figure 6 shows the present day depth section that will data were estimated from both density and velocity
be reconstructed. The Magnus oilfield (De'Ath and measurements and plotted against depth for different
Schuyleman, 1981) comprises an east dipping tilted lithologies. The depth trends, such as that shown in
fault block structure, with the reservoir sandstone, Fig. 4, were fitted with exponential curves, to define
the Magnus Sandstone Member (MSM), within the surface porosity and rate of decrease of porosity with
Kimmeridge formation, truncated in the west by the burial depth (Sclater and Christie, 1980; and Barton
Base Cretaceous unconformity. The Magnus Field and Wood, 1984).
lies within the blue coloured sequence in Fig. 6
at a depth of 3 km between 60 and 80 km on the Estimating thermal subsidence
section. All the sediments above the Base Cretaceous
unconformity are interpreted as post-rift infill. Three of the main basinal processes, loading,
The objective of the modelling work is to try compaction and structure, can be removed from
and quantify the relative contributions of the basinal the present day depth section to show the rela-
processes. In particular the palaeo-bathymetry will tive palaeo-geography across the profile using only
be estimated across the section from the present day the input data of compaction constants, matrix den-
to the Early Cretaceous, the effects of compaction sities and the structural interpretation. The ther-
on the Early Cretaceous sequences will be estimated mal subsidence calculation, however, requires some

Fig. 5. Location sketch map for the regional line used in this study. The cross section (Fig. 6) is 100 km long and crosses the Magnus
Field, into the Shetland Trough and up onto the Nordfjord High.
Fig. 6. Regional present day depth section based on a composite seismic line. Cretaceous sequences are red, pink and brown. Tertiary
sequences extend from seabed to top brown. The present day bathymetry across the line averages 200 m with deeper water to the
leftmost edge (400 m). The Magnus Field is located at the crest of the tilted fault block between 60 and 80 km at a depth of 3 km.
The present day relief on the top red sequence (an Early Cretaceous reflector) is 1500 m.
Restoration of a regional profile across the Magnus Field in the northern North Sea 225

Fig. 5.

MAGNUS REGIONAL LINE: Present day depth section

Vertical eiaggeration - x 2

Post-rift
Tertiary I Post-rift
Late Cretaceous
Post-rift
Mid Cretaceous

Post-rift Syn-rift I Pre-rift

Early Cretaceous Upper Jurassic Jurassic to
I Permian
Fig. 6.
226 R. Young

knowledge of the palaeo-bathymetry upon which the

exponentially decaying model can be based. Ideally 800- Modelled u p p e r crustal
several good markers such as coals would be used at E
stretching = 1.12

a variety of positions across the model and over a L 600-

Modelled lithospheric
E stretching = 1.4
significant time span. V
A
The most useful palaeo-bathymetry marker, or T 400-
trace as defined by Bertram and Milton (1989), in I
the present study is the Brent Group which has been o
N 200-1 Top Brent
interpreted as a shallow marine sandstone (Parry et Thermal Subsidence starts

al., 1981). Applying the models of compaction, flexu-

ral loading and structural restoration with accompa- 200
K-t-
"T
160 120
I
80
I
40 0
nying footwall deformation to the present day depth AGE IN MILLIONS OF YEARS
section to obtain a restored section at Top Brent Fig. 7. Thermal subsidence calculated for the cross-section using
time results in palaeo-bathymetry values of 820 m the Brent Formation as a palaeo-bathymetry marker and starting
thermal decay at 138 Ma (the timing of the last phase of
at the rightmost, least deformed, edge of the sec- faulting). The stretching factors used were 1.12 in the upper
tion. The difference between the palaeo-bathymetry crust (as calculated from restoring the section) and 1.4 for the
from modelling three of the basinal processes and lower crust and lithosphere These were required to obtain 820 m
the palaeo-bathymetry interpreted for the Brent subsidence from Early Cretaceous to the present day.
is assumed to be due to the thermal subsidence
from the time of rifting (138 Ma) to the present
day. Reconstruction of the regional line
The structural restoration to Top Brent resulted through time
in a total shortening of the section from 100 km to
89 km which represents a beta factor of only 1.12. Restoration to Early Tertiary times, Fig. 8, pro-
If a simple McKenzie model (McKenzie, 1978a) is duces a section with palaeo-bathymetry (300 m) only
assumed, with uniform stretching in upper crust and slightly greater than the present day depth section
lithosphere, then this gives insufficient thermal uplift (200 m). The pre-Tertiary sediments have "fluffed
to bring the Brent to its interpreted water depth. If a out" to their pre-burial thicknesses. The lithosphere
two layer stretching model is used then a lithospheric has been allowed to rebound by removing the load of
stretching factor of 1.4 must be input to get 820 m of the Tertiary to recent sediments and the proportion
thermal uplift and restore the MSM to its interpreted of the thermal subsidence from 65 Ma to the present
palaeo-bathymetry. Figure 7 shows the modelled day, a total of 200 m, has been added back into the
thermal subsidence for the profile plotted against profile. It was not necessary to restore structurally at
age in millions of years. The model being used here this time.
assumes that thermal subsidence commences at 138 The unloading calculation was estimated by assum-
Ma, which is the interpreted age of the Magnus ing a rigidity for the plate characterised by an EET
Fault. of 15 km. However, this value is not critical, since
The full reconstructions using models for com- the load that was removed extends almost uniformly
paction, loading, structure and thermal subsidence across the model. The calculation is relatively insen-
are shown and discussed for several time steps in the sitive to EET variations for uniform thickness loads.
following sections. The inclusion of the elastic plate in the calculation

Fig. 8. Restoration to Base Tertiary times. The section has been allowed to rebound due to removing the load of the tertiary sediments.
The underlying Cretaceous and older sediments have been restored to their pre-burial thicknesses and the thermal subsidence from
Fig. 7 into the model.
Fig. 9. Restoration to Early compaction, loading and thermal subsidence have been applied and the Magnus Fault the top red sequence
has been removed due to decompaction.
Fig. 10. Restoration to Earliest Cretaceous times results in further uplift of the Magnus Fault block and indicates a deep (3.75 km)
basin, the Shetland Trough, at this time. Calculated sub-aerial exposure is supported by geochemical evidence of porosity enhancement
by meteoric water leaching.
Fig. 11. Reconstruction to Base Cretaceous before movement occurs along the Magnus fault. The Kimmeridge Formation (coloured
blue) thickens into a fault towards the righthand (eastern) edge of the section. Submarine fan sandstones (the Magnus Sandstone
Member) were deposited during Kimmeridge times before finning upwards into the Upper Kimmeridge organic rich mudstones. The
observed uplift on the Magnus fault block (Fig. 10) of 500 m can be modelled as a response to extension (Fig. 1). As the hanging wall
block is replaced to its pre faulted position the footwall uplift must be removed.
Restoration of a regional profile across the Magnus Field in the northern North Sea 227

MAGNUS REGIONAL LINE: Restored to Base Tertiary

North 0 20 40 60 80 Km South

Vertical exaggeration - x 2
Fig. 8.

MAGNUS REGIONAL LINE: Restored to Early Cretaceous

North 0 20 40 60 8 0 Km South
East
West _| , | i ,

Vertical exaggeration - x 2

Fig. 9.

MAGNUS REGIONAL LINE: Restored to Base Cretaceous (after faulting)

North 0 20 40 60 80 Km South
l E a s
West I I i

Vertical exaggeration - x 2
Fig. 10.

MAGNUS REGIONAL LINE: Restored to Base Cretaceous (before faulting)

North 0 20 40 60 80 Km South
East
West _i | | , ,

Vertical exaggeration - x 2
Fig. 11.
228 R. Young

serves to iron out the local bumps that would occur A structural restoration has been attempted to
if local isostatic response was used. replace the major movement on the Magnus Fault
The implication of the modelling work at this time and the smaller movements on the other faults.
suggests that a small hole, namely 300 m palaeo- During faulting, footwall uplift will occur as the
bathymetry, can be filled with a considerable thick- hanging wall load is removed from the lithosphere as
ness of sediments (in this case 1500 m). The hole was discussed earlier in the simple structural model
in which the sediments are being deposited gets outlined in Fig. 2.
deeper through time due to a combination of thermal To replace the hanging wall of the Magnus Fault
subsidence (200 m), lithospheric response to load- we must include deformation of both the hanging
ing and compaction of underlying sediments during wall and the footwall. Figure 11 shows the structural
burial. restoration of the hanging wall. Gaps and over-
Reconstruction to Early Cretaceous times (Ap- laps are not considered as serious errors during the
tian-Barremian, 125 Ma) is shown in Fig. 9. The restoration, but serve to highlight the simplicity of
same three models of the basinal processes have the models used here.
been accounted for as in the previous restoration The syn-rift sequences are shown to thicken into
step. However, there is now a considerable depth a fault to the east of the Magnus Field. There
of water across the section with shallow regions have been two phases of extension in this area. The
associated with the tilted fault block structures of earlier, Callovian/Oxfordian, phase was controlled by
Magnus and the Nordfjord High. The Magnus Fault approximately north-south orientated faults (Fig. 5)
block, between 60 and 80 kilometres in the fig- into which the sequences containing the Magnus
ures, is sub-aerially exposed with a steep drop into Sandstone Member were deposited. A later phase
the Shetland Ttough, down the Magnus Fault, into of rifting in Base Cretaceous times orientated more
palaeo-bathymetries of 2000 m. north-east to south-west formed the main Magnus
Palaeo-bathymetries of the order of 2000 m have Fault that cuts the earlier syn-rift sequence and
been interpreted by Bertram and Milton (1989) from uplifts them on the footwall block. Thus, Upper
seismic stratigraphy in the North Viking Graben. Jurassic submarine fan sandstones are preserved at
Nelson and Lamy (1987) also interpret large depths the top of a fault block tilted during the Early
at this time from a depth section close to that used Cretaceous.
here. However, they interpret sediments at the base
of the Shetland Trough as being Upper Jurassic
rather than Early Cretaceous as in this study. Conclusions
There has been a dramatic change in the topog-
The results of this study show that simple models
raphy of the top of the red sequence between the
of the main basinal processes can be used to produce
present day depth section (relief was 1500 m) and
quantitative reconstructions starting from a present
the reconstruction shown in Fig. 9. The top of the
day geological model.
red sequence is now flat. The observed topography
(1) The palaeo-bathymetry during the Tertiary was
on the present day depth section can be attributed
little different from that observed in the present
to differential compaction of the red sequence as it
day. The 1500 m of Tertiary sediments have been
is buried by almost 5 km of sediment. This is consis-
deposited into a 300 m hole due to a combination of
tent with a seismic stratigraphic interpretation that
thermal subsidence, loading and compaction.
the Early Cretaceous sequences are a deep marine
passive infill phase. (2) Relief observed on Early Cretaceous sequences
can be attributed to compaction effects when buried
Reconstruction to Base Cretaceous is shown in
by up to 5 km of later sediments.
Fig. 10. The loading and compaction effects of the
(3) The main result of footwall uplift of the Mag-
Lower Cretaceous sediments have been removed and
nus Fault block could have been interpreted from
more thermal subsidence has been added back into
seismic stratigraphy or geochemistry. The basin mod-
the model (total of 800 m). The result is that the
elling carried out in this study increases confidence
Magnus Fault block is raised approximately 500 m
in the geological model by putting the observations
above sea level. Geochemical evidence (Emery et al.,
into a quantitative regional setting.
submitted to Geology) suggests that the sandstones
towards the top of the Magnus Sandstone Mem-
ber have been flushed by meteoric water resulting Acknowledgments
in porosity enhancement of the order of 5%. The
geochemical evidence is consistent with sub-aerial BP gave permission for this paper to be published.
exposure (Shanmugan, 1987). The author would like to acknowledge the following
Restoration of a regional profile across the Magnus Field in the northern North Sea 229

co-workers Iain Hutchison, Mark Shann, David Jack- J. Geophys. Res., 89: 11180-11194.
son and Liz Lasseter with whom the regional study, Nelson, P.H.H. and Lamy, J.-M., 1987. The More/West Shetlands
area: a review. In: J. Brooks and K.W Glennie (Editors),
of which this is only a part, was carried out. Petroleum Geology of North West Europe. Graham and
Trotman, London, pp. 775-784.
References Parry, C.C., Whitley, P.K.J. and Simpson, R.D.H., 1981. Integra-
tion of palynological and sedimentological methods in facies
Barton, P. and Wood, R , 1984. Tectonic evolution of the North analysis of the Brent Formation. In: L.V. Illing and G.D.
Sea Basin: crustal stretching and subsidence. Geophys. J. R. Hobson (Editors), Petroleum Geology of the Continental
Astron. Soc, 79: 987-1022. Shelf of North West Europe. Institute of Petroleum, London,
Bertram, G.T. and Milton, N.J., 1989. Recontructing basin pp. 342-351.
evolution from sedimentary thickness; the importance of Parsons, B. and Sclater, J.G., 1977. An analysis of the variation
palaeobathymetric control, with reference to the North Sea. of ocean floor bathymetry and heat flow with age. J. Geophys.
Basin Res., 1: 247-257. Res., 82: 803-827.
De'Ath, N.G. and Schuyleman, S.F., 1981. The geology of the Royden, L. and Keen, C.E., 1980. Rifting process and thermal
Magnus oilfield. In: L.V. Illing and G.D. Hobson (Editors), evolution of the continental margin of Eastern Canada deter-
Petroleum Geology of the Continental Shelf of North West mined from subsidence curves. Earth Planet. Sei. Lett., 51:
Europe. Institute of Petroleum, London, pp. 342-351. 363-361.
Dewey, J.F., 1982. Plate tectonics and the evolution of the British Sawyer, D.S., 1986. Effects of basement topography on sub-
Isles. J. Geol. Soc. London, 139: 371-412. sidence history analysis. Earth Planet. Sei. Lett., 78: 427-
Gibbs, A.D., 1983. Balanced cross-section construction from 434.
seismic sections in areas of extensional tectonics. J. Struct. Sclater, J.G. and Christie, P.A.F, 1980. Continental stretching:
Geol., 5: 153-160. an explanation of the Post-Mid-Cretaceous subsidence of the
Hellinger, S.J. and Sclater, J.G., 1983. Some comments on two- Central North Sea. J. Geophys. Res., 85: 3711-3739.
layer extensional models for the evolution of sedimentary Shanmugan, G., 1987. Origin, recognition, and importance of
basins. J. Geophys. Res., 88: 8251-8269. erosional unconformities in sedimentary basins. In: K. Klein-
Leader, M.R. and Gawthorpe, R.L., 1987. Sedimentary models spehn and C. Paola (Editors), New Perspectives in Basin
for extensional tilt-block/half-graben basins. In: M.P Cow- Analysis, pp. 83-108.
ard, J.F. Dewey and P.L. Hancock (Editors), Continental Sleep, N., 1971. Thermal effects of the formation of Atlantic
Extensional Tectonics. Geol. Soc. London, Spec. Publ., 28: continental margins by continental breakup. Geophys. J. R.
139-152. Astron. Soc, 24: 325-350.
McKenzie, D.P., 1978a. Some remarks on the development of Watts, A.B., Bodine, J.H. and Steckler, M.S., 1980. Observations
sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32. of flexure and the state of stress in the oceanic lithosphere. J.
McKenzie, D.P., 1978b. Active tectonics of the Alpine- Geophys. Res., 85: 6369-6376.
Hymalayan belt: the Aegean Sea and surrounding regions. Watts, A.B., Karner, G.D. and Steckler, M.S., 1982. Lithospheric
Geophys. J. R. Astron. Soc, 55: 217-254. flexure and the evolution of sedimentary basins. Philos. Trans.,
McKenzie, D.P. and Bowin, C , 1976. The relationship between R. Soc. London Ser. A, 305: 249-281.
bathymetry and gravity in the Atlantic Ocean. J. Geophys. White, N.J. and McKenzie, D.P., 1988. Formation of the "steer's
Res., 81: 1903-1915. head" geometry of sedimentary basins by differential stretch-
McNutt, M.K., 1984. Lithospheric flexure and thermal anomalies. ing of the crust and mantle. Geology, 16: 250-253.

R. YOUNG BP Exploration, London, U.K.

Present address: BP Norway Limited U.A., Fonisbeen 35, P.O. Box 197, N-4033 Forus, Norway
 231

Patterns in the development of extensional fault-block

shapes from comparison of outcrop-scale faults and
experimental physical models

S.E. Laubach, B.C. Vendeville and S.J. Reynolds

The geometry and kinematics of normal fault blocks and subsidiary faults within fault blocks are illustrated by centimeter-
to meter-scale faults in an exposure of Tertiary sedimentary rocks in Arizona and are interpreted using results from previously
published physical models. In map view, the outcrop fault population consists of anastomosing curved and planar faults. In
cross-section there are planar faults and curved faults with either smooth or abrupt (ramp-flat) curvature. The fault patterns
resemble those of physical models and larger faults in other extended terrains. Ramp-flat faults with planar and curved faults
and tilt-block domains are the most prominent structures. Ramp-flat geometry prevails where faults intersect bedding at low
angles. Rats are commonly zones of distributed faults rather than a simple throughgoing fault. In tilt-block arrays curved
faults that are convex-up are as numerous as concave up (listric) faults and contribute to the overall geometry of the fault
array. Subsidiary planar and curved faults prevail in the toes of tilt blocks, whereas horst-and-graben patterns occur in the
less faulted block interiors. Internal deformation of rollover structures due to both accommodation and extension of the
upper plate was accomplished mainly by slip on arrays of smaller faults in synthetic tilt-block arrays and by crisscrossing
horst-and-graben-bounding faults.

Introduction interpretation of fault kinematics. The results lend

insight into the way in which normal faults interact to
The structural evolution of normal fault blocks produce fault block shapes and how, in larger fault
governs the shape, size and distribution of potential blocks, fault complexity that is below seismic resolu-
reservoir rocks in many extended terrains. Seismic tion may control reservoir compartmentalization and
data offer good information on the configuration heterogeneity.
of major faults, but resolution limits observation
of structures that accommodate displacement along Geologic setting of fault example in Arizona
major fault planes and distortion within major fault
blocks. Restored sections indicate that subsidiary The study area in west-central Arizona is in
structures must contribute to the structural evolution the Basin and Range province of the southwest-
of fault blocks, but such complexity is difficult to ern United States, a region transected by numerous
document using seismic data alone. One way to gain Tertiary detachment faults (e.g., Wernicke, 1981).
a better understanding of the development of the These faults are large-displacement, low-angle nor-
fault-block shape is to study the geometry of small- mal faults that have been uplifted and warped into
displacement fault arrays in naturally deformed rocks broad, undulating surfaces. The extension and lat-
and the geometry and kinematics of faults in scaled eral displacement of upper-plate rock was accom-
physical models. modated by the formation of numerous upper-plate
This paper describes faults in an outstanding fault zones. The Lincoln Ranch basin of the eastern
three-dimensional exposure in Tertiary sedimentary Buckskin Mountains (Fig. 1) is above the Buckskin
rocks from the Basin and Range province in Ari- detachment fault in a domain of northeast-dipping
zona and compares them with fault geometries in normal faults.
physical models. Because experimental models are Within the Lincoln Ranch basin there is a canyon
created under known conditions of displacement and exposure (Fig. 2) where the geometry of centimeter-
imposed bulk strain, similarities in geometry between to meter-scale normal faults can be examined lo-
models and the outcrop can be used to refine the cally in 100% outcrop in both map view and in

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 231-241. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
232 S.E. Laubach, B.C. Vendeville and S.J. Reynolds

EXPLANATION
Quaternary surficial deposits

Upper-plate rocks

Lower-plate rocks

/..... High-angle fault, dashed where

/ ' approximate or inferred, dotted
where concealed

/"\ Buckskin-Rawhide-Bullard
detachment fault, dotted
where concealed
QA12539c

Fig. 1. Structural setting of the Lincoln Ranch basin in the Buck- | | Q / Anticline
skin Mountains, west-central Arizona (modified from Spencer j >/ Fault.with downthrown side,
and Reynolds, 1989). ...!.'... __ /Z striation trend, and dip direction
iiiiiiiiiiii Tue
:":,".""" Detachment fault,
Ijijlllj Tus dotted where concealed
cross-section. The faults are in the lower Miocene | 1 MzXc 2 5 \ Bedding attitude
Chapin Wash Formation, which consists there of j ^ j ^ l Mcs, Ms, Mc Horizontal bedding
fine-grained red to reddish-brown sandstones and Fig. 2. Geologic map and cross-section of part of the Lin-
siltstones. Sandstone beds range in thickness from coln Ranch basin showing the setting of the normal fault
exposures (modified from Spencer and Reynolds, 1989). Q =
centimeters to decimeters, but are typically 10 cm undiiferentiated Quaternary gravel and sand; T = undifferenti-
thick or less. Siltstone beds and laminae range from ated Tertiary rocks; Tue = upper conglomerate (Tertiary Chapin
2 mm to 3 cm thick. Beds do not thicken across Wash Formation); Tus = upper sandstone (Tertiary Chapin Wash
faults. Grain fracturing and gouge formation along Formation); MzXc = crystalline rocks of upper plate, undivided
(Tertiary[?], Mesozoic[?] and Proterozoic[?]). Lower-plate rock:
faults show that the unit was lithified prior to exten- Mcs = highly tectonized metasedimentary and crystalline rocks;
sion. Faulting in this exposure of the Chapin Wash Ms = tectonized metasedimentary rocks; Mc = mylonitic crys-
Formation occurred at burial depths of less than talline rocks.
2 km (Spencer and Reynolds, 1989).
(1 to 10 cm) faults and as much as 2 to 3 cm thick on
General characteristics of the faults proportionally larger faults. Slickenline striations on
fault surfaces plunge down-dip and consist of straight
The faults we measured have offsets that range to slightly curved polished ridges and grooves that
from several millimeters to 50 cm; larger faults have match on opposed fracture surfaces.
offsets of several meters to tens of meters. Faults
range from planar to curved. Individual faults are Fault pattern in map view
exposed in cross-section for as much as 35 m and
can be traced in map view for tens of centimeters to In map view, faults range from straight to con-
several meters. Spacing ranges from tens of meters tinuously curved to abruptly curved traces. Abruptly
for large faults, meters to tens of centimeters for curved traces are composite surfaces comprising in-
medium-sized faults, to centimeters for small faults. terconnected planar and curved parts. Traces are
The mean fault strike is northwest. concave and convex toward the hanging wall, with no
Faults have distinct, striated surfaces with no evi- predominance of either shape. Some faults change
dence of soft-sediment remobilization or injection of from concave to convex along strike. Some faults
fault gouge into their wall rocks. The indurated fault have irregular or arborescent patterns consisting of
gouge consists of aligned, scaly, silt-sized particles short, planar parts and curved parts. Faults are com-
and is less than 1 mm thick on small-displacement monly segmented, with segments in subparallel en
Patterns in the development of extensional fault-block shapes 233

indicates a gradual change from dip slip to oblique

slip along curved fault surfaces.
The faults in map view have a characteristic elon-
gated diamond-shaped, rhomboid, or lenticular pat-
tern created by intersecting planar faults and curved
faults (Fig. 3). Aspect ratios of the fault-bounded
lozenges are generally 3:1 to 5:1. The long axes
of the fault-bounded lozenges within an outcrop are
colinear and perpendicular to the striation trend on
the fault surfaces. Diamond patterns in sandstone
can result from anastomosing faults that dip in the
same direction, consistent with the predominance of
domino-style fault sets observed in cross-section, but
crisscrossing faults with opposed dips also form these
patterns.
Faults terminate along strike by losing slip or by in-
tersecting another fault. The slip decreases abruptly
near the fault termination itself. Along-strike loss of
slip is locally accompanied by development of splays
that branch from the main fault trace at a low angle
(Fig. 3b). Anastomosing faults locally abut and ter-
minate against other faults along strike. Some faults
also crosscut and offset other faults at a low angle.
The mutual crosscutting faults with different strike
suggest simultaneous development of multiple faults.
This fault pattern resembles those common among
large (kilometer-scale) faults in other extended ter-
rains (e.g., Donath, 1962).

Fault pattern in vertical section

The main fault types are curved faults, planar

faults and faults with ramp-flat geometry. Curved
faults display both downward-shallowing (concave-
up) listric and downward-steepening (convex up)
"negative listric" shapes (Figs. 4 and 5). Curved
faults have curvature radii of 10 cm to several meters.
Curved faults and planar faults are approximately
equally abundant and concave-up and convex-up
curved faults are also equally represented. Concave-
Fig. 3. (a) Map and cross-section view of anastomosing faults in up and convex-up faults produce relative stratal rota-
domino-style array. Tracing from photograph, (b) Map view of
fault termination. Scale is 30 cm. (c) Map view of fault pattern
tions of opposite sense, and the equal abundance of
illustrating branching faults and crosscutting faults. F = main both fault types results in approximately constant bed
faults. Coin is 1.7 cm wide. dips across the exposure. Convex-up and concave-up
faults that occur in the same area may effectively
echelon or relay sets. Faults locally intersect and cancel out each other's stratal rotations even on a
crosscut their neighbors at low and moderate angles, local (meter) scale (Fig. 4a). Curved faults exist in
but no faults were observed paralleling the extension both sandstone and siltstone, and many curved faults
direction that could be interpreted as transfer faults. cut across bedding without changing geometry. This
Adjacent faults interact with each other by form- is true for both large- and small-displacement faults.
ing complex systems of lateral ramps. Within lateral Ramp-flat geometry occurs where faults cut bed-
ramps, small faults locally have trends that are highly ding at a low angle (less than 45), whereas faults
oblique to the main fault trend. Numerous small with steep dip (50-70) tend to cross bedding with
faults are localized where segments of large faults no deflection as they traverse different rock types.
overlap. Slip sense derived from slickenline trends Ramps are dominantly in sandstone, and flats are
234 S.E. Laubach, B.C. Vendeville and SJ. Reynolds

(a)
Southwest Northeast

Fig. 5. (a) Curved normal fault in sandstone. Curvature changes

from concave up to convex up. Scale is in centimeters, (b)
Progressive changes in bed tilt across several curved (steepening-
(c) downward) faults. Scale is 10 cm.
Southwest Northeast

mutually crosscutting arrays of curved and domino-

style faults. These faults locally are superimposed on
early formed fault-bounded lenses. Older faults in
flats, which are crosscut and displaced by younger
faults, are much less continuous and more difficult to
recognize than younger faults. Subsidiary faults that
have both gentle and steep dips are particularly com-
mon within and above flats. Steeply dipping faults are
Fig. 4. Convex-up (downward-steepening) faults in cross section.
(a) Decrease in bed tilt across downward-steepening fault, (b) truncated by gently dipping faults that define lenses
Small faults localized at bend in curved fault, (c) Fault with in which bedding is rotated. Mutually crosscutting
concave and convex segments in profile. faults and block rotation can cause abrupt thinning
of beds. Closely spaced superposed faults can be so
pervasive that the layers within the flat undergo 50%
mostly in siltstone (Figs. 6 and 7). Ramps are gen- to 60% thinning.
erally distinct faults or zones of several subpar- Continuously curved and abruptly curved (ramp-
allel faults. Subsidiary faults within ramps define flat) faults have antithetic and synthetic faults in
fault-bounded lenses that are elongated parallel to their upper plates that define rollover, tilt-block and
the strike of the master fault. Typically, ramps are horst-and-graben structures. Most common are sub-
convex-up (negative listric) in their upper part and sidiary faults that dip in the same direction, but
concave-up (listric) in their lower part. Ramps with more steeply than the master low-angle slip surface,
greater displacement are straight and are associated defining synthetic tilt-block domains. Synthetic and
with numerous subparallel fault slivers. Some faults antithetic faults locally crosscut one another. Upper-
that flatten and become parallel to siltstone layers do plate rocks above listric or ramp-flat faults commonly
not reemerge from siltstone layers within the scale of contain rootless graben and horsts bounded by pla-
the outcrop because displacement is accommodated nar or (locally) ramp-flat faults (Fig. 7). These horsts
by bedding-plane slip on multiple surfaces. separate fault domains with opposed fault vergence
Flats are complex zones of distributed shear that is and curvature, and they are less deformed than adja-
manifested in bedding-parallel slip and intersecting, cent parts of the upper plate.
Patterns in the development of extensional fault-block shapes 235

|j*#W^

i^MS^
r
^%^%^ "i: w^^

Flat
Krasses^
a^ * ^

Fig. 6. (a) Fault with ramp-flat geometry resulting in overall fault curvature. R ramp; F = flat. Scale is 35 cm. (b) Concave-up fault
with subsidiary faults and rollover in the hanging wall. Scale is 10 cm.

Internal deformation of rollover structures due to ther uniform or variable dip. Where dips locally
both accommodation and extension of the upper change within a fault domain, fault blocks are in-
plate was accomplished mainly by slip on arrays of ternally deformed by smaller faults or by truncation
smaller faults in synthetic tilt-block arrays and by and rotation of domino faults along younger cross-
crisscrossing horst-and-graben-bounding faults. Fault cutting, and locally curved, faults. In some arrays,
sets in upper-plate rocks cannot be interpreted as extensive internal deformation is confined to only a
either vertical (Gibbs, 1983) or nonrotating inclined few domino blocks. The tops and bases of domino
(White et al., 1986) shear planes, but rather they zones are marked by interstratal slip. Domino and
suggest accommodation along a set of shear planes horst-and-graben zones in siltstone are bounded lo-
rotating in domino fashion (Wernicke and Burch- cally above and below by unfaulted sandstone. Some
fiel, 1982). Moreover, upper-plate deformation is not faults within domino arrays of planar faults are
restricted to the area of rollover accommodation, in- curved (concave or convex up) rather than planar.
dicating that the bed-parallel stretch also contributes Faults of a much smaller scale than those of the
to upper-plate faulting. domino blocks occur within individual dominos and
Arrays of planar normal faults with domino (Fig. commonly within only a part of the domino block.
8) and horst-and-graben (Fig. 9) patterns are com- Subsidiary planar and smoothly curved faults prevail
mon at all scales we studied. Domino patterns are in the toes of domino blocks (Fig. 8), whereas horst-
particularly common in sandstone. The angle be- and-graben patterns occur in the less faulted block
tween fault plane and bedding is typically large, interiors. Domino-block edges, where weak bedding
ranging from 60 to 90, with 70 to 80 most com- planes are present, locally display small-offset reverse
mon. Within an array, domino faults may have ei- faults.
236 S.E. Laubach, B. C. Vendeville and SJ. Reynolds

(c)

(b) X > Negative

^listric fault W
^24

15
3>>
Nc^vNfV

\50

\ Listric
^ V \ fault
0 10 cm \ \ \ \ \ V
_1 I "V

Fig. 7. Ramp-flat fault with rollover anticline, (a) Location overview, (b) R = ramp; F = flat; A = rollover anticline; S = upper-plate
fault slivers containing superposed faults; G = upper-plate rootless graben; H = rootless horst. (c) Deformation of hanging wall block
by steepening-downward faults in area distant from rollover. Tracings from photographs.

Fig. 8. Vertical section of tilted fault block within domino-style

array. Small faults accommodate deformation within the block.
S = basal slip surface of array. Width of block is 10 cm.

Some outcrops contain crisscrossing arrays of pla-

nar faults that typically are high-angle faults forming Fig. 9. Criss-crossing faults in outcrop. Tracing from photograph.
horst-and-graben patterns (Fig. 9). Crisscrossing pla- The dotted areas show bedding plane. For experimental analogs
nar faults offset faults with opposed dip or terminate see models presented by Horsfield (1980) and Woods (1988).

against them. Arrays of small planar faults with

horst-and-graben patterns locally abut larger planar in experimental models conducted under controlled
faults. The acute bisectrix of paired faults in horst- boundary conditions (e.g., imposed displacements or
and-graben patterns is at an angle of about 30 to bulk strain). Comparison of outcrop examples and
the associated master faults and is oblique to layer- experimental results help elucidate the fault kine-
ing. Small horst-and-graben patterns are commonly matics in this area.
localized adjacent to larger faults; they are common
in upper- and lower-plate rocks. interpretation of map pattern
The outcrop displays three major fault patterns
Comparison with physical models in map view, with the different patterns confined to
specific areas or domains. In one type of pattern,
Fault geometries in the Lincoln Ranch basin out- lozenge-shaped blocks have fault traces that intersect
crops are similar to geometries of faults produced at small angles; in another, lozenge-shaped blocks
Patterns in the development of extensional fault-block shapes 237

et'al. (1969), where four sets of faults intersecting

at high angles accommodated extension along two
perpendicular directions. In these experiments, the
extension magnitudes in two stretching directions
controlled the attitude and slip along each of the
fault sets. Fault traces crosscutting at high angles in
the field examples (e.g., Fig. 3b) can indicate areas of
L,= 1.5xL 0 local biaxial stretching.
In a third pattern fault traces locally curve toward
and merge with the direction of regional extension
or have a sinuous pattern. Similar geometries are
(b) common in physical models (either sand or clay)
where extension varies laterally across the model,
causing different units of the model to undergo
differential displacements (Cloos, 1968; McGill and
Stromquist, 1979; Vendeville, 1987, 1991; Brix et al.,
1985; Serra and Nelson, 1988). This effect is also
commonly seen near the side walls of models with
Fiiedunit : Moving unit :|: Fixcd unil
lateral friction. In these models, normal faults in the
center of the model or distant from a unit boundary
propagate laterally as oblique- or strike-slip faults
* with curved traces (Fig. 10b). In the field example
Fig. 10. (a) Increase in obliquity between fault traces induced by this is illustrated by complex splay patterns along and
uniform uniaxial stretching (map view). Initial obliquity is 20, at the lateral terminations of curved faults (Fig. 3c).
and becomes 39 with a uniform stretch of 1.5. (b) Fault trace Fault traces are convex toward the unit with greatest
curvature induced by lateral friction. Ticks on the downthrown
displacement and face either the hanging wall or
side of fault. Based on Vendeville (1987).
footwall block, depending on the fault dip within the
unit. Slip sense is synthetic to the shear displacement
have faults that intersect at larger angles (>30; (and shear stress) along the unit lateral boundary
Fig. 3). A third pattern has parallel faults that locally (Fig. 10).
curve toward the direction of regional extension. The Although fault traces that are convex toward the
domains of fault traces intersecting at a small angle footwall block are common in nature (e.g., Shel-
are similar to patterns created in several clay models ton, 1984), fault traces that are concave toward the
subjected to uniaxial stretching (e.g., Cloos, 1955; footwall block have been documented in coal mines
Oertel, 1965; Freund and Merzer, 1976; Kautz and (Walsh and Watterson, 1988) and in transfer zones
Sclater, 1988). Both clay experiments and mathemat- between rift offsets in experimental models (e.g.,
ical simulations (Oertel, 1965; Reches, 1978) showed Faugre et al., 1986; Vendeville, 1988). Our field
that lozenge-shaped fault blocks defined by four sets example shows curved fault traces that are both
of normal faults acting simultaneously can accom- concave and convex to the hanging wall. Both field
modate plane-strain stretching (uniaxial extension). examples and experiments suggest that faults with
Such faults commonly intersect at small angles (15 curved traces are not necessarily listric in vertical
or less), suggesting that outcrop areas with slightly section.
oblique fault traces indicate plane-strain stretching,
not necessarily separate phases of multidirectional Fault shapes in vertical section
stretching. This interpretation is consistent with the Both planar and curved faults exist in the outcrop
unidirectional trend of slip-sense indicators on the example. We define two main categories of curved
fault surfaces. faults. The first category comprises faults that curve
Locally the angles between crisscrossing fault sets on a scale of 10 m or more in the outcrop; the
in the exposure are higher (30 or greater). Such entire fault plane is composed of multiple ramp and
large angles can be interpreted in two ways. First, flat segments, leading to an overall listric geometry.
high angles may be caused by further uniform uni- The second category includes smaller faults (meter
axial stretching of initially slightly oblique crisscross- scale) that are either concave or convex upward;
ing faults that originally intersected at low angle the curvature can be either mechanically controlled
(Fig. 10a). Second, high angles may reflect biax- (initial) or induced by internal distortion of fault
ial stretching, as in the clay models of Hoeppener blocks.
238 S.E. Laubach, B.C. Vendeville and S.J. Reynolds

block deformation can increase curvature of initially

planar or slightly curved faults (Figs. 13 and 14).
A vertical change in the magnitude of horizontal
extension and associated shear strain within blocks
progressively bends the fault plane upward or down-
ward (Fig. 13b, c).
In the outcrop example, convex-up and concave-up
faults are equally abundant and affect both highly de-
formed shale layers and sandstone layers that behave
more rigidly. Determining whether fault curvature
Fig. 11. Fault curvature associated with ramp-flat geometry. (A) is initial or acquired is important in interpreting
Overall fault-plane curvature. Inset shows detail of curved fault
block kinematics. For example, in sandstone, fault
segments. (B) Concave-down segment. (C) Concave-up segment.
curvature controls rotation of adjacent rigid fault
blocks. Downward-steepening faults can drastically
affect the overall geometry of an extended terrain by
VN / / counteracting stratal rotations caused by concave-up

Downward increase
in horizontal stretch
Fig. 12. Curved faults experimentally produced in rollover (in
cross-section) (McClay and Ellis, 1987).

Large-scale (decameter) concave-up fault planes

are formed by a series of ramp and flat segments.
Fault planes climb upsection into sandstone layers,
but they can propagate along bedding in shale layers.
The proportion of the length of the flat to the total
Upward increase
fault length can vary vertically in the outcrop section, in horizontal stretch
leading to overall fault curvature (Figs. 6a, 11a). Ini-
tially, when displacement approximately equals bed Fig. 13. Fault curvature induced by inhomogeneous strain of ini-
thickness, beds influence fault geometry, and there tially planar fault planes. (A) Initial stage. (B) Stretch increases
are abrupt changes in fault dip between ramp and downward. (C) Stretch increases upward.

flat. Faults with ramp-flat geometry have displace-

ment on the same scale as the mechanically signif-
icant bed thickness. If fault displacement is large
compared with bed thickness, faults are more planar;
the fault is smoothed by tectonic erosion. This may
result from formation of duplexes and splays adja-
cent to ramps. Faults that can link ramps into a more
planar surface also may develop from flats (Fig. 7).
Such faults have not been modelled experimentally
in extension so far.
Some small-scale faults bound rigid, unstrained 2 cm
blocks and formed with initial curvature. Examples
are faults that link ramps and flats (Fig. lib, c) and
faults that form in sandstone in rollovers above major
faults. Such small-scale faults can be either convex-
up or concave-up. Experiments of rollover faulting
by McClay and Ellis (1987; Fig. 12) show similar
geometries and suggest that curved stress trajectories
cause initial fault curvature. Fig. 14. Physical models of curved faults that were initially
Other physical models (W.T. Horsfield, in Mandl, planar, (a) Vendeville et al., 1987. (b) Based on W. T. Horsfield
1988, p. 33; Vendeville, 1991) show that internal in Mandl, 1988.
Patterns in the development of extensional fault-block shapes 239

listric and tilt-block faults and, therefore, they can

profoundly affect estimates of extension magnitude
based on the dip of rotated planar faults and bedding
(Wernicke and Burchfiel, 1982). However, in shale
and some siltstone beds, faults are more likely to be
distorted by block strain during extension and, there-
fore, exert little control on block rotation. In contrast
to initially curved faults, curved faults in shale may
indicate block strain.

Block kinematics and strain

In the outcrop, significant fault-block deforma-
tion takes place in three settings: above and below
tilt-block domains in penetratively faulted shale lay-
ers, locally within tilt blocks, and adjacent to major
curved faults in rollover anticlines and zones of dis-
tributed faults.
Tilt blocks have been recognized in many exten-
sional terrains (e.g., Morton and Black, 1975; Prof-
fett, 1977), and they have been reproduced in physi-
cal models (e.g., Vendeville et al., 1987). In models, Fig. 15. Kinematics of block rotation with even (A, B) or uneven
(C-F) fault dips.
domino-style kinematics are generally regarded to
indicate the influence of an undeflectable boundary
(Mandl, 1987, 1988; Vendeville, 1987) along which In the hanging wall block of curved faults, sub-
some shear stress occurs (Brun et al., 1985). In the sidiary structures such as crosscutting synthetic and
outcrop example domino tilt blocks comprise sand- antithetic fault fans, rollover anticlines and crestal
stone layers bounded above and below by shales. collapse graben occur that are closely similar to fault
Zones with tilt blocks are bounded above and below patterns created in analog experiments with sand
by unfaulted layers (generally sandstone) that act as with a rigid footwall (e.g., McClay and Ellis, 1987)
rigid boundaries (Fig. 8). (Figs. 7 and 12). However, the outcrop example dif-
Deformation at the base and top of domino blocks fers from McClay and Ellis' model in two respects.
is concentrated along the edges of sandstone fault First, restoration of cross-sections from the outcrop
blocks where intervening shale layers are thin. In suggests that the shape of the rollover evolved partly
contrast, thick adjacent shale layers accommodate by propagation of synthetic faults into the footwall
potential volume problems created by domino rota- block. Hanging wall deformation that results from
tion by concentrating small-scale faulting. Similarly, this process has not yet been modelled, since ex-
models made entirely of brittle layers of sand or periments have been carried out with undeformable
clay show penetrative deformation of the base of footwall ramps. Second, in physical models defor-
the block (McClay and Ellis, 1987; Vendeville et al., mation of the hanging wall was solely caused by
1987; Kautz and Sclater, 1988), whereas fault blocks sliding along the major listric fault plane and did
undergo little internal deformation in models under- not involve superimposed extension of the hanging
lain by deformable ductile putty (Brun et al., 1985; wall. Hence, no faults occurred away from the main
Vendeville, 1987). fault beyond the crestal graben (McClay and Ellis,
In outcrops, some faults have attitudes that differ 1987). Instead, most listric normal faults in the Lin-
either in amount or sense of dip from that which coln Ranch basin display numerous faults that affect
dominates within the array (Fig. 15). Because fault the hanging wall away from the main fault and the
dip, rotation and stretch are related, changes in fault rollover crest (Fig. 7). This suggests that a signifi-
dip within domino domains imply local changes in cant component of stretching is superimposed on the
stretch along the basal detachment and local gaps rollover deformation (Fig. 16).
and overlaps at the base and tops of tilt blocks.
In the outcrop, rigid layers above and below tilt Implications for interpretation of extended
blocks prevent any significant gap or overlap between terrains
layers. Instead, fault blocks bounded by antithetic
or abnormally steep or gently dipping faults are Excellent exposures that illustrate the details of
internally deformed by closely spaced small faults. fault-block shape and their kinematic interpretation
240 S.E. Laubach, B. C. Vendeville and SJ. Reynolds

Fig. 16. Rollover profiles without (B) and with (C) superimposed extension of the hanging wall. (A) Shows configuration prior to fault
movement. Accommodation is by vertical shear.

based on experimental results lead to the following mic detection (i.e., tens of meters or less) but could
remarks. influence reservoir compartmentalization and het-
Analysis of three-dimensional fault geometries erogeneity.
shows that the upper plate of the detachment fault
is compartmentalized into several mesoscale units Acknowledgments
on the scale of tens of meters that accommodate
the bulk extension. Within most units deformation We thank M.P.A. Jackson, D.D. Schultz-Ela and
remains plane-strain extension accommodated along J.E. Spencer for discussions and reviews, and Bjorn
four sets of faults that intersect at low angles. Units Torudbakken and Rob Young for reviews. This paper
are separated from each other either by oblique- is based upon work supported by the Texas Advanced
slip faults with curved traces or by units that have Research Program under Grant No. 2174. Publica-
undergone biaxial stretching. tion authorized by the Director, Bureau of Economic
The outcrop example illustrates the significant Geology, The University of Texas at Austin.
role of curved faults in the development of over-
all fault shape and kinematics. Some faults have an References
overall curvature that in detail consists of straight
Brix, M., Schwarz, H-.U. and Vollbrecht, A., 1985. Tektonische
segments and sharp bends (ramps and flats). Others Experimente als Beitrag zu Strukturanalysen im Ruhrkarbon.
are smoothly curved and either bound rigid blocks or Glucksauf-Forschungshefte, 46: 4-11.
significantly deformed blocks. Convex-up (steepen- Brun, J.P., Choukroune, P. and Faugere, E., 1985. Les
ing downward or negative listric) faults are abundant discontinuity significatives de l'amincissement crustal: ap-
plication aux marges passives. Bull. Soc. Geol. Fr., 8: 139-
and play a significant role in the overall geome-
144.
try of the faulted terrain. Such faults may be more Cloos, E., 1955. Experimental analysis of fracture patterns. Geol.
important in extended terrains than is currently ap- Soc. Am. Bull., 66: 241-256.
preciated. When they bound rigid blocks, negative Cloos, E., 1968. Experimental analysis of Gulf Coast fracture
listric faults can counteract tilting caused by listric pattern. Am. Assoc. Pet. Geol. Bull., 52: 420-444.
Donath, F.A., 1962. Analysis of basin-range structure, south-
faults or domino-style rotation. Hence they cause
central Oregon. Geol. Soc. Am. Bull., 73: 1-16.
underestimation of the amount of extension deduced Faugere, E., Brun, J.P. and Van Den Driessche, J., 1986.
from block tilt. Underestimation of extension may Bassins asymetriques en extension pure et en decrochement:
also result from internal deformation at the base and Modeles experimentaux. Bull. Centres Recherche Exploration
top of rotating fault blocks as shown in the models Production Elf-Aquitaine, 10: 13-21.
Freund, R. and Merzer, A.M., 1976. The formation of rift valleys
of Kautz and Sclater (1988). Similar problems in
and their zigzag fault patterns. Geol. Mag., 113: 561-568.
extension estimation and section restoration arise in Gibbs, A.D., 1983. Balanced cross-section construction from
stretched hanging wall blocks of major listric faults. seismic sections in areas of extensional tectonics. J. Struct.
Fault-block shapes reflect crisscrossing and su- Geol., 5: 153-160.
perposed faults having a range of sizes. Complex Hoeppener, R., Kalthoff, E., Schrader, P., 1969. Zur phys-
icalischen Tektonik Bruechbildung bei verschieden affinen
within-block fault geometry, at medium scale, can Deformation im Experiment. Geol. Rundsch., 59: 179-193.
result from simple kinematics. Such complex fault Horsfield, W.T., 1980. Contemporaneous movement along cross-
geometries are likely to be below resolution of seis- ing conjugate normal faults. J. Struct. Geol., 2: 305-310.
Patterns in the development of extensional fault-block shapes 241

Kautz, S.A. and Sclater, J.G., 1988. Internal deformation in clay Vendeville, B.C., 1987. Champs de failles et tectonique en exten-
models of extension by block faulting. Tectonics, 7: 823-832. sion: modelisation experimentale. Rennes, France, Memoires
Mandl, G., 1987. Tectonic deformation by rotating parallel faults du Centre Armoricain d'Etudes Structurales des Socles, 15,
the "bookshelf" mechanism. Tectonophysics, 141: 316-327. 392 pp.
Mandl, G. (Editor), 1988. The Mechanics of Tectonic Faulting. Vendeville, B.C., 1988. Fault geometries and location in sed-
Elsevier, Amsterdam, 407 pp. imentary cover during basement-controlled deformation: an
McClay, K.R. and Ellis, P.G., 1987. Analogue models of exten- experimental investigation. Am. Assoc. Pet. Geol., Bull., 72:
sional fault geometries. In: M.P. Coward, J.F. Dewey and RL. 1029 (abstract).
Hancock (Editors), Continental Extensional Tectonics, Geol. Vendeville, B.C., 1991. Mechanisms generating normal fault
Soc. London, Spec. Publ., 28: 109-127. curvature: a review illustrated by physical models. In: A.
McGill, G.E. and Stromquist, A.W., 1979. The graben of Canyon- Roberts, G. Yielding and B. Freeman (Editors), Geometry of
lands National Park, Utah: geometry, mechanics and kine- Normal Faults. Geol. Soc. London, Spec. Publ., 56: 241-249.
matics. J. Geophys. Res., 84: 4547-4563. Vendeville, B.C., Cobbold, PR., Davy, P., Brun, J.P. and
Morton, W.H. and Black, R., 1975. Crustal attenuation in Afar. Choukroune, P., 1987. Physical models of extensional tec-
In: Pilger, A. and Rosier, A. (Editors), Afar Depression of tonics at various scales. In: M.P. Coward, J.F. Dewey. and PL.
Ethiopia, Inter-Union Comm. Geodyn. Sei. Rep., No. 14, pp. Hancock (Editors), Continental Extensional Tectonics. Geol.
55-65. Soc. London, Spec. Publ., 28: 95-109.
Oertel, G., 1965. The mechanism of faulting in clay experiments. Walsh, J.J. and Watterson, J., 1988. Dips of normal faults in
Tectonophysics, 2: 343-393. British Coal Measures and other sedimentary sequences. J.
Proffett, J.M., Jr., 1977. Cenozoic geology of the Yerrington Geol. Soc. London, 145: 859-875.
district, Nevada and its implications for the nature and origin Wernicke, B.P., 1981. Low-angle normal faults in the Basin and
of Basin and Range faulting. Geol. Soc. Am. Bull., 88: 247- Range province: nappe tectonics in an extending orogen.
266. Nature, 291: 645-648.
Reches, Z. 1978. Analysis of faulting in three-dimensional strain Wernicke, B.P. and Burchfiel, B.C., 1982. Modes of extensional
field. Tectonophysics, 47: 109-129. tectonics. J. Struct. Geol., 4: 105-115.
Serra, S. and Nelson, R.A., 1988. Clay modelling of rift asym- White, N.J., Jackson, J.A. and McKenzie, D.P., 1986. The
metry and associated structures. Tectonophysics, 153: 307- relationship between the geometry of normal faults and that
312. of the sedimentary layers in their hanging walls. J. Struct.
Shelton, J.W., 1984. Listric normal faults: an illustrated summary. Geol. 8: 897-910.
Am. Assoc. Pet. Geol., Bull., 68: 801-815. Woods, E.P, 1988. Extensional structures of the Jabiru terrace,
Spencer, J.E. and Reynolds, S.J., 1989. Geology and min- Vulcan sub- basin. In: P.G. Purcell and R.R. Purcell (Editors),
eral resources of the Buckskin and Rawhide Moun- The North West Shelf, Australia. Proceedings, North West
tains, West-Central Arizona. Ariz. Geol. Surv. Bull., 198, Shelf Symposium, Petroleum Exploration Society Australia,
279 pp. Perth, W.A., pp. 311-330.

S.E. LAUBACH Bureau of Economic Geology, University of Texas at Austin, University Station BoxX, Austin, TX 78713-7508, U.S.A.
B.C. VENDEVILLE Bureau of Economic Geology, University of Texas at Austin, Austin, TX 78713-7508, U.S.A.
S.J. REYNOLDS Arizona Geological Survey, Tucson, AZ 85719, U.S.A.
 243

Complex hanging-wall deformation above an extensional

detachment example: Gulifaks Field, northern North Sea
A.G. Koestler, A.G. Milnes and A. Storli

Analysis of the geometry and kinematics of the Gulifaks Fault System, based on the Statoil interpretation of the 1985 three-
dimensional seismic survey, itself constrained by abundant well data, shows that the overall structure can best be understood
in terms of hanging-wall collapse in a pull-apart zone above a sub-Triassic extensional detachment. The "domino"-style faulting
in the Western Gulifaks reservoir horizons is part of this pull-apart zone and shows an unusually well-constrained geometry,
which, independent of the proposed kinematic model, allows important conclusions to be drawn. In particular, the shape of the
irregular fault surfaces and the constancy of bedding can only be understood in terms of bulk "plane strain", with a vertical
E - W deformation plane containing the E-plunging slip vector. E - W depth-converted cross-sections can therefore be restored
using simple balancing techniques, taking into account the probable original dip of the faults (about 70) and bedding planes
(sub-horizontal) and the probable effect of syn- to post-faulting compaction (shown to be of minor importance). The restored
sections show that horizontal extension in the pull-apart zone amounted to about 50%, and that considerable homogeneous
shear strain must have affected the reservoir rocks in the fault blocks, a fact which may be significant for reservoir behaviour
and modelling.

Introduction tonic interpretations have been difficult, however,

because of the discontinuity of the main reflectors
The Gulifaks oil-field lies close to the main axis and the uncertainty of fault correlation between
of the Viking Graben in the northern North Sea different levels. Nevertheless, there is general agree-
(Fig. 1). The geology of the field is well-known ment that the Jurassic reservoir units show complex
because of high-quality three-dimensional seismic normal faulting, related to extensional deformation
surveys and extensive borehole control (Saeland and in latest Jurassic and early Cretaceous times. Al-
Simpson, 1981; 1982; Erichsen et al., 1987; Fossen, though there are some indications that the structure
1989; Petterson et al, 1990). Structural and tec- may have been affected by later non-extensional

100 km

Fig. 1. Tectonic sketch map of the northern North Sea, showing the location of the Gulifaks Field (based on Faerseth and Pederstad,
1988).

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 243-251. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
244 A.G. Koestler, A.G. Milnes and A. Storli

see Petterson et al., 1990; Fig. 6). The quality of

the seismic data can be judged by consulting the
published lines (uninterpreted and interpreted) in
Fossen (1989) and Petterson et al. (1990). In the
present work, non-reflective faults are only con-
sidered as hard data where they can be fixed in
orientation by alignment of three or more reflector
terminations. Our aim has not been to reinterpret the
seismic data itself, but rather to identify, and to carry
out an evaluation of, those structural features of a
well-established seismic interpretation which can be
regarded as definitive. The quality of the geophysical
and well data allows an unusually clear distinction
to be made between observation and interpretation,
comparable to onshore geological mapping.

Reservoir level, Western Gullfaks

As documented by several prominent Jurassic re-
flectors (top Statfjord, top Dunlin, top Rannoch, top
Brent), the seismic structure of the Gullfaks reservoir
horizons west of 215/ (e.g., the area surrounding A
Fig. 2. Structure map of the Gullfaks Field at top Dunlin level,
showing the position of Figs. 3 and 6 and of the three production
and B platforms, Fig. 2) is dominated by a prominent
platforms ^4, B, C (based on Fossen, 1989). system of roughly N-S-trending normal faults, with
displacements in the range of a few hundred metres,
all with downthrow to the east. In horizontal sections
movements (Fossen, 1989), such modifications seem (Fig. 3), the fault traces are sinuous and branching
to have been slight compared to the overall exten- (i.e., the fault surfaces are non-planar/non-parallel).
sional strain which characterizes the internal struc- They delineate lozenge-shaped fault blocks with N-S
ture of the Gullfaks field. In this paper, we focus long axes. Geometric analysis of the individual fault
on the latter, the main deformation of the reser- surfaces shows them to be strongly "cylindrical" in
voir units and, hence, the main determinant of three-dimensions, i.e., the poles to the fault surfaces
internal reservoir structure. As we shall see, our measured at various positions along the faults show
analysis suggests that it represents a pull-apart and a great circle distribution in stereographic projection
collapse zone above a major sub-horizontal exten- (Fig. 4). This "cylindricity" is repeated throughout
sional detachment, whose regional significance is at the western part of the field, giving a mean "axis of
present unknown. In Gullfaks, this complex zone cylindricity" plunging 30-35 due east (Fig. 5). At
of hanging-wall deformation can be subdivided into the same time, the statistical averages of bedding
two contrasting structural regimes, Western Gullfaks orientation within the various fault lozenges remain
and Eastern Gullfaks, with a gradational bound- constant and tightly confined, with an overall mean
ary approximately along longitude 215/ (Fig. 2). In dip of 10-15 west (Figs. 4 and 5). This remarkable
the following, we first summarize the basic seismic geometry puts tight constraints on the kinematics of
structure of these two regimes, before presenting fault movement, which can only have taken place
an analysis of the geometrical relationships and a by slip parallel to the mean "axis of cylindricity".
tentative kinematic model. Appreciable slip in any other direction would have
caused complex and heterogeneously distributed in-
ternal adjustments within the lozenges, which the
Seismic structure
constancy of bedding belies. Hence, the Western
The data base for our analysis is the Statoil in- Gullfaks fault system at reservoir level corresponds
terpretation of the 1985 three-dimensional seismic to the "domino" fault model (e.g., Wernicke and
survey, in which all the main reflectors have been Burchfiel, 1982; Axen, 1988; Barr, 1987), although
identified and mapped on the basis of extensive well the word "domino" in this context is somewhat mis-
correlations. The data was available as interpreted leading, since we are obviously dealing with very
seismic time-sections (work station, paper copies) irregularly shaped fault blocks. In three-dimensions,
and as depth-converted structure contour maps at the Western Gullfaks structure is an almost
several different stratigraphic levels (scale 1 :25 000, ideal example of bulk "plane strain", with a ver-
Complex hanging-wall deformation above an extensional detachment example: Gullfaks Field, northern North Sea 245

20' 215
615

WESTERN GULLFAKS
HORIZONTAL SECTION
AT 2100 M DEPTH

CRETACEOUS

TOP BRENT

TOP RANNOCH

TOP DUNLIN

LARGE FAULT (s > 150m)

SMALL FAULT (s < 150m)

Fig. 3. Horizontal section of the western part of the Gullfaks field at a depth of 2100 m (for location see Fig. 2), based on the Statoil
interpretation of the 1985 three-dimensional seismic survey (ST 8511). The section is constructed from 1:25 000 depth-converted
structure contour maps for top Statfjord, top Dunlin, top Rannoch and top Brent, with stratigraphic control from about 40 wells. Ticks
on the fault traces indicate the downthrown side and also mark the positions at which the fault orientation and displacement were
measured (for stereographic projection see Fig. 5).

tical E-W-trending deformation plane. This means Reservoir level. Eastern Gullfaks
that all movement has taken place within the defor- The seismic structure of the Gullfaks reservoirs
mation plane (fault slip, internal deformation of fault east of about 215/ (e.g., the area surrounding C plat-
blocks) and there has been no dimensional changes form, Fig. 2) does not show such neat geometrical re-
in the N-S direction. Whether by chance or by lationships. Figure 6 shows the reflector reconstruc-
design, the lines of the 1985 three-dimensional seis- tions from three lines of the 1985 three-dimensional
mic survey parallel precisely the deformation plane, survey, taken from the Statoil seismic interpretation
hence the ideally "domino" aspect of the faults in the as published in Fossen (1989). On such sections,
western part of the Gullfaks Field on seismic sections only identified reflector traces and those segments
(Fig. 6). of fault traces which can be "pinned" by lines of
reflector terminations are relevant to the present
246 A.G. Koestler, A.G. Milnes and A. Storli

20'
GULLFAKS
WESTERN FAULT

Fig. 4. Example of a structure contour and true slip variation map of a large fault from the Gullfaks field (western fault, see Fig. 2).
The inset is a stereogram of the fault surface, showing poles to the fault plane at the points of slip determination. The partial great
circle distribution of the fault plane poles results from the "cylindricity" of the fault surface, and the axis of the girdle (X) marks the
true slip direction (for explanation see text).

study. The "domino"-style faulting of Western Gull- fault with a downthrow of a few hundred metres
faks is seen to change eastward into an irregular to the west, i.e., antithetic to the Western Gullfaks
association of both eastward- and westward-dipping faults (Fig. 6). This fault marks the western margin of
normal faults, possibly with some small reverse faults the so-called Gullfaks "horst" (Saeland and Simpson,
and flexures. Some of these complications may be 1982; Fossen, 1989), a deeply eroded high along the
due to minor superimposed non-extensional defor- eastern edge of the field. To the east, this high is
mation (transpression, cf. Fossen, 1989), but they are bounded by the Main Boundary Fault, a major struc-
also of the type encountered in any area of exten- ture, with a throw of 1-2 km, which can be traced
sional deformation where the boundary conditions on deep seismic profiles far down into the crust (cf.
do not allow compatible movements (e.g., Coffield Beach, 1986; Gibbs, 1987). The seismic structure of
and Schamel, 1989). The most constant structural the Eastern Gullfaks high differs from the rest of the
feature of Eastern Gullfaks is a large subvertical field in that it shows only small, widely spaced dis-
Complex hanging-wall deformation above an extensional detachment example: Gullfaks Field, northern North Sea 247

Kinematic model

The seismic structure described above (see also

/ \ Petterson et al., 1990) suggests a kinematic inter-
pretation in terms of hanging-wall collapse, in a
W." / GIRDLE OF FAULT POLES
\\
region of pull-apart above an extensional detach-
" " \ ment (Fig. 7). The interpretation is based on the
*/ \ remarkable similarities between different parts of
/ '
the Gullfaks structure and several published geo-
metrical and experimental models of similar fault
# /. DEFORMATION PLANE SLIP DIRECTION
systems (e.g., Wernicke and Burchfiel, 1982; Ram-
'-{4& , > lg)= say and Huber, 1987; McClay and Ellis, 1987; Barr,
J # .rt 4n-50 IPv 3 35
- ' 1987; Vendeville et al., 1987; Axen, 1988; Groshong,
Ml 1988; Kautz and Sclater, 1988). However, the single
# T% , BEDDING POLE
* * , VARIATION FIELD
ramp fault of most detachment models, which leads
*\
% \# /
to the asymmetrical structure of the trailing edge
of the hanging-wall ("reverse drag", "roll-over" an-
/
ticline, half-graben, etc.) is replaced in our model
/
by two separate faults (ramp fault and trailing-edge
\ LOWE *RL HEMISPHER
EQUA ARE A NET
E
/ fault) bounding a broad region of pull-apart which
. - \ N = 126 / collapses gravitationally to fill the space, as the de-
Fig. 5. Compilation stereogram
* \ for fault plane measurements
/
in the western part of the \ / 3). The
Gullfaks field (area of Fig. tachment becomes active. In the case of Gullfaks,
\ \\
great circle distribution \ /
\ fault plane poles/strongly suggest
of the we suggest that this collapse took the form of exten-
a plane strain movement picture for the western Gullfaks fault sion and thinning of the more competent formations
system, with a sub-vertical E-W deformation plane and slip
direction marked by the girdle axis (plunge 30-35 towards E).
by "domino"-style faulting, possibly accompanied by
concomitant ductile deformation in adjacent less
competent units. The constant eastward dip of the
faults in Western Gullfaks suggests that they were
continuities in the reflector pattern, mostly normal originally a system of small faults within the pull-
faults both synthetic and antithetic to the fault along apart zone, synthetic to the main ramp fault (which
its western margin (Fig. 6). lies, on this interpretation, to the west of Gullfaks,
presumably between it and the Statfjord Field). The
Intra-Triassic level more complex relations in Eastern Gullfaks would
The prominent array of reflectors at intra-TYiassic be caused by the accommodation of this fault sys-
level (e.g., top Carnian, Fig. 6) are not as well con- tem to the antithetic trailing-edge fault, which is
strained as the Jurassic reservoir horizons described the western marginal fault of the Eastern Gullfaks
above (no well control), and the constructed fault high, as mentioned above. The high itself represents
orientations are less reliable (fewer reflector ter- the trailing edge of the upper "plate" above the
minations). Hence, data from this level cannot be detachment. Because of the apparent congruence in
analysed so definitively. Nevertheless, it is interesting tectonic style between the Jurassic reservoir horizons
to note that it can best be interpreted as a simi- and the intra-THassic horizons (see Fig. 6), we would
lar structure to that documented above, i.e., as a place the detachment at a sub-THassic level, possibly
"domino"-style fault system to the west and more at the basement/cover contact, which in this area is
complex relations of similar geometry to the east supposed to lie at a depth of 5-6 km (Ediriweera,
(as drawn in Fig. 6). The apparent lack of continu- 1988; see also Swallow, 1986).
ity between the faults at Jurassic and intra-TYiassic A somewhat different interpretation of the Gull-
levels may be structurally significant (e.g., implying faks structure in terms of detachment tectonics is
decoupling and ductile deformation in intervening given by Fossen (1989), although the main exten-
incompetent horizons) or it may be an artifact of sional deformation of the Gullfaks field was not the
processing and interpretation. Be that as it may, subject of his paper. On his interpretation, the de-
the congruence of structural style at the two levels tachment ramps downwards into the basement along
suggests that they both belong to the same tectonic the Main Boundary Fault (Fossen, 1989; Fig. 7).
regime. This point leads us to place the main detach- Our interpretation of the time relations, however,
ment of our kinematic model below the intra-Triassic suggests that a large part of the movement on the
reflector array (see below). Main Boundary Fault took place during the deposi-
248 A.G. Koestler, A.G. Milnes and A. Storli

WESTERN GULLFAKS I EASTERN GULLFAKS TWT

(sees)

Fig. 6. Reflector reconstructions for three profiles from the 1985 three-dimensional seismic survey, taken from the published seismic
sections in Fossen (1989). The indicated faults are not reflectors: they are only indicated where they can be reconstructed definitively
from the displacement of the stratigraphic reflectors. The reflectors S (top Statfjord) and D (top Dunlin) are controlled by abundant
well data, whereas reflector C (intra-Trias, presumably top Carnian) is uncontrolled. The profiles are not depth-converted and the
faults appear steeper than they actually are (cf. Fig. 8). Other symbols: K = base Cretaceous, P = near top Paleocene, MBF = Main
Boundary Fault.

tion of the lower Cretaceous formations in the basin tachment had ceased to be active, truncating it and
to the east, at the time when the Gullfaks high was its hanging-wall east of the trailing edge fault and
being uplifted and eroded (creating the base Creta- thus isolating the Eastern Gullfaks high (compare
ceous angular unconformity). There is stratigraphic Figs. 6 and 7). That extensional tectonics often take
evidence from the Gullfaks South Field, where the place in distinct phases, producing intersecting fault
post-Brent formations are better preserved, that ac- systems, is well known, both from the northern North
tivity on the Gullfaks Fault System had ceased by Sea (e.g., Badley et al., 1988; Gibbs, 1987) and from
uppermost Jurassic times (Heather formation un- other continental margins (e.g., Mutter and Larson,
faulted). Also, the sharp truncation of several faults 1989; Williamson et al., 1989). More data from sur-
of the Gullfaks system by the Main Boundary Fault rounding areas may prove that the Gullfaks structure
along the southern boundary of the Gullfaks field is a is of a similar type.
further indication of temporal separation. Hence, we As indicated in the last section, the fault and
feel that it is likely that most movement on the Main reflector geometry of Western Gullfaks (Figs. 3-5)
Boundary Fault took place after the sub-Gullfaks de- shows that almost ideal "plane strain" conditions
Complex hanging-wall deformation above an extensional detachment example: Gullfaks Field, northern North Sea 249

REGION OF PULL-APART
w
\ \ \ \ \ \ /
FUTURE MAIN RAMP FAULT ZONE OF FUTURE HANGING-WALL

Jbi HANGING-WALL COLLAPSE ' ("UPPER PLATE")

r - FUTURE TRAILING EDGE FAULT

a)
FOOT-WALL FUTURE EXTENSONAL DETACHMENT
("LOWER PLATE")

LEVEL OF LATER EROSION

DECOUPLING HORIZON

Fig. 7. Schematic illustration of the proposed kinematic model for the Gullfaks Field in terms of detachment tectonics (for explanation
see text).

must have reigned during the development of the by the lack of a discernible angular difference be-
Gullfaks Fault System. The deformation plane, con- tween the fault traces in Dunlin and the underlying
taining the E-plunging slip vector, was orientated E - and overlying sandy formations (Statfjord, Brent),
W, so that E-W depth-converted sections are ideally i.e., faulting took place after the reservoir horizons
suited to applying balancing techniques (cf. Gibbs, had reached a depth regime in which differential
1983). As an example, straightforward area balance compaction of sandstone and shale was small, (af-
has been applied to the Dunlin unit on part of a ter a few hundred metres of burial, Xiao and Suppe,
depth-converted E-W section by restoring the fault 1989). Hence, although compaction clearly continued
planes to an original dip of about 70 and the bed- during and after faulting, it can be assumed to have
ding planes to horizontal (Fig. 8). The original fault had only a minor effect on the observed geometrical
dip of 70 is an assumption based on field observa- relations.
tions and theoretical considerations (cf. Ramsay and Two interesting facts emerge from the restoration
Huber, 1987; Walsh and Watterson, 1988; Xiao and shown in Fig. 8. Firstly, it gives a quantitative esti-
Suppe, 1989). The other assumption in the procedure mate of the amount of stretching in the pull-apart
is that the cut-off trace length (distance between the zone. The horizontal extension in this cross-section
intersections of the top and bottom of the forma- (AT to X'Y\ Fig. 8) is approximately 50%, a value
tion and the fault traces) has remained constant. which is typical for the Western Gullfaks structure as
Physically, this means that most compaction is as- a whole. If the width of the pull-apart zone and the
sumed to have taken place before faulting occurred. depth to detachment were known, as may be possi-
The following arguments indicate that this is likely ble with increasing knowledge of surrounding areas,
to be a good approximation to the true situation. this extension estimate will allow the amount of slip
Porosity reduction in shales, which make up most on the detachment fault to be estimated. Secondly,
of the Dunlin Group, takes place mainly during the the restoration shows that the fault "blocks" them-
first 200-300 m of burial. Below this, compaction selves must have undergone a significant amount
occurs much more slowly, at about the same rate of internal deformation, since the bedding/fault an-
as in sandstone, and can be shown to alter the gle has changed from about 70 to 40-50 in the
dip angles of pre-compaction faults by only about course of the faulting (Fig. 8). Since compaction, as
17km (Xiao and Suppe, 1989). Since the Gullfaks shown above, is unlikely to have modified the ge-
faulting is certainly post-Brent in age, possibly even ometry to any significant extent, this indicates that
later, and since the Brent formation in Gullfaks is the Gullfaks reservoirs have undergone considerable
about 270 m thick (Petterson et al., 1990), the Dun- homogeneous shear strain, which must have been
lin shales can be assumed to have reached depths accommodated by adjustments at the level of in-
of 300 m before faulting started. This is supported dividual grains and/or by movement on small-scale
250 A.G. Koestler, A.G. Milnes and A. Storli

SEA FLOOR

TERTIARY
L_1km

.3km

1km
PRE-FALTING RESTORATION
.4km
HORIZONTAL DATUM

Fig. 8. Depth-converted E-W section through the reservoir horizons in part of Western Gullfaks (approximately along the line of
section C, Fig. 6). Inset: pre-faulting restoration assuming that the original fault trace dip was about 70 and the original bedding was
horizontal (for explanation see text).

(sub-seismic) distributed fractures. The combination collapse in a pull-apart zone above an extensional de-
of movement on distinct fault surfaces and homo- tachment. The detachment fault lies below the intra-
geneous block-internal "ductile" deformation is a THassic reflectors, possibly at the basement/cover
common feature of sand-box and clay models of contact (here at 5-6 km depth). The Eastern Gull-
"domino"-style faulting (McClay and Ellis, 1987; faks high ("horst") represents the trailing edge of the
Kautz and Sclater, 1988; Vendeville and Cobbold, upper plate: it is a high relative to the faulted units
1988), but as far as we are aware, this is the first to the west because it did not collapse, and it looks
time it has been proved on a field example, using like a "horst" because it was later truncated to the
site-specific data. east by the Main Boundary Fault, here interpreted as
mainly post-detachment in age.
Conclusions (3) The tightly constrained geometry of the West-
ern Gullfaks faults allows pre-faulting restoration of
Our analysis of the geometry and kinematics of depth-converted E-W cross-sections by simple area
the Gullfaks Fault System is based on the Statoil balancing. We show that most compaction took place
interpretation of the 1985 three-dimensional seismic before faulting and that compaction during and after
survey, itself constrained by abundant well data, and faulting is unlikely to have had a major effect on the
leads to the following conclusions. structural relations. Comparison of the restored and
(1) Fault surface geometry and the constancy of present cross-sections allows a quantitative estimate
bedding orientation show that the overall defor- of horizontal extension in the pull-apart zone (50%).
mation of the "domino"-style faulting in Western It also shows that the Gullfaks reservoirs must have
Gullfaks is "plane strain", with a vertical E-W defor- undergone a considerable amount of internal defor-
mation plane containing the E-plunging slip vector. mation (homogeneous shear strain, accommodated
(2) This faulting is part of a fault system which by grain-scale adjustmen s and/or movement on sub-
can be interpreted in terms of complex hanging-wall seismic discontinuities) curing faulting. This may be
Complex hanging-wall deformation above an extensional detachment example: Gullfaks Field, northern North Sea

of significance for reservoir behaviour and produc- seismic sections in areas of extensional tectonics. J. Struct.
tion performance and needs to be taken into account Geol., 5: 153-160.
Gibbs, A.D., 1987. Deep seismic profiles in the northern North
in reservoir modelling.
Sea. In: J. Brooks and K.W. Glennie, (Editors), Petroleum
Geology of North-West Europe, Vol. 2. Graham and Trotman,
Acknowledgements London, pp. 1025-1028.
Groshong, R.H., 1988. Half-graben structures: balanced models
The above ideas are based on the interpretation of of extensional fault-bend folds. Geol. Soc. Am. Bull., 101:
96-105.
the Gullfaks ST 8511 survey by J.O. Henden and H. Kautz, S.A. and Sclater, J.G., 1988. Internal deformation in clay
Fossen. We would like to thank them for generously models of extension by block faulting. Tectonics, 7: 823-832.
allowing us to use the results of their work, and for McClay, K.R. and Ellis, P.G., 1987. Geometries of extensional
many discussions. Our thanks are also due to an fault systems developed in model experiments. Geology, 15:
341-344.
anonymous reviewer and Rob Young, for stimulating
Mutter, J.C. and Larson, R.L., 1989. Extension of the Ex-
criticism which resulted in significant improvements
mouth Plateau, offshore northwestern Australia: deep seismic
to the manuscript, and to Statoil a.s., for publication reflection/refraction evidence for simple and pure shear mech-
permission. We are, however, solely responsible for anisms. Geology, 17: 15-18.
the ideas presented and opinions expressed. Petterson, O., Storli, A., Ljosland, E. and Massie, I., 1990. The
Gullfaks Field-Geology and reservoir development. In: A.T.
Butler, E. Berg, O. Hjelmeland, J. Kleppe, O. Torsaeter and
References J.O. Aasen (Editors), North Sea Oil and Gas Reservoirs
II. Graham and Trotman, London, pp. 67-90.
Axen, G.J., 1988. The geometry of planar domino-style normal Ramsay, J.G. and Huber, M.I. (Editors), 1987. The Techniques
faults above a dipping basal detachment. J. Struct. Geol., 10: of Modern Structural Geology, Vol. 2. Folds and Fractures.
405-411. Academic Press, London, 700 pp.
Badley, M.E., Price, J.D., Rambech Dahl, C. and Adgestein, T, Saeland, G.T and Simpson, G.S., 1981. 3-D seismic data help
1988. The structural evolution of the northern Viking Graben to delineate subconformity trap in Norwegian North Sea. Oil
and its bearing upon extensional modes of basin formation. J. Gas J., 79(45): 381-398.
Geol. Soc. London, 145: 455-472. Saeland, G.T. and Simpson, G.S., 1982. Interpretation of 3-D
Barr, D., 1987. Structural/stratigraphic models for extensional data in delineating a subconformity trap in block 34/10,
basins of half-graben type. J. Struct. Geol., 9: 491-500. Norwegian North Sea. Am. Assoc. Pet. Geol., Mem. 32: 207-
Beach, A., 1986. A deep seismic reflection profile across the 216.
northern North Sea. Nature, 323: 53-55. Swallow, J.L., 1986. The seismic expression of a low angle
Burchfiel, B.C., Wernicke, B., Willemin, J.H., Axen, G.J. and detachment (sole fault) from the Beryl Embayment, central
Cameron, C.S., 1982. A new type of decollement thrusting. Viking Graben. Scott. J. Geol., 22: 315-324.
Nature, 300: 513-515. Vendeville, B. and Cobbold, PR., 1988. How normal faulting
Coffield, D.Q. and Schamel, S., 1989. Surface expression of an and sedimentation interact to produce listric fault profiles
accommodation zone within the Gulf of Suez rift, Egypt. and stratigraphic wedges. J. Struct. Geol., 10: 649-659.
Geology, 17: 76-79. Vendeville, B., Cobbold, PR., Davy, P., Brun, J.P. and
Ediriweera, K.K., 1988. A geological and geophysical investiga- Choukroune, P., 1987. Physical models of extensional tec-
tion of the Viking Graben area, North Sea. Dr. Ing. Thesis, tonics at various scales. Geol. Soc. London, Spec. Publ., 28:
NTH, Trondheim. 95-107.
Erichsen, T, Helle, M., Henden, J. and Rognebakke, A., 1987. Walsh, J.J. and Watterson, J., 1988. Dips of normal faults in
Gullfaks. In: A.M. Spencer et al. (Editors) Geology of British Coal Measures and other sedimentary sequences. J.
the Norwegian Oil and Gas Fields. Graham and Trotman, Geol. Soc. London, 145: 859-873.
London, pp. 273-286 Wernicke, B. and Burchfiel, B.C., 1982. Modes of extensional
Faerseth, R.B. and Pederstad,. K., 1988. Regional sedimentology tectonics. J. Struct. Geol., 4: 105-115.
and petroleum geology of marine, late Bathonian-Valanginian Williamson, P.E., Collins, C.D.N. and Falvey, D.A., 1989. Crustal
sandstone in the North Sea. Mar. Pet. Geol., 5: 17-33. structure and formation of a magnetic quiet zone in the
Fossen, H., 1989. Indication of transpressional tectonics in the Australian southern margin. Mar. Pet. Geol., 6: 221-231.
Gullfaks oilfield, northern North Sea. Mar. Pet. Geol., 6: Xiao, H.-B. and Suppe, J., 1989. Role of compaction in listric
22-30. shape of growth normal faults. Am. Assoc. Pet. Geol., Bull.,
Gibbs, A.D., 1983. Balanced cross-section construction from 73(6): 777-786.

Paper submitted October 1989.

A.G. KOESTLER Geo-Recon A IS, Bernhard Herres vei 3, 0376 Oslo 3, Norway
A.G. MILNES Department of Earth Sciences, ETH, Zrich, Switzerland
Present address: Geological Institute, Department A, University of Bergen, Allegt 41, N-5007 Bergen, Norway
A. STORLI Statoil AIS, Bergen, Norway
 253

Physical changes in fault blocks caused by rotation and

their implications for hydrocarbon accumulations
J.E. Iliffe, I. Lerche and K. Nakayama

Physical changes that occur when a fault block is rotated in the Earth's gravitational field are assessed using a computer
model. Parameters such as lithology, height, width, angle of tilt and fault dip are varied. Results describe how and where the
physical properties of the block are changed by rotation, e.g., seismic reflection coefficients, geothermal gradients and bulk
density of the block.
Fault blocks preferentially compact on their downthrown side, up to and including, the crest. Reflection coefficients in layers
in the uppermost 1 km are moderated at particular depths in the compressed field of the block. The footwall area of the block
acts as a conductive rod to heat; the juxtaposed low conductive areas making the latter a more attractive hydrocarbon generating
area. Dewatering fluids are predicted to flow from the hotter compacted hanging wall to the cooler uncompressed footwall,
which may contribute to shale diapirism seen in muddy deltaic sequences and improve the hydrocarbon potential of the footwall
area. Also, calculations of extension estimates may be erroneous, due to either shortening of the surface layer or changes in
deeper bed lengths by compactional folding. The stress configuration of the rotated compacted block suggests a plane of failure
may occur in lieu of compactional strain. The angle this secondary fault makes to the original fault is consistent with field
observations.

Introduction Model description

The importance of block faulted extensional basins The model consists of three unrotated identical
as hydrocarbon producing areas is exemplified in the blocks of width, W and vertical height, 77z, bounded
North Sea, Gulf of Suez and Renconcavo Basin. by faults with an angle, r, with layers of equal
The common element in extensional regimes is the thickness () (Fig. la). All these parameters may be
existence of rotated fault blocks. Geohistory mod- varied in experiments.
els exist to approximate the change in the physical The characteristics of different lithology blocks
properties of sediments through time in these areas may be broadly recreated by varying the scaling
(e.g., Guidish et al., 1984), but analysis of what hap- constant (i) and surface porosity (0) in the poros-
pens to the intrinsic properties of the fault blocks ity/depth function of Magara (1976) (Fig. 2).
themselves due to just the rotation i.e., without
syn-sedimentary deposition, needs to be examined. () = 0(^) (1)
This paper assesses the physical changes which oc-
cur in fault blocks, tilted in the Earth's gravitational
where () is the calculated porosity at depth z.
field, by means of a forward mathematical model.
The model calculates changes in porosity, thermal
Compaction
conductivity, reflection coefficient, thicknesses and
The most important change in physical properties
lengths, overall block shape and internal stress field
of the blocks due to rotation is of porosity due to
caused by the rotation, for the different regions of
compaction. The compaction effect is determined by
the block. Forward models have several advantages
solving the differential equations relating the poros-
over field analyses in that the designer knows exactly
ity depth function (eqn. 1) to the overburden in terms
what all the parameters of the system are and how
of the original porosity (), original frame weight
they are manipulated. A vast array of variability of
Wt(x',') and the resultant frame weight Wt(x,z):
situations can be tested and the results analyzed to
gain a better understanding of the factors influencing _ <t,(z')\Wt(x,z)-Wt(x',z')]
a wide range of variations. [W(z')2W {l)

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 253-267. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
254 J.E. Iliffe, I. Lerche and K Nakayama

POROSITY (/)

^ 30 40 50

Fig. 2. Diagram comparing porosity depth curves of sandstone

Fig. 1. (a) Diagram showing parameters used to define the pre- and shale using values of 0 and I from Sclater and Christie
rotated blocks. W = block width; Th = block height; z l = strata (1980) in eqn. 1 of Magara (1976).
thickness; r = fault angle, (b) Diagram illustrating the method
of post-rotation sampling of depth (z) in increments of Ax, and
parameters used to describe the block in the model, (c) Model Geothermal effects
output illustrating the results of post-rotation compaction on When the porosity of a region within the block
shale blocks tilted 30. The dashed line with crosses denotes the
is decreased as a result of compaction, the thermal
pre-compacted block outline, the solid line denotes the resultant
compacted block. conductivity of that region is increased and may be
calculated using the relationship

where is the amount of compaction, pm is the

=~{){) (3)
fully compacted rock density and g is the acceleration where KR is the whole rock thermal conductivity at
due to gravity. depth z, Km is the thermal conductivity of the matrix
The pre-rotation total frame weight [(',')] and K{ is the fluid thermal conductivity (Lewis and
acting along each stratal boundary is calculated by Rose, 1970). An estimate of the temperature field
summing the weight between the top of the block within the block is then made using Fourier's law of
and the depth to the stratal boundary (Fig. la). The heat conduction:
new (post-rotation) total frame weight [Wif(x,z)] is
assessed at each point along a stratal boundary by V (KVT) = 0 (4)
summing the weights calculated for a set of depth
values measured perpendicular to the strata over with q = (/) on the basement and:
the final vertical depth (Fig. lb). If the new weight
[^(,)] is less than, or equal to, the old weight (5)
[Wt(x,)z,)]9 no compaction takes place. Compaction dz~ K
occurs when Wt(x,z) is greater than Wt(x','). where q is heat flux through the basement (assumed
If the blocks are assumed to have no horizontal constant); K is thermal conductivity; and dT/dz is
strength, compaction is limited to the vertical plane. the vertical temperature gradient. In order to cor-
The values for each z (vertical) point at each x rect for unrealistic topographic effects, in the one-
(horizontal) location are summed cumulatively from dimensional case we take the area between the tops
the base layer upwards. The resulting displacement of the block and a theoretical horizontal surface
() is now subtracted from its original vertical (e.g., seafloor with a temperature of 6C) (A-B in
position to yield the compaction corrected section Fig. 3a) to be "filled" with a weightless material with
(Fig. lc). the same geothermal properties as the sediments in
the blocks. With this boundary condition, the tern-
Physical changes in fault blocks caused by rotation and their implications for hydrocarbon accumulations 255

A Surface

c)

Fig. 3. (a) A column of sediment A, B, C, D is considered as units of rock of homogenous properties. The block is composed of
shale, 4 km high, 4 km wide with a fault angle of 80, rotated 30. Each layer is 500 m thick, (b) The average porosity of each layer
is calculated, yielding an average whole rock thermal conductivity and consequently an average geothermal gradient. The temperature
profile with depth is then calculated from an input surface temperature in a cumulative fashion, (c) Resulting pattern of isotherms due
to rotation and compaction. Note that regions where isotherms are close together correspond to upper sections of the block being
buried, whereas those areas with isotherms more widely spaced correlate with the exhumed deeper parts of blocks with lower porosity.
The blocks are the same as those in (a).

perature at each layer surface may be calculated as Reflection coefficients

follows. The calculation of a reflection coefficient is usually
The porosities within each layer are calculated undertaken for the boundary between two layers of
from eqn. 1 and the layer's average porosity found. different density and seismic velocity. In the present
Next, the average thermal conductivity of the layer form of the model only homogeneous blocks of rock
is calculated from eqn. 3. The average geothermal whose density varies as a function of the porosity
gradient (dT/dz) of each layer is calculated from eqn. of the rock are considered (Wyllie et al., 1956), in
5, using an assumed value for heatflow. The temper- accord with the porosity-depth function (eqn. 1). We,
ature field is now determined from the horizontal therefore, calculate the change in reflection coeffi-
surface down, by adding the product of the thickness cients of rotated layers relative to their unrotated
of the (compacted) layer and its geothermal gradient state.
(Fig. 3b). Isotherm contours are then drawn by linear To calculate the reflection coefficient, a relation-
interpolation (see Fig. 3c). ship of velocity with depth is required or, specifically,
Figure 3c shows that isotherms towards the tops change of Lame constants with depth. These may be
of the blocks are closer together than those which estimated from the formula:
are deeper, a result of the relationship of porosity to Xz + 2 = ( + 20) + [
geothermal gradient. The temperatures in areas of max
high porosity change with depth more rapidly than + 2 - (0 + 20)] tanh(z/L) (6)
those in areas of low porosity. where and are the Lame constants at depth,
256 I.E. Iliffe, I. Lerche and K. Nakayama

denoted by subscripts "o" surface, "z" at depth z; reflection coefficients and stress configurations are
and "max", maximum Lame constants, and L is the investigated by measuring changes in bed lengths,
scaling depth over which the Lame constants vary bed thicknesses, angles and overall shape of the
from surface to deep values. output relative to the input.
From an estimate of the maximum (VmdiX) and In the tests performed, two lithologic types were
minimum seismic velocities (VQ) of a shale as, e.g., considered, those of pure shale and of homogeneous
6000 and 2000 m/s, respectively, it is possible to sandstone, the values of t and 0 being taken from
construct a formula for the velocity at any depth, z, Sclater and Christie (1980; Fig. 2).
of the type constructed by Wyllie et al. (1956) (i.e., The size effect of the blocks, specifically the ratio
()2 = [() + 2()]/()). of the height of the block to its width, was tested
The reflection coefficient is calculated from: using three basic types. The long flat "warehouse
type" block (Fig. 4) was compared with a perfectly
ldlnVp
RC (7) square block (Fig. 5) and with a tall, thin "skyscraper
~2 dz
type" block (Fig. 6).
i.e., As the lithology and block dimensions were varied
the fault angle and degree of rotation were also
1 9 dV2 _ l/>m -Pwd[0(z)]
RC = -V~2 x - - (8) incrementally changed.
4 dz 2 p(z) dz

The only unknown in this equation is dK2/dz, Compaction

which may be calculated numerically. After the The distribution of the compaction effect is shown
blocks have tilted, the reflection coefficients cal- in Fig. 7. Compaction is restricted to the downthrown
culated along each layer are then compared to those hanging wall of the block and to the upper 2500 m
of the unrotated state and the changes are plotted. at that. The footwall side of the block remains unaf-
fected, except for a few meters down the fault from
Experiments
the surface/fault intersection. When a point at the
top of a formation on the hanging wall side of the
The effects of variation in lithology, height to width block is rotated the load acting on it is increased by
ratio of the block, angle of the faults and degree the secant of the angle of rotation (in this case 30).
of tilt, on compaction, geothermal characteristics, A point at the top of the same formation towards

f "

Fig. 4. (a) Fault block of shale, width (W) of 6 km, height (Th) of 2 km and angle of tilt () of 10; (b) = 30; (c) = 45
Physical changes in fault blocks caused by rotation and their implications for hydrocarbon accumulations 257

a)

b)

Fig. 5. (a) Fault blocks of shale of width (W) of 4 km, height (Th) of 4 km and angle of rotation (0) of 10; (b) = 30; (c) fault blocks
of sandstone, = 30; (d) fault block of shale, = 45.

the footwall side of the block has had its load The fault planes themselves are unaffected by
decreased by the rotation and/or faulting, implying compaction, since they are unloaded.
no compression.
The compaction effect is not only differently dis- Degree of tilt and height to width ratio
tributed laterally, but it also varies with depth. It is In the skyscraper type blocks (Fig. 6) deformation
easier to squeeze a very porous surface mud than in the upper formations is intense, but a large part
fully lithified shale at 3 km depth. The optimal depth of the block is unaffected. The wide warehouse type
for compaction is dictated by the scaling factor I in blocks deform throughout their length and increase
eqn. 1, which is a function of the lithology (Fig. 2) in intensity with increasing tilt (Fig. 4). As the angle
and is related to surface porosity 0 values. of tilt is increased there is always more block to
Although the reduction of thickness due to com- be added to the load (because of its length) in
paction is greater in shale, sandstone has a broader the rough form of sec# times the original weight.
zone of impact (Fig. 8). Increasing the angle of ro- The skyscraper type blocks, because of their narrow
tation amplifies the compaction effect (Fig. 8). The width, have only a very small maximum potential
reduction of the post-rotated (30) total height of increase in load (Fig. 6).
the block by compaction is 4.6% for sandstone and The most compaction occurs in blocks rotated 90.
3.3% for shale. For blocks rotated 45 these values The maximum change in weight takes place below
are 9.7% for sandstone and 7.3% for shale. the top left hand corner of the blocks, the excess
There is also a point on a formation where great- weight acting on this point can be crudely described
est compaction occurs. This point is very close to as PRwgW. Since /9Rav (bulk density of rock) and g
the change-over from compacting to non-compacting (gravity) are constants, the limiting factor is the width
sides of the block and is important in the subsequent of the block, W. The depth to which compaction is
discussion. evident is also controlled by the width of the block
258 J.E. Iliffe, I. Lerche and K. Nakayama

/o Change in a p p a r e n t thickness ( T )

10 15 20 25 30 35 40 45 50

shale
1 b) sandstone
0 = 45
: 30

Fig. 8. Graph illustrating the percentage change in apparent

thickness of layers 200 m thick, due to compaction following
rotation of 30() and 45 ( x ) of a block of sandstone and shale.

width of the block and angle of rotation.

When considering a block wider than the vertical
scaling factor, , compactional effects are minimal
until a tilt of 30 is reached. The rate of change
of the compaction effect is roughly proportional to
sin#.
Fig. 6. (a) Fault block of shale, width (W) of 2 km and height
(Th) of 6 km tilted by = 10; (b) = 30; (c) = 45.
Fault dip
The distribution and maximum depth of com-
e< ,*^ /Compactional folding paction vary with the changing fault angle (r). As is
decreased from 90, the area of the block in the zone
of compaction is decreased by its physical rotation
movement laterally out of the compacting region. In
this case the maximum depth, zmax, at which com-
paction can occur is limited once again by the scaling
factor, i, (eqn. 1) and width of the block, with:
Hangingwall Footwall
(Compacted) (uncompacted )
zmax = W sin + W cos tan(r - ) (10)
Fig. 7. One post-rotation compacted block exhibiting features of The results of changing r, zmax and are shown
folds, line of maximum compaction, an uncompacted footwall
and compacted hanging wall. The block is of shale 4 km high, 4
in Fig. 9. Consider a block 4 km wide, tilted at an
km wide with a fault angle of 80, rotated 30. Layers are 500 m angle of 45, and of infinite depth. Vary the fault
thick. angle by 90 (Fig. 9a), 80 (Fig. 9b) and 60 (Fig. 9c)
then zmax is 5.6, 4.75 and 3.6 km, respectively. The
(compare Fig. 4b and Fig. 6b). The maximum depth fault angle, (r), therefore plays a significant role in
at which compaction occurs may be written as: determining the amount of compaction taking place
and the depth range over which it occurs.
= Wsine (9)
Extension calculations
where zmax is the maximum depth of compaction. The surface of the blocks modelled appear short-
The compaction distribution of the blocks is related ened, whereas compactional folding occurs at depth
not only to the surface porosity, 9 and scaling (Figs. 4, 5 and 6). This type of deformation is
factor, i, i.e., lithologic type, but also, in part, to the not always obvious in field measurements. Areas
Physical changes in fault blocks caused by rotation and their implications for hydrocarbon accumulations 259

numbers are 0 (angle of H i t ) in degrees

W(km)

1 2 3 ^ 5 6 7

are 0 ( angle of t i l t ) in degrees

Fig. 9. Diagram showing the variation in maximum depth of compaction (zmax) with changing fault angles (r).

of extension are somewhat quantified by extension considering angle of tilt and the depth of the forma-
estimates and, where subsurface control is limited, tion in the block were measured and was found to
this extension is generally obtained by comparing the be low in blocks tilted less than 20, but in blocks
bed lengths of pre-faulted layers with those of post- tilted 45 the error at the surface was of the order
faulted layers. Compaction due to rotation will alter of 20%. However, at depths below 1000 m the error
the estimate of original bed length because origi- decreased to less than 5%.
nal bed length is now deformed, and differentially It is, therefore, recommended that bed length be
so depending on whether this deformation is at the measured at a depth of greater than 1000 m if block
surface or at depth (Fig. 6c). rotation is greater than 20 in the section being
In this study the error in estimates of extension assessed.
260 J.E. Iliffe, I. Lerche and K. Nakayama

a)

]b)
*1^%

6
^
9*

8*

' t '

I I I I I I I I

Fig. 10. Diagrams showing the rates of change in reflection coefficients with depth along formation boundaries as a result of block
rotation and compaction. All blocks are 4 km wide and 4 km high with a fault angle of 80. (a) Shale block = 10, (b) shale block
= 30, (c) sandstone block, = 30, (d) shale block = 45. Note: values are multiplied 105 times.

Reflection coefficients to give () at various depths. For this reason the

This model considers blocks of a homogeneous model was tested in the same way in which com-
lithology in which seismic waves propagate in a paction was considered, by changing the angle of
smoothly varying depth dependent medium. rotation () and comparing the results of pure shale
The most important factor affecting reflection co- and homogeneous sandstone blocks (Fig. 10).
efficients (RC) is the change in porosity with depth, Compaction is essentially a porosity decrease (Fig.
()/, which is dependent on the input lithologic 2) which causes an increase in bulk density (p) and
parameters t and 0 which are used through eqn. 1 velocity (V) of pressure waves (P waves). The values
Physical changes in fault blocks caused by rotation and their implications for hydrocarbon accumulations 261

isotherm over a depth range of 400 meters. The same

exercise was carried out for the 150C isotherm. The
maximum departure from the curve using a one-
dimensional calculation was of the order of 15C,
again only 10% of the value of the isotherm. This
temperature variation occurred over a range of 600
meters down from a depth of 3100 meters. Since
bottom hole temperatures measured in wells are
=
-- =
Shale
Sandstone accurate only to a confidence level of 10%, a two-
30 =
R =
D e g r e e s of t i l t ( t h e t a )
C h a n g e in r e f l e c t i o n dimensional analysis would not yield any more useful
coefficient
information.
The undulations of the isotherms in Fig. 3c are
Fig. 11. Graph illustrating the difference between ARC before
a direct result of rotation, since topographic effects
compaction and after compaction at each layer interface for are removed by assuming the downthrown basin
varying angles of tilt and lithology. is filled with material with a thermal conductivity
appropriate to a depth relationship consistent with
the unrotated blocks (eqn. 1). The compaction effect
of RC at each interface are decreased by compaction. on changing thermal conductivity is minor relative to
The areas of blocks where the greatest change in the rotational effect.
ARC has occurred are predicted to become more Areas of high thermal conductivity are rotated
transparent to P waves (Fig. lOa-d). But the maxi- closer to the surface in the footwall, whereas less
mum change in RC caused by compaction (Fig. lOd) compacted sediments are buried on the hanging wall.
is only 2-3% of the ARC variation over a depth of The hanging wall areas are, therefore, cooler relative
4 km. to footwall areas. The oil kitchen would therefore
RC values do not change laterally beyond the be on the downthrown side of the block (hanging
line of maximum compression (lmc) and the change wall) at a depth of greater than 1600 meters. The gas
occurs rapidly in the upper layers and peters out kitchen is at a depth of 3100 meters in the lowest
gradually below the optimum depth (Fig. 11). The part of the footwall and hanging wall.
values of ARC are an order of magnitude greater As the fault angle is increased, the amplitude of
in blocks tilted 45 than in those tilted 30. Also, the the isotherms also increases (compare a and b of Fig.
zone of greatest change in RC is much narrower in 12). Since sandstone has a higher conductivity than
the vertical dimension in the 45 tilted blocks than in shale (compare b and c of Fig. 12) the isotherms
those rotated 10. These results are implicitly tied to are spaced further apart i.e., the geothermal gradient
changes in compaction, consequently RC values do is less for sandstone. A small change in basement
not significantly change in blocks tilted by less than heat flow from 1.2 (Fig. 12b) to 1.5 HFU (Fig. 12d)
30. greatly affects the position of the isotherms, although
Comparing blocks of pure shale and sandstone, the the general pattern is maintained.
region of greatest change in RC (not to be confused
with the optimal region for reflections) is at about Stress considerations
1000 meters for shale (see also Fig. 11) and in all the During the calculations to determine the change
shale cases (Fig. 10a, b and d) the maximum ARC in position of a formation due to compaction, the
occurs in a relatively narrow band. In the sandstone original weight Wt(x',') and final weight Wt(x,z)
case (Fig. 16c), the maximum ARC is not only are calculated. The difference between these weights
deeper (between 1500 and 2500 meters), but is also is an indication of the change in magnitude of stress
much wider than the shale cases a direct result along each bedding plane. The compaction of shale
of the lithologic effect, i.e., the ARC distribution is and sandstone is therefore strain (in the form of
responding to the d^(z)/dz curves (eqn. 1, Fig. 2). alignment of shale grains and dewatering) as a result
of this stress. Beds act as passive strain markers.
Geothermal effects Consider a block of identical dimensions, tilted
The first assessment of the geothermal effects was the same amount, but consisting of homogeneous
to determine whether a two-dimensional calculation unjointed granite. Although the magnitude of the
was necessary. A horizontal line was drawn across changes in overburden have changed, the points at
the tops of the "waves" of the 100C isotherm at which the new weight is greater are in approxi-
1600 meters depth in a shale block tilted 30 (Fig. mately the same positions as in the compressible
3c). The temperature varied 10C or 10% of the sedimentary blocks. The pattern of stress increase is
262 IE. Iliffe, I. Lerche and K. Nakayama

(b)

(c)

(d)

Fig. 12. Temperature distribution caused by differential compaction after rotation calculated using the one-dimensional method. All
blocks are 4 km wide and 4 km high, (a) Shale, = 10, Q = 1.2HFU; (b) shale, = 30, Q = 1.2 HFU; (c) sandstone, = 30,
Q = 1.2HFU; (d) shale, = 30, Q = 1.5 HFU; (e) shale, = 45, Q = 1.2 HFU.

insensitive to block composition. (line of maximum compaction or lmc). If this block

A phenomenon inherent in the rotated blocks of were of a brittle rock (such as granite) and the
sediment is a point along each formation at which increase in stress was large enough to overcome
compaction is greatest (Fig. 7). These maximum the shear strength of the material, then a fault may
compactional points for each of the formations may appear in this position. If the granite were indurated
be joined to form a line from the top of the block with joints in various directions, the pre-existing joint
down. This line represents the region in the block which lies in the most favorable orientation (closest
where the stress has undergone the greatest change to that predicted here) will release the rotation-
Physical changes in fault blocks caused by rotation and their implications for hydrocarbon accumulations 263

created strain by acting as a fault, but requiring less in an interbedded sandstone and shale block, would
stress than before. compactional folding occur?
Tests to date suggest that the angle of intersection Since the amount of compaction of layers differs
between this secondary fault and the main fault only slightly between lithologies in the zone between
below is a nearly constant 35 regardless of the 1000 and 1600 m, folding in the layers in an inter-
degree of tilt, for a fault angle of 80. If the fault layer situation would be just as prevalent as in the
angle is 60, the angle of the intersection becomes monolithologic blocks.
45. When considering the change in reflection coeffi-
This relationship has been independently observed cients in an interbedded scenario it is first appropri-
in the field. Miller et al. (1983) working in the Snake ate to consider how reflection coefficients will vary
Range, Nevada, U.S.A., describe "second genera- with depth in the unrotated state. Following the dis-
tion" faults which intersect older faults at an angle of cussion above of change in thickness and porosity of
40. They also describe the faults as beginning in a layers, it is possible to sketch an approximate vari-
near vertical orientation as predicted by this model. ation of reflection coefficients with depth (Fig. 13).
Similar observations have been made by Angelier If these blocks are subsequently buried by rotation,
and Colletta (1983) in the Gulf of Suez, Egypt; West- the effect of compaction will change the reflection
ern Gulf of California, Mexico; and the Southern coefficient by decreasing them in the top of the block
Basin and Range, U.S.A. They describe the occur- (above 1000 m), since the shale compacts towards
rence of near vertical (80-90 dip) tension gashes the porosity and density of the sandstone. The region
formed in the early stages of extension, when blocks between 1000 and 1600 m will be largely unchanged,
are tilted 5-10, which later become active as faults whereas directly below this zone, the reflections will
as tilting progresses. decrease once more. Because the more porous sand-
The 80-90 dip of the tension gash with a tilt of stones compact more than the less porous shales,
20 in a fault with an original dip of 60 yields an the result is a tendency of layers to be closer to-
intersection of fault and tension gash at 40. gether. Reflectors should become more transparent
Thorough testing of different block dimensions, in areas of compaction (i.e., footwall) in the zone
fault angles and angles of tilt has been carried out, above 1000 m and below 1600 m, whereas those
providing firm corroborative support for the ruggedly between these depths will be largely unchanged, but
stable nature of the effect. The model suggests a very transparent (Fig. 14).
mechanism for the development of secondary faults Disruption of reflectors in the areas predicted by
that is not limited to compressible sediments and is
substantiated by reported field accounts.
RC

Discussion

The problem with all synthetic models lies with

their intrinsic assumptions. These may be of a for-
givable oversimplification, such as assuming a fault L yX// Precompaction
block to be composed of a single lithology or of a s. Y
more serious nature when the actual physical phe-
oj
|
I
Postcompact ion
c I
nomena are not fully understood. It is therefore
important to question the limitations of assumptions ^ I
assessing the validity of a model.
r I\
Interbedded scenarios 11 \\
Consider the limitations and validity of the as- * \\
sumption of a lithologically homogeneous block. \\
Clearly, situations with interbedded varying litholo- I
gies would be more geologically appropriate. A '
Above a depth of about 1000 m, shale layers I //
compact more than sandstone (Fig. 8), between 1000 //
m and 1600 m they compact by the same amount //
and their porosities are quite similar (Fig. 2), but I
Fig. 13. Sketch illustrating the suggested change in reflection
below 1600 m, sandstone layers actually compress coefficients with depth pre- and post-compaction for a block of
more than shale and their porosity is lower. Thus, interlayered sandstone and shale.
264 IE. Iliffe, L Lerche and K. Nakayama

Interbedded layers will not influence the basic

trend of isotherms, but because the geothermal
gradient is dependent on the thermal conductivity
(which varies with lithology), a mixed block will have
a lower average geothermal gradient than pure sand-
stone, but higher than homogenous shale. The result
will be intermediate between the sandstone and shale
cases (Fig. 12b and c).
The effect of stress change in an interlayered
situation would not be very different to that in
homogenous blocks, except that the depth at which
Strong reflectior
Transparent reflector
the stress change occurs would be deeper than in
LMC Line of maximum compaction shale.
Fig. 14. Sketch illustrating the distribution of reflectors in
a compacted block of interlayered sandstone and shale. The Fault geometry
intensity of the lines represents the relative visibility of the
Another assumption which requires validation is
reflectors.
that of the boundary conditions of planar faults as
opposed to listric faults. No compaction occurs on
the model is visible on seismic sections (Fig. 15a the faults themselves, although the rotation could
and b), although other explanations of the same just as easily be accommodated by listric faults. The
phenomenon may be possible. simplicity of the mathematical functions required to

Fig. 15. (a) Seismic section from Wernicke and Burchfiel (1982, their fig. 15), exhibiting folding of upper layers of "skyscraper" type
fault blocks, sequential secondary fault development and regions of transparent seismic reflectors in the hanging wall, (b) Seismic line
3630-1/2-85 across the Emerald Basin, Nova Scotia. Note the less contiguous nature of the reflectors in the hanging walls, and the
occurrence of secondary faults in the approximate positions predicted by the model (from Tankard and Welsink, 1989). They suggest
that diapirism has also occurred in the footwalls of these fault blocks.
Physical changes in fault blocks caused by rotation and their implications for hydrocarbon accumulations 265

rotate blocks on planar faults does not preclude

extrapolation of results into listric faulted regions
5ml
(Fig. 15a and b). Indeed, the problem of the void
5km
beneath the rotated blocks would be accommodated
by listric faults, whereas vertical foundering of the
situations modelled here may transpose the effects o E
o
of compaction. The speculation of a set of secondary *""" ^ ^ B&h o

faults may be more applicable to the listric fault
setting. Shale Shale jfljSa'lti|

Instantaneous rotation and lack of

syn-tectonic deposition
First, let us consider the effect that subsequent
deposition might have. Selected sediment deposition Fig. 17. The observed relationship of shale diapirs, faults and salt
movement in offshore Texas and Louisiana based on Woodbury
on the hanging wall side of the block, as in growth (1973).
faults, will cause increased compaction in that region,
and the changes described above will be accentuated.
If, however, the basin as a whole subsided and a "sag If blocks rotated and compacted gradually, the
fill" (Harding, 1984) type der ition ensued, all the changes in stress applied at certain points in the
effects would be moderated, since compaction effects blocks would also constantly change, affecting the
are decreased with depth, making natural examples orientation and frequency of secondary faults, per-
of phenomena described here more difficult to find. haps in a sequential fashion not unlike the model
of development proposed by Angelier and Colletta
(1983) (Figs. 15 and 16).
In general, therefore, we conclude that the ob-
served results of the model are consistent with natu-
rally occurring examples.

Diapirism
A further implication of the model is a possible
cause of diapirism. Rotation causes compaction. If
=10 the water loss occurs after rotation and if the blocks
were composed of interbedded shales, the fluid mo-
tion would be curtailed by impermeable shale layers
andflowwould be more along the bedding to the un-
compacted sides of the block, rather than upwards.
This concentration of fluids in the hanging wall of
the block might be a causative mechanism for mud
= 20
diapirism and/or areas of accumulation of hydrocar-
bons. The effect would be amplified by the dehydra-
tion of smectite to illite at around the depth at which
the temperature attains 100C (Bruce, 1981). This ef-
fect is, of course, concordant with the accepted ideas
of diapir formation and shale diapirs in the footwalls
of faults in muddy deltaic settings, as was observed in
Tertiary of offshore Louisiana, U.S.A. by Woodbury
0=30 et al. (1973) (Fig. 17) and in the Upper Carbonifer-
ous of the West Coast of Ireland (Rider, 1978), and
would seem to provide observational support of the
conjecture.
Salt diapirism may also be triggered by fault block
rotation. If a salt layer acts as an extensional zone
Fig. 16. Sketches suggesting a possible sequential formation of
secondary faults towards the footwall during progressive rotation, of detachment for blocks to tilt on, the changes in
as suggested in the seismic section of Fig. 15a Wernicke and variable load beneath different regions of the block
Burchfiel (1982). may encourage the salt to flow "up gradient" into
266 IE. Iliffe, I. Lerche and K Nakayama

Fig. 18. Diagram summarizing the effects of rotation and compaction on temperature and fluid flow of possible hydrocarbons. Row is
from the hanging wall to the footwall as shown by arrows. Blocks are 4 km high and 4 km wide composed of shale, tilted 30 with a
fault angle of 80.

the uncompacted and unloaded region below the themselves are not affected by compaction, since they
footwall of the block, as also observed by Woodbury are exhumed and unloaded.
et al. (1973). (4) Strata lengths are increased by low ampli-
tude folding due to compaction. Increases in bed
Implications for hydrocarbon accumulations length cause underestimation errors in extension es-
The implication that fluid will migrate from a com- timates of up to 10%. This value decreases with
pacted area in the hanging walls of fault blocks to depth and reliable bed lengths can be obtained at
the uncompacted region in the footwall suggests that depths greater than 1 km. Surface measurements are
hydrocarbons reservoirs may be prevalent in these reduced, suggesting an overestimation of extension.
areas. Such reservoirs are widely recognized. What is (5) The change in reflection coefficient with depth
interesting is that the compacted hanging walls are due to compaction is a minor effect (<3%) compared
also the hottest areas, and thus may be kitchens for with the cumulative change with depth. Reflection
hydrocarbon maturation, with migration to the foot- coefficients in interlayered sand and shale scenarios
walls (Fig. 18). It is also possible that the upfaulted are reduced above 1000 m depth and below 1600
high thermally conductive areas may act as conduits m, whereas the zone between remains largely un-
for heat to reach upper, less conductive areas of the changed. These effects are increased with increasing
block and heat them above their expected tempera- tilt of the blocks.
tures, much like salt diapirs do in the Gulf of Mexico, (6) Seismic evidence is consistent with the pre-
U.S.A. (Vizgirda et al., 1985). This principle can be diction of the occurrence of compactional folds,
compared to putting a skewer in a potato to make it secondary faults and zones of more transparent re-
bake more quickly. flectors in the region between 1000 and 1600 m
and more coherent reflectors in the footwall side
Conclusions compared to the hanging wall.
(7) Model results show that the hanging walls
(1) The maximum depth at which compaction may of rotated blocks have an elevated temperature
occur is governed by the scaling factor (), the width compared to the footwall, mainly due to rotation.
of the block (W), the angle of tilt () and dip of Isotherm trends do not change with lithology al-
faults (r). The dominant factors are and W. though isotherm values do change. The act of ro-
(2) Compaction is confined to the hanging wall tation brings the higher conductive regions of the
side of the blocks and is greater in the upper 1000 block to the surface which may act to focus heat from
meters and is also greater for changes of tilt in excess below and dissipate it higher in the lithologic section.
of 30 and results in an overall rounding of the shape This effect can cause local elevation of temperatures,
of the block. Layers are reduced in thickness by up particularly in syn-rift deposits, if present, which may
to 10% in moderately tilted blocks. have important connotations on hydrocarbon play
(3) Faults dipping more steeply yield more com- designs in such tectonic regions.
paction than those dipping gently, although the faults (8) Analysis of the change of overburden in the
Physical changes in fault blocks caused by rotation and their implications for hydrocarbon accumulations 267

areas where compaction takes place indicates that Bruce, C.H., 1981. Smectite dehydration its relation to
the stress configuration is changed due to rotation. A structural development and hydrocarbon accumulation in
Northern Gulf of Mexico Basin. Am. Assoc. Pet. Geol., Bull.,
line of maximum compaction can be drawn through 66: 217-224.
points along each formation which have the greatest Guidish, T.M., Kendall, C.G.St.C, Lerche, I., Toth, D.J. and
compaction. This line of change in stress is a likely Yarzab, R.F., 1984. Basin evaluation using burial history
position for strain release in the form of a secondary calculations: an overview. Am. Assoc. Pet. Geol., Bull., 69:
fault. This result is consistent with the independent 92-105.
Harding, T.P., 1984. Graben hydrocarbon occurrences and struc-
field observations. Secondary faulting will occur even tural style. Am. Assoc. Pet. Geol., Bull., 68: 333-362.
in incompressible rocks provided the increase in Lewis, C.R. and Rose, S.C., 1970. A theory relating high
stress is sufficient to overcome the shear strength of temperatures and overpressures. J. Pet. Technol., 22: 11-16.
the material. Magara, K., 1976. Water expulsion from elastic sediments during
(9) Fluid migration will take place from the com- compaction direction and volumes. Am. Assoc. Pet. Geol.,
Bull., 60: 543-553.
pacted side of the block up-gradient to the uncom-
Miller, E.L., Grans, P.B. and Garing, T, 1983. The Snake
pacted side and then vertically upwards. Such flow Range decollement: an exhumed mid-Tertiary ductile-brittle
may manifest itself in nature as observed shale di- transition. Tectonics, 2: 234-263.
apirs in muddy deltaic regions or serve to redistribute Rider, M.H., 1978. Growth faults in Carboniferous of Western
any hydrocarbons present. Local changes in overbur- Ireland. Am. Assoc. Pet. Geol., Bull., 62: 2191-2213.
Sclater, J.G. and Christie, P.A.F., 1980. Continental stretching:
den over an originally horizontal salt layer may also
an explanation of the post-mid-Cretaceous subsidence of the
cause halokinesis. central North Sea basin. J. Geophys. Res., 85: 3711-3739.
(10) Finally, further burial of the rotated blocks, as Tarkard, A.J. and Welsink, HJ., 1989. Mesozoic extension and
commonly seen in extensional basins, will reduce the styles of basin formation in Atlantic Canada, in extensional
effects discussed above. tectonics and stratigraphy of the North Atlantic Margin. In:
A.J. Tarkard and H.R. Balkwill (Editors), Am. Assoc. Pet.
Geol., Mem., 46, 12, pp. 175-195.
Acknowledgements Vizgirda, J., O'Brien, J. and Lerche, I., 1985. Thermal anomalies
on the flanks of a salt dome. Geothermics, 14: 553-565.
Thanks and appreciation go to the Industrial As- Wernicke, B. and Burchfiel, B.C., 1982. Modes of extensional
sociates of the University of South Carolina Basin tectonics, J. Struct. Geol., 4: 105-115.
Woodbury, H.O., Murray, LB., Pickford, P.J. and Akers, W.H.,
Modelling Group for financial support, and also to 1973. Pliocene depocenters, outer continental shelf, Louisiana
Donna Black for typing the manuscript. The paper and Texas. Am. Assoc. Pet. Geol., Bull., 57: 2428-2439.
has benefited greatly by the critical and insightful Wyllie, M.R.J., Gregory, A.R. and Gardner, L.W., 1956. Elastic
reviews of Bjorn T0undbakenin and Rob Young. wave velocities in heterogenous and porous media. Geo-
physics, 21: 41-46.

References
Angelier, T. and Colletta, B., 1983. Tension fractures and exten-
sional tectonics. Nature, 301: 49-51.

J.E. ILIFFE Department of Geological Sciences, University of South Carolina, Columbia, SC 29208, U.S.A.
I. LERCHE Department of Geological Sciences University of South Carolina, Columbia, SC 29208, U.S. A.
K. NAKAYAMA Department of Geological Sciences University of South Carolina, Columbia, SC 29208, U.S.A.
 269

Techniques for the geometrical restoration of sections: an

example from the Bjornoya Basin, Barents Sea shelf
Alastair Beach and Mark G. Rowan

Structural analysis and modelling are important aspects of seismic interpretation for both exploration prospect evaluation and
field development studies. One of the principal methods for validating structural interpretations of seismic data is by carrying
out systematic restoration of sections, and the routine application of these techniques greatly improves the understanding of the
geological development of a structure, including the fault geometry, the stratigraphic development and the reservoir geometry.
Without the availability of sophisticated software, such methods are time consuming and often inaccurate, but utilising such
software allows the techniques to be carried out objectively, with accuracy and in a way that can be repeated. This paper
describes the application of these methods to evaluate a specific section from the Bj0rn0ya Basin on the continental shelf of the
Barents Sea.
The principal structures shown in the section are a result of two stages of fault movement, firstly during Jurassic extension
and secondly during Cretaceous subsidence. The greatest uncertainty in the original interpretation relates to the geometry of
the faults, the correlation of horizons across the faults and the thickness and position of units within each fault block. The
restorations enable modifications to be made which improve the consistency and viability of the interpretation and reduce the
uncertainty.

Introduction example, it is recognised that hanging wall rollovers

can develop against planar faults, making it difficult
A seismic line has been selected for analysis from to predict fault profiles from hanging wall geome-
a data set covering the Bj0rn0ya Basin, to the west try (Gibson et al., 1989). Similarly, there is usually
of the Loppa High on the continental shelf of the enough uncertainty in some of the factors that affect
Barents Sea. This analysis illustrates several aspects the results, such as decompaction, determining the
of section restoration, carried out with the com- undeformed regional geometry and identifying the
puter software Geosec-20 from Geo-Logic Systems, appropriate algorithm for fault projection (Rowan
Inc. It is only since the development of software and Kligfield, 1989), so much so that application of
packages that restoration of sections can be at- the technique can be quite misleading.
tempted routinely and with precision. The concept Two important points emerge from the above con-
of the balanced section was described sometime ago siderations. The first is that section balancing in
(Dahlstrom, 1969; Hossack, 1979), and more recently extensional terranes is, at the present time, a poorly
developed computer techniques have concentrated understood process. It is more instructive to carry
largely on methods for testing and verifying that a out sequential restorations of interpreted sections,
section is balanced. In compressional terranes, these back through geological time, rather than trying to
include checking that bed lengths and cross-sectional produce unique, balanced sections. The distinction
area are preserved, predicting the depth to detach- between restoration and balancing is important: in
ment and checking that the geometry is a viable balancing, a set of rules and assumptions, which may
one that can be restored to its unstrained state. In or may not be valid, are applied to constrain an in-
extensional terranes, sections are often balanced by terpretation, whereas in restoration, the deformation
predicting linked listric fault trajectories from the is removed so that the undeformed and intermedi-
shapes of the hanging walls (Gibbs, 1983; White et ate geometries can be evaluated. The second point
al., 1986; Williams and Vann, 1987). Recent geologi- is that the computer software used for carrying out
cal studies of normal faults, however, raise questions the restorations should make no geological assump-
about the assumptions made in employing this geo- tions about fault geometry in the section and that
metrical technique in drawing balanced sections. For it should be equally applicable to listric or planar

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 269-276. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
270 A. Beach and M. G. Rowan

2-\

4
H- ^

Fig. 1. Line drawing from a seismic section on the flanks of the Loppa High and the Bj0rn0ya Basin. The line and its exact location
are confidential. In this figure, seismic events have been traced from the original section without any interpretation and in an attempt
to show objectively the reflections present.

faults. The sequence of restorations then provides a are both consistent and acceptable to the geologist.
precise record of the geometrical evolution of the It, thus, provides for a more objective assessment of
section through successive geological stages. the structure, stratigraphy and reservoir geometry,
The analysis of the example line from the Bj0rn0ya although it does not necessarily provide a unique
Basin (Fig. 1) illustrates the fundamental steps in solution.
restoring a section and is typical of many exploration Prior to carrying out any restorations, it is neces-
problems where the seismic data are of reasonable sary to establish the most accurate depth geometry
quality, but with some ambiguities, and where few for the section. Any errors in the geometry at this
well data are available. The starting point is an initial stage will be maintained and carried forward in the
interpretation of the horizons and faults (Fig. 2) and restored sections, and may give rise to incorrect geo-
sequential restoration of this section will help answer logical interpretations. The amount of data available
questions such as: for depth conversion will vary enormously according
- is the interpretation of the general structural to whether any wells have been drilled and if so, how
style acceptable? many. In association with the line studied here, no
- what is the most consistent interpretation of the wells are available, and depth conversion has been
fault geometry? based on the use of stacking velocities. Seismic veloc-
- are the horizon correlations across the faults ities used for depth conversion are: seawater, 1480
acceptable? m/s; Tertiary, 2600 m/s; Cretaceous, 3000 m/s at the
- is the reservoir geometry within each fault block top increasing to 4000 m/s at the base; Jurassic-
valid? THassic, 4000 m/s at the top increasing to 4250 m/
- what is the deformation history of the interpreted s for the deepest layer interpreted. As more data
structures? become available, a more accurate set of restored
Geometrical restoration of sections is primarily sections becomes possible, as a result of having an
concerned with checking the validity of interpreta- improved depth section to start with.
tions of seismic data and with verifying that these The depth geometry is also affected by compaction
interpretations are based on geological models that of sediments and, therefore, decompaction must be
Techniques for the geometrical restoration of sections: an example from the Bj0rn0ya Basin, Barents Sea shelf 271

\N E

___^^^^^==^>^^^ j f e ^ i g g
1- - lertiary - ^

^^II^^"^ljrr^
- ' "

L__~~ V ^ ^ y
_ ~ zz==
f^~ , r rr^aceous
==^- ' %^I^3~//
2 Z --" "~ _ Xba' r "' / X
r-" . -- -&- - f~' ! * f ~
j_ Cretaceous -J^Tt^^^^T~ -J ~~w "' " ' ""*^ c r - ^ ^ ^ ^ ^ ^ ^~^~ - > ^ ^ ^
5 ,n r
- y^j^MHtoS^ij^ _ v ^ J 'ass/c^><-L ^S^-r^^^^ ^ ^ ^
(0
a>
S^^\^J^
^^^ ^ I _ ,- ~^r
^-^^ "~
3- JIT"

^^^^~ ~^-^^
^ ^ ^ " ^ ~~~~^>-
^^^f2^^~ /jS^ ^~~^*^^
~ ^
4-
?

5-

/- 0^ _ 5 10

I
km

Fig. 2. A preliminary interpretation of the line drawing shown in Fig. 1, based on event correlations and cut-oifs. The base Tertiary and
base Cretaceous are readily identified, but the stratigraphy in the pre-Cretaceous section is not as clearly understood.

carried out during restoration. Few data are available restore, but any changes must be converted back to
for the line studied here and decompaction functions time to check for compatibility with the data. Finally,
from Sclater and Christie (1980) have been used, a more powerful method is to make changes to the
assuming that the lithologies are claystone in the Tbr- restored section, then compact and forward model to
tiary and Cretaceous and a 50/50 mix of interbedded the deformed geometry. The resulting interpretation
sandstone and claystone in the Jurassic and TViassic.

INTERPRET
Section restoration SEISMIC

\
> DEPTH
A single routine restoration of a section is a rel- CONVERT

atively simple process involving depth conversion, i

MODIFY DEFORMED CHOOSE HORIZON
STATE(DEPTH)
decompaction and then restoration to the chosen ge- FOR RESTORATION
i
ological time horizon. However, if the restoration is DECOMPACT

to be used to evaluate the interpretation and to make 1

RESTORE
modifications where deemed necessary, the proce- CHOOSE METHOD
I >f MORE
dure becomes much more complicated and involves I - FOR ADJUSTMENT * yes
T
^ RESTORATIONS? yyes
in
interactive manipulation of both the deformed and I ^^^^-^^
restored geometries (Fig. 3). After decompaction MODIFY
RESTORATION
<
' ~^0U T P LJTJ-^*"
and restoration to one horizon, the resulting ge- I
ometry must be evaluated in order to identify any FORWARD MODEL

problems arising from a poor interpretation. Any COMPACT

such problems can subsequently be modified in one I

CONVERT TO TIME
of three ways. The time interpretation can be ad-
justed and then depth converted, decompacted and i
COMPARE
WITH DATA
restored again, so that a better interpretation even-
Fig. 3. Flow chart for the simple restoration and modification
tually results through trial and error. A variation is to of seismic interpretations in extensional terranes. See text for
modify the section in depth and then decompact and discussion.
272 A. Beach and M. G. Rowan

is one which restores to an acceptable geometry, but Figure 1 shows a line drawing of the seismic line,
it too must be converted to time for comparison with made by tracing reflections from the original data,
the data. Usually, some combination of all these pro- and Fig. 2 shows an initial interpretation of this
cedures is employed in modifying and improving an line. Figure 4 is the result of depth-converting the
interpretation, and only the use of appropriate soft- interpretation and is simplified in that the minor,
ware, with its inherent speed and precision, makes it intra-block faults are not shown. The depth section
a practical proposition. shows a westerly dipping, low-angle detachment fault
A standard simple restoration assumes an origi- with a sequence of rotated fault blocks in its hang-
nal horizontal depositional surface and, thus, uses a ing wall. Although the data could be interpreted
flat geometry as the template for the horizon being with non-linked, more planar faults, other nearby
restored. Alternatively, if the structure is a linked lines support at least the local existence of a dip-
fault system, the basal detachment can be used as ping detachment at the shallow, eastern end of the
the restoration template. Neither of these methods, line. Whether or not it continues to the west un-
however, takes into account possible original topog- der the deeper fault blocks is unknown, but should
raphy and paleobathymetry at the time of the chosen not change the main conclusions derived from the
restoration. For example, whereas a formation de- restorations. The fault blocks are truncated by the
posited regionally at or near sea level can be restored base Cretaceous unconformity and overlain by Cre-
accurately using a flat template, a formation that only taceous and Tertiary sequences that thicken to the
partially filled semi-starved basins bounded by fault west. After depth conversion, the basal detachment
scarps would be restored incorrectly by the same is seen to have a more gentle dip to the west than is
method. Where the data are available, it is possible apparent in time (compare Figs. 2 and 4). In addi-
to incorporate isostatic effects and thermal subsi- tion, the anticlinal folding of events in fault blocks 2
dence models (e.g., McKenzie, 1978) to calculate the and 3 (Fig. 2) disappears in depth (Fig. 4) and hence
paleobathymetry at any given point and time (e.g., these are considered to be velocity anomalies in the
Bertram and Milton, 1989). This can be done for nu- time section.
merous points along a section, thereby recreating the Figure 5 shows the first stage of section restoration
paleotopographic profile which forms the template where the Cretaceous and Tertiary sequences have
for a more accurate restoration. In the case of the ex- been removed and the section restored to a contin-
ample from the Barents Sea, an unknown amount of uous and horizontal base Cretaceous unconformity.
Tertiary uplift and erosion affected the area (Wood This unconformity is interpreted to have been caused
et al., 1989), so that such calculations can not be by erosion, at or near sea level, of the rotated and
made. Restorations, therefore, were confined to the uplifted footwalls to the major faults during ex-
simple flat regional and basal detachment templates. tension and deformation. Making the unconformity
For both of the horizons restored, however, these horizontal in the restoration is probably a valid ap-
form adequate approximations, and potential errors proximation for most of its length, as evidenced by
should not affect the conclusions drawn from the the abundance of truncated events. In reality, how-
restorations. ever, there may have been paleotopography at the

km
Fig. 4. Depth conversion of the interpretation shown in Fig. 2. Numbers are fault blocks referred to in the text and x is the upper
Jurassic horizon to be restored.
Techniques for the geometrical restoration of sections: an examplefromthe Bj0rn0ya Basin, Barents Sea shelf 273

km
Fig. 5. Restoration of the depth section shown in Fig. 4 to a horizontal base Cretaceous unconformity, thereby removing the overlying
differential subsidence of the Tertiary and Cretaceous, the small fault offsets at base Cretaceous and decompacting the underlying
sequences. Numbers refer to fault blocks and x marks the next horizon to be restored.

end of the Jurassic in the form of small relict basins The bulk deformation has been modelled in both
in the immediate hanging walls of the major faults. restorations by vertical simple shear, as opposed to
Three stages are involved in constructing the the inclined simple shear suggested by White et al.
restoration shown in Fig. 5: (1) the removal of the (1986). Since changing the angle of shear used in
Cretaceous and Tertiary sequences and the accom- restoration affects the total amount of extension in
panying decompaction of the underlying sediments; the section, one way of evaluating which angle to use
(2) the restoration of fault offsets at the base Cre- is to restore on a fault template and choose the angle
taceous; and (3) the restoration of the unconformity which brings the chosen restoration horizon back to
to horizontal. The post-Jurassic regional differential regional by reversing the appropriate amount of ex-
subsidence and associated faulting has been removed tension. For example, if removing fault offsets using
and, hence, the fault geometry at the end of the main vertical simple shear on a given fault template leaves
phase of Jurassic extension has been established. the horizon below the level regional, a shear angle
This shows a series of rotated and truncated fault inclined antithetic to the fault will raise the horizon
blocks above a roughly horizontal detachment. and thereby improve the restoration. However, it is
The next stage restores the section to one of the crucial that the appropriate fault template be used.
upper Jurassic events (marked x in Figs. 4 and In the example presented here, if the upper Juras-
5), using the basal detachment and its footwall as sic event is restored in one step from the present
the template onto which successive fault blocks are deformed geometry, using the template provided by
restored by removing the displacement at the chosen the dipping detachment of the original depth section
horizon (Fig. 6). Where the restored horizon has (Fig. 4), a shear angle of 70 antithetic to the fault is
been truncated by erosion, its geometry has been found to work best. If, instead, the base Cretaceous
projected using underlying events as a guide, so that unconformity is first restored as described above, so
the missing material has been reconstructed in the that the Jurassic event is subsequently restored using
restoration. the level fault template of Fig. 5, vertical simple

w
>*

/ 7 6^^
/ 4 / y/ ly"^
5 3 S ^ 2

km
Fig. 6. Restoration of the geometry shown in Fig. 5 to the upper Jurassic horizon ( x ) using the basal detachment of Fig. 5 as the
template. See text for discussion of problems apparent in the restoration. Numbers refer to fault blocks.
274 A. Beach andM.G. Rowan

iTertiary

-10
km

km

Fig. 7. Final interpretation in depth, generated by multiple cycles of restoration and modification, showing a consistent geometry of
faults, horizons and unit thicknesses (compare with Fig. 4). Numbers are fault blocks referred to in the text.

shear works better. The implication is that, at least modifications are (compare Figs. 2 and 8, and Figs. 4
for this example, if the section is restored properly in and 7): a steepening of the basal detachment; minor
a sequential manner, vertical simple shear is a better adjustments of the deeper portions of some of the
algorithm for approximating the bulk deformation fault trajectories; a reinterpretation of the unconfor-
(see also Rowan and Kligfield, 1989). mity in fault block 6; significant adjustments to the
Figure 6 may be used as a preliminary check on the interpreted horizons, especially in fault blocks 2 and
validity of the interpretation, for example, by evalu- 4; and a raising of the entire stratigraphic sequence
ating fault trajectories, the geometry of the restored in blocks 6 and 7.
horizon and the correlation of horizons across faults. It is, of course, always possible to continue making
The principal ways in which the section does not re- minor modifications to a section in order to per-
store properly are (Fig. 6): firstly, in the irregularity fect its restoration. However, the limitations of the
of the uppermost Jurassic event from which fault off- method are exceeded after a certain point and it is
sets were removed; secondly, in the reversals of dip felt that with the data available here, the restorations
along the restored basal detachment; and thirdly, the and modifications made are sufficient to remove the
mismatches between the same horizons across the most important inconsistencies in the initial interpre-
restored faults. The latter case is especially obvious tation and adequately produce an acceptable section
for fault block 2, which has a thinner sequence at one with horizon correlations that are internally consis-
level than the blocks to either side and for the fault tent. Restoration of the section provides a clear and
between blocks 4 and 5, which shows reverse offset objective assessment of the structural geometry and
in the restored geometry. the stratigraphic development of the area. This can
Iterative modifications can be made to the ge- be used as the basis for an exploration prospect
ometry to produce a more acceptable interpretation evaluation and reduces considerably the geological
and restoration. These can be carried out in the uncertainty in such an evaluation.
deformed state and their effect tested by restora-
tion, or they may be made to the restored geometry, Conclusions
followed by forward modelling, compaction and con-
version to time to check for compatibility to the Fault movements occurred during two stages,
original data (Fig. 3). The first method has been firstly during Jurassic extension and secondly dur-
used to make major changes and the second method ing Cretaceous subsidence. At the eastern end of the
for minor adjustments. Numerous such modifications section, Jurassic-THassic sequences occur at seabed
were made, with interactive manipulation of both the and the Cretaceous and Tertiary thicken from zero
deformed and restored geometries. The net effect to over 4 km to the west. This is a period of re-
of all changes is an improved depth interpretation gional differential subsidence, primarily thermal, of
(Fig. 7) and its time equivalent (Fig. 8). The major the western flank of the Loppa Ridge. There is no
Techniques for the geometrical restoration of sections: an example from the Bj0rn0ya Basin, Barents Sea shelf 275

Fig. 8. Final interpretation of the seismic data which corresponds to the depth section of Fig. 7. The differences between this and the
original interpretation are discussed in the text.

obvious evidence on the seismic line studied here and represents the locus of maximum footwall uplift
for compressional structures of either Cretaceous or during the Jurassic faulting. Successive fault blocks
Jurassic age. However, some slight compressional re- further west all show rotation of Jurassic sequences,
activation along the basal detachment in the Tertiary but less marked truncation below the unconformity.
may explain the present lack of rotation of beds in Analysis of the truncations and the syn-rift sediments
fault block 1. indicates that blocks 4 and 5 form a major coherent
The restoration to the base Cretaceous uncon- unit divided by a relatively minor fault and that the
formity shows that a pre-Cretaceous geometry of same is probably true for blocks 6 and 7.
extensional fault blocks was established. These fault In conclusion, the geological modelling carried out
blocks lie above a basal detachment fault that prob- on the seismic line has resulted in an interpretation
ably dipped gently to the west at this time. Rotation that depicts a structural geometry and horizon corre-
of the fault blocks occurred and some footwall crests lation that is internally consistent. Although it is not
underwent erosion. The structure restores and recon- the only possible valid interpretation, it is certainly
structs in a way that suggests that it formed largely more acceptable than one which does not restore to
by dip-slip extensional movements. Although some an acceptable geometry. The restoration procedure
strike-slip movement may have occurred, it is not provides a clear and objective basis for assessing the
necessary to explain the current geometry. structure and stratigraphic sequences in any poten-
From east to west across the section studied, there tial exploration target and the results can easily be
are differences in the structure between adjacent extended to adjacent areas.
fault blocks. At the eastern end, THassic-Jurassic
sequences are interpreted to be sub-horizontal and
there is some element of rotation in fault block Acknowledgements
2, though Jurassic is still present at sea floor. The
third fault block shows a marked rotation and the We thank Geo-Logic Systems, Inc., of Boulder,
strongest angular discordance of Jurassic sequences Colorado, for access to the GEOSEC-20 software,
below the base Cretaceous unconformity. This cor- and several anonymous reviewers for their com-
relates well with the greater width of the block ments.
276 A. Beach and M.G. Rowan

References Rowan, M.G. and Kligfield, R., 1989. Cross-section restoration

and balancing as an aid to seismic interpretation in exten-
Bertram, G.T. and Milton, M.J., 1989. Reconstructing basin sional terranes. Am. Assoc. Pet. Geol., Bull., 73: 955-966.
evolution from sedimentary thickness: the importance of Sclater, J.G. and Christie, P.A.E, 1980. Continental stretching:
paleobathymetric control, with reference to the North Sea. an explanation of the post mid-Cretaceous subsidence of the
Basin Res., 1: 247-257. Central North Sea Basin. J. Geophys. Res., 85: 3711-3739.
Dahlstrom, C.D.A., 1969. Balanced cross sections. Can. J. Earth White, N.J., Jackson, J.A. and McKenzie, D.P., 1986. The
Sei., 6: 743-757. relation between the geometry of normal faults and that of
Gibbs, A D . , 1983. Balanced cross-section construction from the sedimentary layers in their hangingwalls. J. Struct. Geol.,
seismic sections in areas of extensional tectonics. J. Struct. 8: 897-910.
Geol., 5: 152-160. Williams, G. and Vann, I., 1987. The geometry of listric normal
Gibson, J.R., Walsh, J.J. and Watterson, J., 1989. Modelling of faults and deformation in their hanging-walls. J. Struct. Geol.,
bed contours and cross-sections adjacent to planar normal 9: 789-798.
faults. J. Struct. Geol., 11: 317-328. Wood, R.J., Edrich, S.P. and Hutchinson, I., 1989. Influence
Hossack, J.R., 1979. The use of balanced cross-sections in the of North Atlantic tectonics on the large scale uplift of
calculation of orogenic contraction: a review. J. Geol. Soc. the Stappen High and Loppa High, Western Barents Shelf.
London, 136: 705-711. In: AJ. Tankard and H.R. Balkwill (Editors), Extensional
McKenzie, D.P., 1978. Some remarks on the development of Tectonics and Stratigraphy of the North Atlantic Margins,
sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32. Am. Assoc. Pet. Geol., Mem., 46: pp. 559-566.

A. BEACH Alastair Beach Associates, 11 Royal Exchange Square, Glasgow Gl 3AJ, Scotland, UK.
M.G. ROWAN Alastair Beach Associates, 11 Royal Exchange Square, Glasgow Gl 3AJ, Scotland, UK
 277

Numerical simulation of jointed and faulted rock behaviour

for petroleum reservoirs

A.G. Koestler, A. Makurat, N. Barton and K. Monsen

Shearing along fractures and faults, and changes in fracture and joint conducting apertures are key elements of the behaviour
of petroleum reservoirs and in the evaluation of cap rock integrity. Improved joint characterization, constitutive joint modelling
and numerical block interaction codes have recently been developed, which can be utilized for parametric studies in these
complex engineering fields. The program UDEC-BB (Universal Distinct Element Code with Barton-Bandis joint behaviour
model) can be applied on simulations of reservoir compaction behaviour, in studies of gas storage in lined caverns and in
larger-scale studies of tectonic behaviour of faulted sediments under extensional deformation. The paper describes the input
data requirements, some special features of the scale-dependent joint model, and some recent results of numerical modelling of
faulted rock mass behaviour.
Fault configurations from offshore Norway are still little understood with respect to stress, displacement and rotation
distributions within fault blocks and fault zones. An improved understanding would provide a useful tool for the prediction of
minor faulting, which often cannot be mapped with traditional seismic surveys. This study concentrates on a large-scale fault
configuration similar to parts of the Gullfaks field.
The importance of non-linear joint behaviour in cases involving large changes of fluid pressure are emphasized. The use of
linear joint models with constant shear and normal stiffness and linear frictional properties are clearly limited when effective
stress changes of several tens of megapascals take place.

Introduction gravitational extension. This type of geological set-

ting is based on analogue situations reported from
The attempts in petroleum geology to predict actual hydrocarbon fields offshore Norway, such as
and model tectonic heterogeneities at reservoir scale the Gullfaks Field (see Fig. 9; Fossen, 1989; Pet-
cover a wide range of approaches, such as studying terson et al., 1990; Koestler et al., this volume).
onshore analogues, analyzing fault patterns at differ- The large fault blocks above the detachment were
ent scales and characterizing them e.g. with fractal designed in a way to give indications on the de-
geometry, modelling of fault patterns with complex formation behaviour of blocks of similar size, the
random parameter distributions, incorporating theo- influence of antithetic faults, and the differences in
ries from other scientific fields such as percolation a very narrow fault slice. The hanging wall with a
theory, cluster distribution theory, chaos theory, etc. total thickness of about 1600 m consists of homo-
The interest of simulation of fault patterns focuses geneous sandstone fault blocks, while the underlying
mainly on two points: (1) modelling of fault networks detachment is designed as a shale layer of about 330
for areas with limited and poor data; and (2) mod- m thickness. The footwall consists of sandstone of
elling of fault networks below seismic resolution for the same properties as the hanging-wall sandstone.
specific hydrocarbon reservoirs. Additionally the model is covered by 500 m of sed-
The better characterization of reservoirs in terms iments. The two-dimensional modelling is carried
of horizontal heterogeneities (sedimentological ori- out in the profile section representing the section of
gin) and vertical heterogeneities (tectonic origin) is plane strairi.
of vital importance for hydrocarbon production and The application of a recently developed numerical
for enhanced oil and gas recovery. simulation technique (UDEC-BB: Universal Distinct
In this study an attempt was made to simulate Element Code with Barton-Bandis joint behaviour
stress changes in a fault block configuration using model), which allows the realistic modelling of joint
numerical methods. A tectonic setting was chosen deformation behaviour as a function of normal and
with fault blocks above a low-angle detachment and shear stress, and modelling of the intact block de-

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 277-288. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
278 A. G. Koestler, A. Makurat, N. Barton and K. Monsen

formation and translation along joint due to forces

acting on these blocks, is supposed to lead to a better
understanding of the stress distribution inside the
intact blocks and along the faults.
The advent of discontinuum modelling tech-
niques pioneered largely by Cundall (1971, 1980)
has opened up a new field of geomechanics. This
new field is being actively applied to a variety of
geomechanics problems. An especially interesting
application has been the modified Universal Dis-
tinct Element Code termed FRIP (Fluid-Rock In-
teraction Program) reported by Pine and Cundall
(1985). These authors and other have applied the
code to the Cambourne School of Mines hot dry
rock geothermal project (CSM-HDR) in Cornwall,
England, in attempts to explain fluid-rock interac- Fig. 1. Classic hydraulic fracturing to make hydraulic connection
tion in the fractured granite at approximately 2 km between two wells may be impossible if pre-existing jointing is
depth. Anisotropie in-situ horizontal stresses, and non-parallel to an anisotropic principal stress system.
sub-vertical jointing striking at some 30 to the prin-
cipal stress, causes joint shearing when the effective The decrease in reservoir pressure due to deple-
normal stress is sufficiently reduced by fluid injection. tion and subsequent increase from modelled injec-
Figure 1 can be used as a simple illustration of this tions was simulated in a more simplified manner
mechanism, where the hydraulic connection between than Pine and Cundall (1985) or Harper and Last
well 1 and 2 will never be achieved. (1989). Row was not modelled since near-well effects
Pine and Cundall (1985) used an over-simplified were not of direct interest. Therefore it was possible
linear sub-routine to model joint shearing, dila- to utilize a more sophisticated non-linear joint sub-
tion, opening and closure, and their coupling of routine in the calculations. Some important aspects
joint aperture with hydraulic conductivity was also of this BB (Barton and Bandis) joint model will be
over-simplified. They justified their simplified joint described here to illustrate differences with respect
geometry and the joint behaviour model with the to commonly used linear models.
shortened calculation time achieved.
Harper and Last (1989) utilized Cundall's Uni- Linear and non-linear joint modelling
versal Distinct Element Code (UDEC) directly in
theoretical investigation of the influence of various A linear joint model as utilized by Harper and Last
joint patterns in rock-fluid interaction. They referred (1989) is illustrated in Fig. 3. A single shear stiffness
to modelling of fractured sandstone petroleum reser- (Ks) defines the slope of the shear stress-shear
voirs but utilized several of their results in attempts displacement lines (see Fig. 2a). No differentiation
to explain some of the phenomena observed or is made in this value of Ks with increasing normal
deduced from the CSM-HDR project referred to stress. Similarly, a single dilation angle (d) is used
above. These authors also used over-simplified, lin- to define the slope of the normal displacement-
ear joint behaviour models, which may have influ- shear displacement lines, independent of normal
enced the validity of their conclusions. They also stress (see Fig. 2b). In normal closure, a single value
used a direct coupling between joint aperture and of normal stiffness (Kn) defines the slope of the
hydraulic conductivity which is erroneous for small aperture-normal stress line up to some critical stress
apertures, due to the areas of rock-to-rock contact level, at which point a constant residual aperture (aT)
required for normal stress transfer. is attained, which does not reduce with increasing
In studies of reservoir compaction in connection normal stress.
with the subsidence of the North Sea Ekofisk reser- Most of the features illustrated in Fig. 3 are
voir Barton et al. (1986) utilized both linear an fundamentally incorrect. Work conducted by Bandis
non-linear joint behaviour sub-routines in UDEC. (1980), Barton and Choubey (1977), Gale (1982)
Reservoir fluid pressure draw-down of 20 MPa re- and others has shown that the shear stiffness (Ks)
sulting from 15 years of production were simulated, and normal stiffness (Kn) increase significantly with
and were shown to result in shearing along steeply normal stress. A further error is the assumption that
dipping conjugated fractures in the chalk. Figure 2 hydraulic conductivity (k) is equal to 1/12 of (a^),
illustrates the principal set up of the model. where ar is defined in Fig. 3c.
Numerical simulation ofjointed and faulted rock behaviour for petroleum reservoirs 279

62 MPa

* P

zero y-displacement
1 m

oora\v
(C) (d)
y y y

I I
I I I i I 1
1
i I i 1
1
i . ii

t I

Fig. 2. Results of non-linear distinct element modelling with UDEC-BB in studies of the Ekofisk reservoir compaction (Barton et al.,
1986). (a) Model set up. The 1 x 1 m high porosity chalk is surrounded by roller boundaries at the left, right and bottom boundary.
62 MPa vertical stress is applied along the top boundary, (b) Hydraulic fracture aperture after pore pressure reduction from 48 to 18
MPa. Thickness of lines is proportional to fracture aperture. Maximum joint aperture = 31.6 . One line thickness = 0.5 . (c)
Total displacement vectors due to pore pressure reduction (48-18 MPa). Maximum displacement = 4.5 mm. (d) Shear displacement
due to pore pressure reduction (48-18 MPa). Thickness of lines is proportional to the amount of shear. Maximum shear displacement
= 499 /xm. One line thickness = 10 .

An example of a more realistic non-linear joint ductivities is shown in Fig. 5. Physical apertures (E)
model (based on experiments), as developed by Bar- developed at three different normal stress levels (10,
ton et al. (1985), is shown in Figs. 4 and 5. Shear 20 and 30 MPa) are converted to smaller conducting
stress-shear displacement curves show increasing apertures (e) to derive the conductivity (;) given by:
shear stiffness (Ks) with increasing normal stress,
and joint size dependence is also modelled. Dilation k = (1)
curves are "delayed" by joint size (or block size) 12
variation, not by normal stresses. Figure 4c illus- The relationship between the theoretical conduct-
trates conductivity-displacement relationships from ing aperture (e) and the physical aperture (E) is
"origins" v4, B and C. The source of these initial con- derived from the test data presented in Fig. 6. The
280 A.G. Koestler, A. Makurat, N. Barton andK. Monsen

(a) z M
A (0
Q.
A ^ J 0 0 m m
T /^^mrrr 30 MPa
Increasing
normal </)
stress
(/)
LU JJ-^^JOOmm

A*
CC
w ^~~:~~~^===- 20 MPa
Shear stiffness h- JiV 2t>0mm
(/)
CC
SHEAR DISPLACEMENT
< |[V^__^00mm
10 MPa
LU 250mm
X
</>
F
1 1 1 1 1 1 1 I I I I I I I I I I I
o - > r u r o - < t m < o r v G o c n oi - - ' C \ -> (\
J m >* in
(M M
( M( M( \
l
1 .0

0 .9

0.

E 0. 7

SHEAR DISPLACEMENT F 0 .6

z 0. 5
o 0 .4
Dilation component
h-
(with shear) J< 0 .3
Q
0 .2

0. 1

0 .0
c>.-'C\jro'<tm(ONaoa)a'-'<\im'tm(Dr^cocj)o-'<\icr>''tm

-3.00
i Normal stiffness CM
Residual aperture -3.25
E
- -3.50

>- -3.75

-4.00
Fig. 3. The linear joint model assumed by Harper and Last l-
> -4.25
(1989). (a) Shear stress vs. shear displacement. Shear stiffness H
-4.50
Ks is assumed to be constant until peak, (b) Dilation vs. =
Q> -4.75
shear displacement. No dilation occurs until peak shear stress Z
is reached, (c) Joint aperture vs. normal stress. Normal stress o

-5.00
-5.25
is assumed to be independent of normal stress until residual
aperture is reached. o -5.50

o -5.75
-1 I 1 I I I I I L
-6.00

ratio E/e increases with increasing normal stress, and

Fig. 4. Non-linear shear-delation-conductivity model, showing
also shows roughness dependence. stress and joint size dependence (after Barton et al., 1985).
The plots show curves for 100 and 250 mm long joint samples,
sheared under 10, 20 and 30 MPa normal stress (JCSQ = 82.5
Input data for non-linear joint modelling MPa, JRC0 = 9.0, = 29.5). (a) Shear stress vs. shear
displacement, (b) Dilation (increase in mechanical aperture) vs.
The input data needed to generate the behaviour shear displacement, (c) Conductivity vs. shear displacement. A,
curves shown in Figs. 4 and 5 are provided by B, and C correspond to the initial conducting apertures prior to
four easily measurable parameters. The first three shear for 10, 20 and 30 MPa normal stress taken from Fig. 5.
of these parameters describe the frictional strength
of rock joints. The peak friction angle () under an
effective normal stress () is given by the following JCS = joint compressive strength; Schmidt ham-
relationship: mer tests (for weathered joints) or point load or
unconfined compression tests for unweathered rock
fJCS\ joints (where JCS = ac) (Barton and Choubey,
= JRC log ( j + (2) 1977).
= residual friction angle; tilt tests on flat sawn
Simple index tests are used to measure the parame- surfaces of the rock give basic friction angles (<^b)
ters: which are converted to via a simple expression
JRC = joint roughness coefficient; self-weight tilt involving the ratio r/R, where r = Schmidt hammer
tests on jointed core or on naturally jointed blocks of rebound on weathered joint, R = Schmidt hammer
rock (Barton and Choubey, 1977). rebound on unweathered rock surface.
Numerical simulation of jointed and faulted rock behaviour for petroleum reservoirs 281

35
(a)
2 34
30
e = 38.3/cm
a.
25

CO
c/) 20 B E = 101.2/
LU e = 42.2 A m
GC
H 15
0)
E = 1 1 1.4/^m
< 10 A
e = 51.1yO.rn

GC

CLOSURE (microns)

-3.0

E -3.5

o
-4.0
>-
-4.5
>
o -5.0
D
O -5.5
z
o
oI
-6.0

o -6.5 h-

NORMAL STRESS MPa

Fig. 5. Non-linear stress-closure-conductivity model (after Barton et al., 1985). The joint is loaded in four cycles to 30 MPa normal
stress (JCSo = 82.5 MPa, JRC0 = 9.0). The fourth cycle is considered to reflect in-situ conditions, (a) Normal stress vs. joint
closure. The mechanical (E) and hydraulic (e) joint apertures are given for 10, 20 and 30 MPa normal stress. As input for the
shear-dilation-conductivity modelling, e-values are used (see Fig. 4). (b) Joint conductivity vs. normal stress. A, B and C reflect joint
conductivities of 10, 20 and 30 MPa normal stress.

L n = in-situ joint length (m). flow (due to stress variation) through rock joints will
These simple tests, which can be performed not be matched by the e = E model of Witherspoon
cheaply on joint samples from drill cores, are de- et al. (1980); a model also used by Pine and Cundall
scribed in detail by Barton and Choubey (1977). (1985) and Harper and Last (1989) in their fractured
reservoir codes.
Coupled stress flow behaviour The simple index tests used to obtain JRC, JCS
gives only a crude estimate of an initial unstressed
Joint apertures strongly depend on the effective aperture (E0). More sophisticated small-scale tests,
normal stress acting across the joint plane. Due to in which flow through a natural rough joint is mea-
the non-linear relation between joint conducting and sured under normal and shear stress, have shown
joint mechanical apertures (see Fig. 6) variation in that two orders of magnitude increase in conductivity
282 A.G. Koestler, A. Makurat, N. Barton andK. Monsen

1mm 0.1mm 0.01mm 0.001mm

Fig. 6. An empirical relationship incorporating joint roughness (JRC) and aperture (E = mechanical aperture; e = hydraulic aperture)
(after Barton, 1982).

(for a granitic joint) can result from a modest 1-2 in practice than those predicted. Some soft cap rock
mm of shear displacement (Makurat, 1985), broadly shales have shown reductions in conductivity with
in agreement with our predictive model (Fig. 4c). shear due to the low values of JCS/an applied (for
Extensive testing of fractures in weaker petroleum example JCS = 10 MPa, = 10 MPa).
reservoir rocks such as sandstones, chalks and cap
rock shales indicates that gouge production may par-
Application of UDEC-BB to faulted block
tially or wholly block the established flow channels,
behaviour
as shear and asperity degradation occur (Makurat et
al., 1990). The results of a coupled shear-flow test Geometry and material properties
(CSFT) on a reservoir sandstone are shown in Fig.
7. Figure 7a shows that the joint dilates over the full In a first step a faulted block configuration was
range of 5 mm of shear. However, as illustrated in chosen which is supposed to represent a border of a
Fig. 7b, only the first 1 mm of shear lead to a major graben, where several fault blocks are located above
increase in joint conductivity. From 2 to 4 mm shear a deep detachment zone (see Fig. 8c). This model
displacement no further increase of k was observed, could be a highly simplified representation of some
and during the last 1 mm even a slight reduction of areas along the eastern border of the Tmpen Spur
k was measured. In all CSFT-tests on softer rocks (see Fig. 8a).
conducted so far, a certain amount of fines could The footwall of the 4 x 3 km model consists of
be observed at the outlet of the test apparatus. It is sandstones with elastic properties similar to the
assumed, however, that most of the fines produced Berea sandstone (see Tkble 1), which is an often
due to joint surface degradation stay inside the joint used standard sandstone in rock mechanics.
and reduce the effective cross-sectional area avail- The fault blocks, separated by normal faults with
able for fluid flow. The key to gouge production dip angles of 70, consist of sandstone with identical
and flow channel blockage (with e < E) is the ra- material properties. They rest on a double layer of
tio JCS/an appearing in eqn. (2). Low values imply shale (for properties see Table 2), of which the lower
greater gouge production, as observed in laboratory one contains an additional low angle joint set.
tests (Makurat et al., 1990). The conductivity curves The extra low angle joint set was thought to ease
for = 30 MPa in Fig. 4c will probably be lower the onset of slip along the shale-sandstone interface,
Numerical simulation ofjointed and faulted rock behaviour for petroleum reservoirs 283

5> s 1(a)
o
E o
O o

if
i_ O
O Z
e
Z I O
LU
s 1

dilation
UJ

o
Q.
CO 1
Q

1 reverse shear ^^^^

DC
o
Z 1 w ^ ^ ^^^TART

\/^^ forward shear

* -5000 -2000 - 000 4000 5000

SHEAR DISPLACEMENT micrometer

E (b)
o
forward shear

START * *

o(0
(0
(0 (0

u. u.
o O
c

-^000 -5000 -2000

SHEAR DISPLACEMENT micrometer

Fig. 7. Coupled shear-flow test on a reservoir sandstone, (a) Normal displacement vs. shear displacement. Forwards and reverse shear
resulted in 150 reduction of fracture aperture, (b) Conductivity vs. shear displacement. Forward shear starts to increase k due to
dilation, but keeps it almost constant for larger shear displacement due to starting gouge production, which is more increased during
reverse shear resulting in reduced k when movement is back to initial position.

and to give additional information on the influence no cohesion and extremely low friction coefficient
of varying dip angles on the direction of the axis of (see Table 3), in order to create low-friction slip
principal stresses. planes.
The gently dipping discontinuities between the All faults between the sandstone blocks are mod-
sandstone and the shale and all joints inside the elled as non-linear BB-joints with proper ties as
shale are modelled as linear joints (joint set 1) with described in Tfable 4, where a JRC value of 4
284 A. G. Koestler, A. Makurat, N. Barton and K. Monsen

(a) NW SE

| GULLFAKS BLOCK

-_LEQCENE

UPPER CRETACEOUS
J \^^^^\ \ \ \ \\
C2^^^vOWER CRETACEOU
s
TAMPEN SPUR \V S*l TRIASSIC \-^_JK^\^/^~

10 km
| AFTER ZCGLER ET AL. 19 NORTHERN VIKING GRABEN

(b)
GULLFAKS FIELD
SKETCH OF FAULT CONFIGURATION

NOT TO SCALE! MAIN BOUNDARY

AFTER KOESTLER ET AL. 1990 FAULT

(C) + 500 m SEDIMENT OVERBURDEN

SHALE*

FRACTURE SET 2 ^

FOOTWALL

Fig. 8. The Gullfaks block as an analogue to the model set up. (a) Line drawing of a seismic section across the Gullfaks Block (after
Ziegler et al., 1986). (b) Configuration sketch of the hanging-wall fault blocks which are separated from the footwall by detachment
zone (after Koestler et al., this volume), (c) Model set up for the numerical simulation. The footwall block is surrounded by roller
boundaries. Left, bottom and right boundaries are roller boundaries during consolidation. Additional load corresponding to 500 m
sediment is applied to the top boundary. During the extension simulation, an outward directed velocity boundary is applied to the
right-hand boundary of the hanging wall.

TABLE 1 TABLE 2

Material properties of Berea sandstone Shale material properties

3
Youngs modulus 5.0 x 10 MPa Young's modulus 5.6 x 104 MPa
Poisson's ratio 0.38 Poisson's ratio 0.31
Bulk modulus 6.9 x 10 3 MPa Bulk modulus 4.9 x 104 MPa
Shear modulus 1.8 x 10 3 MPa Shear modulus 2.1 x 10 4 MPa
Density 2.1 x 10 3 kg/m 3 Density 2.6 x 10 3 kg/m3
Numerical simulation of jointed and faulted rock behaviour for petroleum reservoirs 285

TABLE 3
PRINCIPAL
Linear joint material properties (joint set 1) STRESSES
Normal stiffness 2.6 x 10 4 MPa/m
Shear stiffness 1.3 x 10 3 MPa/m POST-CONSOLIDATION
Cohesion 0.0 MPa/m EXTENSION = 0 m
Tensile strength 0.0 MPa (50 000 cycles)
Friction coefficient 0.01 tangent 500 MPa
Dilation coefficient 0.0 tangent

TABLE 4

Barton-Bandis joint material properties (joint set 2)

EXTENSION = 2.5 m
Joint roughness coefficient JRC 4.0 (55 000 cycles)
Joint compressive strength JCS 10.0 MPa
Intact rock uniaxial compressive strength 20.0 MPa 500 MPa
Initial joint aperture 0.35 mm
Residual friction angle 20

represents a moderate rough joint. A JRC of 1 would

represent a planar fracture, whereas a JRC of 20 EXTENSION = 8.7 m
(67 000 cycles)
would correspond e.g. to an extreme rough tensile
fracture. 500 MPa

Boundary conditions and simulation sequence

y' V . r-.. rv7-
Under consolidation roller boundaries (zero fric- . \ "\- Ax/
:''\., \ \ \ / / \
tion) are applied to the vertical and the lower hori- : ;;V..
zontal boundaries. Gravity is acting on the model and
an additional load of 10.5 MPa is applied along the
top boundary, corresponding to 500 m overburden
^^^^ EXTENSION = 105 m
(259 980 cycles)

500 MPa
(see Fig. 8c).
After consolidation is finished, which means that
all blocks are in a steady-state condition, extensional
tectonics is simulated by applying a small outward-
directed velocity to the upper 2 km of the right model
boundary, thus limiting rollers to the basement sand- EXTENSION = 130 m
stone (see Fig. 9). Table 5 illustrates the major steps (303 143 cycles)
in the simulation sequence.
500 MPa

Fig. 9. Development of principal stresses during consolidation

(A) and increasing extension (B to E).
TABLE 5

Major steps in the simulation sequence

Results and discussion

Simulation steps Figure/Table Three areas of interest have been identified

Set up geometry Fig. 8
in the numerical simulation of a large-scale fault
Define material and joint properties define Tables 1--4 block model:
boundary conditions (1) Stress variation with increasing extension;
Consolidation to in-situ stresses Fig. 9A block rotation;
Evoke extensional tectonics by applying out- Figs. 9B, 10A distribution of fault slip and block deformation.
ward directed velocity to right model (2) Can the simulation be used for the prediction
boundary
of reservoir structuration?
Delete the right side sandstone blocks after Fig. 10C
they lost contact with the model (3) Is the presented large-scale simulation geologi-
cal reasonable?
286 A.G. Koestler, A. Makurat, N. Barton andK. Monsen

PRINCIPAL STRESSES

200 MPa 500 MPa

A B

BLOCK ROTATION
"V
MAX ROTATION
= 4.7
MAX ROTATION
= 4.7
'v\
-A

DISPLACEMENT VECTORS \\
XV^\T
100 m 500 m ^
-

C D

SHEAR DISPLACEMENT

MAX SHEAR MAX SHEAR

DISP. = 25m DISP. = 115m

EXTENSION = 130 m
(303 143 cycles)

Fig. 10. Principal stresses, block rotation and displacement vectors, and shear displacement after maximum extension close to the
detachment zone at the trailing edge (A, C, E) and at the landing edge (B, D, F).

The principle stresses after consolidation (before (Fig. 11A-D). Minor extension (E = 33 m) only
extension) display a sub-regular pattern, with the activates block displacement at the right-hand model
expected stress increase with depth and the main side (e.g. 10-20 m for the diamond-shaped block).
stress axis oriented vertically and horizontally (Fig. Increasing extension affects more and more the
9A). Minor reorientation due to boundary effects blocks towards the left-hand model side. The final
and the influence of shear fracture orientation can stage of simulated extension (E = 130 m) displays
specifically be observed in the lower shale layer. Al- a regular pattern of curved displacement lines indi-
ready minor extension (55,000 cycles, = 2.5 m) cated by the orientation of the displacement vectors.
produces a more heterogeneous stress distribution. The amount of displacement is continuously decreas-
Stress axes are rotated mainly in the right-hand part ing from a maximum of about 130 m in the eastern
of the model, where extension results in unloading part of the model to about 20 m in the westernmost
of the blocks next to the large block, which is pulled block of the hanging wall. The lower part of the
towards the right-hand border of the model. During sandstone blocks and the small shale blocks show
further extension, stress axes are oriented parallel sub-horizontal displacements. Large displacements
and perpendicular to the fault planes in the vicinity can also be observed in the area where fault blocks
of the major faults (see Fig. 9E). Stress axis orien- are bounded by synthetic and antithetic faults (up-
tation in the detachment horizon is highly irregular. permost part of the right-hand model side). The
Within the shales e.g., stresses are oriented parallel block rotation indicated by the block displacement
to the chosen detachment orientation (see for detail vectors is sinistral for all blocks and decreases from
Fig. 10A). Stress concentrations, with increased V/H east to west for the large sandstone blocks, but in-
ratios, can be observed at the end to the left-hand creases for the small sandstone blocks and the shale
side of the detachment zone. blocks close to the detachment (Fig. 12).
The displacement vectors indicate how contact Shear displacements are mainly concentrated on
points between the blocks move during extension the low-friction interfaces within the shale layers and
Numerical simulation ofjointed and faulted rock behaviour for petroleum reservoirs 287

DISPLACEMENT VECTORS SHEAR DISPLACEMENT

EXTENSION = 33 m
(115 012 cycles) EXTENSION = 2.5 m
(55 000 cycles)
1000 m DISPLACEMENT
MAX SHEAR DISP. = 3 m

EXTENSION = 56 m
(159 980 cycles) EXTENSION = 130 m
1000 m DISPLACEMENT (303 143 cycles)

MAX SHEAR DISP. = 264 m H 3 0 0 m

B
Fig. 13. Slip along shear fractures and faults after very little
extension (A), and after maximum extension (B).

EXTENSION = 105 m
(259 980 cycles) Conclusions
1000 m DISPLACEMENT
(1) Since UDEC-BB is a two-dimensional code, all
faults and fractures are assumed to strike perpen-
dicular to the modelled cross section. The Gullfaks
Field is an example with a well-defined slip plane
and little indications for mass transport normal to it
EXTENSION = 130 m
(303 143 cycles) (Koestler et al., this volume) and has therefore been
1000 m DISPLACEMENT chosen for this pilot study.
(2) The results are considered to be preliminary
Fig. 11. Displacement vectors at different simulation stages. since different fault geometries have to be simulated
and an extension to three-dimensional modelling
could drastically improve the applicability of the
BLOCK ROTATION presented method.
(3) The occurrence of big voids along faults with
0.1 increasing extension is not expected to happen in na-
EXTENSION = 2.5 m
(55 000 cycles)
ture, but could indicate dilated areas with improved
MAX ROTATION = 0.15
fluidflowpotential.
(4) The variation and reorientation of stresses
with increasing extension can be investigated in great
detail. Stress concentrations and stress ratios can be
used to identify possible areas for fault nucleation.
(5) High stress concentrations at the western end
EXTENSION = 130 m
(303 143 cycles) of the detachment might favour the generation of
MAX ROTATION = 6.5 new faults, which illustrates the potential for footwall
B collapse towards the hinterland. High stresses close
Fig. 12. Block rotation after very little extension (A), and after to the detachment zone imply the chance for small-
maximum extension (B). scale faulting.
(6) The deformation/displacement and rotational
behaviour of the fault blocks was shown to be highly
partly at the bottom of the small sandstone blocks dependent on block size, shape and fault block con-
(Fig. 13). Little to no shear displacement happened figuration.
between the small sandstone blocks, except in the (7) Little displacement was observed on faults be-
elongation of the major faults. Large displacements tween blocks of equal size. Antithetic faults favour
are also seen on the major faults mainly in the block rotation and block deformation/displace-
area where syn- and antithetic faults interact and ment.
along the fault which bounds the three equal-sized (8) The applied non-linear joint behaviour model
blocks in the west against the more complex area simulates the stress dependent shear displacement-
in the east. dilation behaviour of faults and fractures in a more
288 A. G. Koestler, A. Makurat, N. Barton and K. Monsen

realistic way than linear joint models. The results Symp. Int. Soc. Rock Mech., Nancy, Vol. 1.
from laboratory coupled shear flow tests (CSFT) Cundall, P.A., 1980. A generalized distinct element program for
modelling jointed rock. Eur. Res. Off., U.S. Army.
indicate a strong influence of the stress/strength Fossen, H., 1989. Indication of transpressional tectonics in the
ratio and shear displacement on the fracture surface Gullfaks oil-field, northern North Sea. Mar. Pet. Geol., 6:
properties. These results have to be incorporated 22-30.
into similar future studies in order to evaluate the Gale, J.E., 1982. The effects of fracture type (induced versus
sealing potential of faults as a function of stress and natural) on the stress-fracture closure-fracture permeability
relationship. Proc. 23rd U.S. Symp. Rock Mech., Univ. Calif.,
displacement.
Berkeley, Calif., pp. 290-298.
(9) The presented method in combination with Harper, T.R. and Last, N.C., 1989. Interpretation by numerical
improved input data (e.g. fault surface properties) modelling of changes of fracture system hydraulic conductivity
can be used as a tool evaluating critical processes for induced by fluid injection. Geotechnique, 39(1): 1-11.
reservoir structuration. Further, the applied method Koestler, A.G., Milnes, A.G. and Storli, A., 1992. Complex
hanging-wall deformation above an extensional detachment
could be used as a stress-displacement/deformation
example: Gullfaks Field, northern North Sea. In: R.M.
controlled back-stripping tool, which allows to iden- Larsen, H. Brekke, B.T. Larsen and E. Talleraas (Editors),
tify critical fault configurations and their reaction on Structural and Tectonic Modelling and its Application to
tectonic stresses at different scales. Petroleum Geology. Norwegian Petroleum Society (NPF),
Special Publication 1. Elsevier, Amsterdam, pp. 243-251 (this
volume).
References Makurat, A., 1985. The effect of shear displacement on the
permeability of natural rough joints. Hydrogeology of rocks of
Bandis, S., 1980. Experimental studies of scale effects on shear low permeability, Proc. 17th Int. Congr. Hydrogeol., Tucson,
strength, and deformation of rock joints. Ph.D. Thesis, Dep. Ariz., pp. 99-106.
Earth Sei., University of Leeds. Makurat, A., Barton, N., Rad, N.S. and Bandis, S., 1990. Joint
Bandis, S., Makurat, A. and Vik, G., 1985. Predicted and conductivity variation due to normal and shear deformation.
measured hydraulic conductivity of rock joints. Int. Symp. In: N. Barton and O. Stephansson (Editors), Rock Joints.
Fundamentals of Rock Joints, Bj0rkliden, pp. 269-280. Proc. Int. Symp. Rock Joints, Loen. Balkema, Rotterdam, pp.
Barton, N., 1982. Modelling rock joint behaviour from in-situ 535-540.
block tests: Implications for nuclear waste repository design. Petterson, O., Storli, A., Ljosland, E. and Massie, I., 1990. The
Off. Nucl. Waste Isolation, Columbus, Ohio, ONWI-308, p. Gullfaks Field Geology and reservoir development. In:
96. Buller et al. (Editors), North Sea Oil and Gas Reservoirs
Barton, N. and Choubey, V, 1977. The shear strength of rock II. Proceedings. Graham and Trotman, London, pp. 67-90.
joints in theory and practice. Rock Mechanics, Springer, Pine, R.J. and Cundall, P.A., 1985. Applications of the fluid-rock
Vienna, No. 1/2, pp. 1-54. interaction program (FRIP) to the modelling of hot dry rock
Barton, N., Bandis, S. and Bakhtar, K., 1985. Strength, defor- geothermal energy systems. Int. Symp. Fundamentals Rock
mation and conductivity coupling of rock joints, Int. J. Rock Joints, Bj0rkliden, pp. 293-301.
Mech. Min. Sei. Geomech. Abstr. 22, No. 3, pp. 121-140. Witherspoon, P.A., Wang, J.S.Y., Iwai, K. and Gale, J.E., 1980.
Barton, N., Hrvik, L., Christianson, M., Bandis, S.C., Makurat, Validity of cubic law for fluid flow in a deformable rock
A., Chryssanthakis, P. and Vik, G., 1986. Rock mechanics fracture. Water Resour. Res., 16: 1016-1024.
modelling of the Ekofisk reservoir subsidence. 27th U.S. Ziegler, W.H., Doery, R. and Scott, J., 1986. Tectonic habitat
Symp. Rock Mech., Key to Energy Production, pp. 267-274. of Norwegian oil and gas. In: A.M. Spencer et al. (Editors),
Cundall, P.A., 1971. A computer model for simulating progres- Habitat of Hydrocarbons on the Norwegian Continental Shelf.
sive large scale movements in blocky rock systems. Proc. Graham and Trotman, London, pp. 3-20.

Paper submitted March 1990

A.G. KOESTLER Geo-ReconAIS, Bernhard Herres vei 3, 0376 Oslo 3, Norway

A. MAKURAT Norges Geoteknisk Institutt, EO. Box 40 Tsen, 0801 Oslo 1, Norway
N. BARTON Norges Geoteknisk Institutt, EO. Box 40 Tsen, 0801 Oslo 1, Norway
K. MONSEN Norges Geoteknisk Institutt, EO. Box 40 Tsen, 0801 Oslo 1, Norway
 289

Footwall uplift during normal faulting implications for

structural geometries in the North Sea

G. Yielding and A. Roberts

In recent years, studies of major normal faults in actively extending regions (Aegean, Basin and Range) have documented the
vertical motions associated with normal faulting. In addition to the expected subsidence of the hanging wall, it has been found
that uplift of the footwall occurs during earthquakes on normal faults. This uplift has been modelled as an elastic effect, treating
the fault as a dislocation in an elastic medium. Over geological time scales, however, creep processes below the upper crust act
to restore isostatic equilibrium, and so a purely elastic model is not appropriate. Instead, a flexural model is used, in which the
upper crust is treated as an elastic layer, overlying a "fluid" substratum. A single fault cutting such a layer gives rise to buoyancy
forces that produce reverse drag in both footwall and hanging wall. The wavelength of the reverse drag depends principally on
the effective elastic thickness of the lithosphere. The relative amplitude of the footwall and hanging wall vertical movements is
dominated by the nature of the fill in the half-graben: loading by sediment pushes the whole structure down, whereas loading by
water (sediment-starved) allows more footwall uplift. When a number of faults occur adjacent to one another, their reverse-drag
fields overlap to produce the familiar tilted-block profile of normal-faulted terrains. If the major faults are evenly spaced, dips
within the fault blocks will be relatively uniform and can be approximated by a "domino" model. Such a fault model can
be readily incorporated into a lithospheric stretching model (which causes subsidence of the whole basin), to predict vertical
motions relative to a fixed datum such as sea level.
We apply these concepts to a variety of fault systems in the North Sea and the mid-Norway shelf. Normal faulting in the Late
Jurassic created sediment-starved half-graben with significant bathymetric relief. The footwalls of major platform boundary faults
are typically eroded, often showing post-rift sediments resting on basement (e.g., East Shetland Boundary Fault, Brae Fault). In
these cases, no direct evidence remains of the amount of footwall rotation, so that the eroded crest might be assumed rigid
and erroneously projected into the basin as a regional datum for "section balancing". However, some boundary faults preserve
a condensed footwall sequence, or contain intra-basement reflections, which demonstrate that flexural uplift of the footwall
accompanied normal faulting. The Fr0ya Fault (Haltenbanken) is an example of these. Erosion of marginal footwalls was an
important source of clastic material, and is, therefore, a major factor in the distribution of Upper Jurassic reservoir formations
(e.g., Fr0ya Formation, Brae Formation, Claymore Sst. Member, Fulmar Sand Formation). Within the graben, arrays of fault
terraces can be described by the domino model, leading to the conclusion that the extent of erosion on intra-basinal footwalls is
largely a function of fault spacing. Thus, in the Brent Province, the Snorre fault block suffered severe erosion during the Late
Jurassic extension, whereas on the Brent fault-block erosion was less marked, approximate fault spacings being 30 km and 15
km, respectively. Erosion of intra-basinal footwalls has also been a significant source of Late Jurassic sediment for reservoirs,
e.g., the Magnus member and Munin member.

Introduction scale of study. (For example, when studying the gross

basin architecture, processes of fault seal can be
Deformation in the earth is complex. The rheol- neglected, but isostasy cannot.)
ogy of rocks is dependent on many factors, of which (2) It should make quantitative predictions. The
temperature, pressure, lithology, strain rate and fluid only reason for performing structural modelling in a
content are probably the most important. It is, there- commercial context is to increase the understanding
fore, apparent that no single "model" can describe of an area and thereby to reduce the risk associated
all aspects of earth deformation. If such a unified with its exploration or development. This requires
model were developed, it would be so complicated that quantitative predictions be made of parameters
as to be of little use on a day-to-day basis in hydro- which cannot be observed directly.
carbon exploration. The essential features of a useful (3) It should be simple to use, so that predictions
model are: can be made without spending months performing
(1) It should take into account the significant the calculations.
geological processes operating at the appropriate Clearly, when addressing a particular structural

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 289-304. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
290 G. Yielding and A. Roberts

problem, the appropriate model should be chosen slip produces measurable deformation of the ground
carefully, depending on the questions that need an- surface over an area extending many kilometres from
swering. A good analogy can be made between mod- the fault.
els and maps. Both are representations of physical The best-documented example of the geodetic
reality, but contain a very limited amount of informa- changes associated with major normal faulting is the
tion about the real world. A road map at a scale of 1983 Borah Peak earthquake (Stein and Barrientos,
1:500 000 may be completely adequate to drive from 1985). This magnitude 7.0 earthquake occurred on
one end of a country to the other. However, it would the Lost River Fault in the Basin and Range Province
be of no use in choosing the best climbing route up of the United States. Geodetic, geological and seis-
a particular mountain, for which a large-scale topo- mological data are summarised in Fig. 1, taken from
graphic map would be needed. Similarly, a structural Stein et al. (1988). The Lost River Fault is a major
model that gives good results for a whole basin may normal fault with over 4 km of displacement, sep-
be useless for a single fault zone, and vice versa. arating the Lost River Range from the Thousand
The purpose of this paper is to review briefly Springs Valley (Fig. lb). The focal mechanism and
a number of selected models concerned with nor- hypocentre of the earthquake indicate that the fault
mal faulting and crustal extension. The emphasis is dip is approximately 45 at a depth of approximately
on those models that make predictions about foot- 15 km (Fig. lc). Since the fault dip at the ground
wall deformation by normal faulting. We specifically surface is also approximately 45, the fault is ap-
exclude listric-fault models that assume a rigid, un- proximately planar through the uppermost 15 km of
deformed footwall. Whilst such models may apply the crust. This thickness corresponds to the brittle,
in areas of gravitational collapse of thick sediments seismogenic layer, as shown by the distribution of
(e.g., Gulf Coast, Niger delta), they are of little aftershocks (Fig. lc).
relevance to areas of crustal extension unless weak Figure la presents geodetic data obtained by
decoupling layers such as salt are present. resurveying a levelling line that happened to cross
The models that we consider can be divided into the Lost River Fault. The co-seismic elevation
four groups on the basis of scale: changes (i.e., those that accompanied the earth-
(1) elastic dislocation models, appropriate for quake) amount to almost 1.4 m of subsidence in the
smaller faults and for single slip events on larger hanging wall adjacent to the fault and almost 0.3 m of
faults; uplift in the footwall. Both hanging wall subsidence
(2)flexural/isostaticmodels, appropriate for larger and footwall uplift display a "reverse drag" profile,
faults, i.e., those that create significant topography/ i.e., they exponentially decrease in magnitude with
bathymetry; distance from the fault. The co-seismic deformation
(3) domino models, where major faults are approx- extends approximately 20 km into the hanging wall
imately evenly spaced; and approximately 15 km into the footwall. The ratio
(4) stretching models, which give the gross basin of subsidence to uplift is approximately 4:1. Also
morphology and ignore the individual faults. shown in Fig. la is a model fit to the geodetic data,
After discussing the elastic modelling in terms of derived by assuming the fault to be a dislocation in
presently active faults, we concentrate on flexural/ an elastic medium (Mansinha and Smylie, 1971). The
isostatic and domino models and their application to model parameters were 2 m of slip over a planar
normal faulting in the North Sea and mid-Norway rectangular fault (23 x 18 km) dipping at 47. The fit
shelf. We then discuss some aspects of the three- between model and data is excellent.
dimensional nature of fault geometry, and the influ- A number of important conclusions can be drawn
ence this has on sediment transport routes within an from the Borah Peak example:
extending basin. (1) In normal faulting earthquakes, there is uplift
of the footwall as well as subsidence of the hanging
Active normal faulting wall.
(2) Uplift and subsidence decrease exponentially
Most, though not all, fault movement occurs dur- from the fault.
ing rapid slip events, i.e., earthquakes. Observations (3) A hanging wall rollover does not imply a listric
of earthquakes on active faults can, therefore, con- fault geometry. A planar dislocation embedded in an
tribute a great deal to our understanding of how nor- elastic medium can produce an excellent match to
mal faults and their associated structures develop. the data.
The largest normal-faulting earthquakes result from Similar observations, though less complete, have
a few metres of fault slip, on faults some tens of kilo- been made following other normal-fault earthquakes.
metres long (Jackson and White, 1989). This fault Geodetic profiles for the 1928 Bulgaria earthquake
Footwall uplift during normal faulting implications for structural geometries in the North Sea 291

r
400
T" 1 i i 1 1i
r 1 i 1

w
I

-a ^_ f a ' *" * ~~H
5

LU U- \ -J
O
z
<
X 400 lu 1 -J
o DATA
z
o ~ ~H
\ | - MODEL
<
>
u
w -800 k-

u
\ I -j

\
-1200
l a "j
J_ L_ I I I i 1 _L 1 1
-30 -20 -10 10

1 r
sw Thousand Springs Valley
Borah Peak NE -J
2
z
o
<
>

I 10
l-
Q.
LU
D

15

VERT - HORIZ

20 1L 10
-30 -20 -10 0
DISTANCE NORMAL TO FAULT CKM]

Fig. 1. Geodetic and geological data from the Lost River Fault in Idaho, on which occurred the 1983 Borah Peak earthquake (from
Stein et al., 1988). (a) Coseismic deformation associated with the earthquake, measured by resurveying a levelling line across the fault.
The model curve was calculated assuming a planar fault in an elastic medium. "Kinks" in the model profile result from changes in
azimuth of the levelling route across the section (see Stein et al., 1988, for a detailed map), (b) Geological cross-section across the
Lost River Fault. Qa = Quaternary alluvial fill; v = Tertiary volcanics; dotted area = Palaeozoic and PreCambrian. (c) Seismological
data for the 1983 earthquake sequence. Error bars bracket the main-shock location, with the focal mechanism shown at the hypocentre.
Small circles indicate aftershock locations. The "model fault" is the fault surface used to generate the geodetic curve in (a).
Fig. 2. Seismic line across the Troll East structure (from Roberts and Yielding, 1991). Both the hanging wall and footwall are dominated by reverse drag. Footwall uplift is highlighted by the
divergence between the hydrocarbon-water flat-spot and the top reservoir reflection. (The apparent minor normal drag in the hanging wall adjacent to the fault is a post-faulting compaction effect.)
Footwall uplift during normal faulting implications for structural geometries in the North Sea 293

(Jankhof, 1945; Richter, 1958), the 1954 Fairview son and White (1989), i.e., a "large" fault is one that
Peak earthquake (Whitten, 1957) and the 1959 Heb- cuts right through the seismogenic layer of the upper
gen Lake earthquake (Myers and Hamilton, 1964) crust (approximately 0-15 km). A "small" fault is
all show a pattern of uplift and subsidence similar one that is confined within the seismogenic layer. On
to the Borah Peak earthquake. Vita-Finzi and King the time scale of a single earthquake (20 s or less),
(1985) report a footwall uplift proportion of approx- the whole of the crust and mantle is an elastic solid,
imately 10% (0.2 m in a total throw of 1.7 m) for which is why elastic models account so well for the
the 1981 Corinth earthquakes which occurred on co-seismic deformation (Fig. 1). On a longer time
near-planar faults dipping at approximately 42. (No scale, however, the mantle and lower crust (beneath
geodetic profile crosses this fault system; uplift and the seismogenic layer) deform by thermally activated
subsidence were measured locally relative to the sea creep processes (e.g., Brace and Kohlstedt, 1980;
surface.) Kusznir and Park, 1987). Their response is more akin
The geological structure around faults is built to a viscoelastic solid, i.e., they show an instanta-
up through many seismic cycles. Each earthquake neous elastic response and then exponential relax-
represents one small increment of slip on the fault ation by viscous flow. Therefore, after an earthquake
surface. Thus, to some extent the geological structure on a large normal fault, there is a slow rebound to a
can be modelled by summing together the effects of new equilibrium position. This post-seismic rebound
many successive earthquakes (e.g., King and Vita- takes the form of a broad up-doming centred on the
Finzi, 1981; Jackson et al., 1982). fault (but with no further fault slip). Modelling of
Although the co-seismic deformation is most read- post-seismic movements (e.g., Rundle, 1982; Cohen,
ily measured around large faults, it is believed that a 1984) suggests that the rebound is complete within
similar deformation field is produced by seismic slip approximately 1000 years of an earthquake, with
on small faults. As a small fault nucleates and grows, movement being fastest immediately after the seis-
the geological structure that forms around it will be mic slip. This is less than or comparable to a typical
the cumulative result of all the slip events during its earthquake recurrence time (e.g., Wallace, 1984), and
growth history. Gibson et al. (1989) used an approx- so the rebound following one earthquake is probably
imation to the elastic dislocation model to describe completed before the next occurs. Reilinger (1986)
bed geometry adjacent to evolving planar normal documents approximately 30 cm of uplift in the 24
faults. They used a realistic displacement distribu- years following the 1959 Hebgen Lake earthquake
tion on the fault surface, and obtained a somewhat (approximately 7 m co-seismic throw); post-seismic
tighter "reverse drag profile" than that seen in the rebound here is presumably far from complete.
earthquake geodetic data. The importance of post-seismic movement is that,
Such reverse drag, on both sides of a normal for large faults cutting the upper crust (seismogenic
fault, is very commonly observed in geological cross- layer), the long-term (geological) structure is not
sections. A number of examples are given by Barnett simply the sum of all the earthquake slip events.
et al. (1987) and Gibson et al. (1989), from both More appropriate models for this scale of faulting
mine plans and seismic reflection data. Another case are discussed in the next section. However, for small
where the footwall reverse drag is particularly clear faults, (i.e., those which produce the internal defor-
is in the eastern part of the TYoll Field on the mation of blocks bounded by large faults), elastic
Horda Platform of the northern North Sea (see Fig. dislocation models appear to work well. This is be-
2). Here, a flat internal datum is present within cause small faults do not reach the zone where creep
the fault-block in the form of the seismic "flatspot" becomes significant. Beds cut by such faults show a
produced by the hydrocarbon/water contact (Birtles, local reverse drag profile consistent with the accumu-
1986; Hellem et al., 1986). Reflections from the lated earthquake slip on the fault surfaces (Gibson et
Viking Group reservoir show the steepest dip to be al., 1989).
adjacent to the main fault, decreasing to zero at
a distance of approximately 8 km. Footwall uplift Flexural/isostatic models
above the flatspot datum is approximately 200 m at
the fault. The term flexure is used to describe the bending
Using an elastic dislocation model to describe the of the uppermost layer of the Earth (the seismo-
geological deformation around faults is only strictly genic layer, or "elastic lid") in response to loads. A
appropriate if the co-seismic deformation is the dom- rapidly imposed load will not be in isostatic equilib-
inant structural process. This is probably so for small rium. On a geological time scale thermally activated
faults. Here, we are using the distinction between flow at depth will act to restore isostatic equilibrium.
"small" and "large" faults in the same sense as Jack- Only the relatively cool seismogenic layer retains sig-
294 G. Yielding and A. Roberts

HALF-GRABEN (-ve load)4 MOUNTAIN (+ve load)! King et al., 1988; Fowler and McKenzie, 1989). It
deposition (+ve load) \ erosion (-ve load) t
is not yet clear whether a thin, truly elastic, plate
persists in the upper crust, or whether the apparent
rigidity is diminished over a broad depth range (so
Elastic/brittle that a thin layer is simulated). In either case, the
upper crust
effective elastic thickness is a parameter that is esti-
mated by comparing observed structural wavelengths
with those predicted by modelling. Thus, the wave-
Ductile lower crust lengths (approximately 15 km) of the reverse drag
MOHO
deformation around major normal faults are consis-
tent with an effective elastic thickness of less than 5
Fig. 3. Schematic diagram illustrating the isostatic loads gen-
erated by a large normal fault. Uplift of the footwall and
km (e.g., King et al., 1988; Marsden et al., 1990).
subsidence of the hanging wall generate positive and negative The relative amplitude of the footwall and hanging
loads, respectively. Erosion of the footwall, and sediment de- wall vertical movements is dominated by the sedi-
position in the hanging wall graben, act to reduce these initial ment loading in the half-graben. A sub-aerial fault
loads. The isostatic loads are supported by the flexural rigidity
of the elastic/brittle upper crust (seismogenic layer). Modified system with relatively little sediment input will have
from King et al. (1988). approximately equal uplift and subsidence (King et
al., 1988). This is the case for the Lost River Fault
shown in Fig. 1 (Stein et al., 1988), though the co-
nificant strength. The mathematical model used to
seismic movements were unequal. If the footwall is
investigate flexure is that of an elastic layer (repre-
subaerial and the hanging wall covered by sediment,
senting the upper crust) overlying a fluid substratum
the footwall uplift is only 5-25% of the total throw
(representing the lower crust and mantle). (See Tur-
(cf. "circa 10%" quoted by Jackson and McKenzie,
cotte and Schubert 1982 for a general discussion of
flexural models.) 1983). More footwall uplift occurs if the hanging wall
is loaded mainly by water (i.e., a sediment-starved
Figure 3 shows how large normal faults generate
loads that in turn cause flexure. Co-seismic uplift of basin), since water is less dense than sediment. This
the footwall creates a topographic load at the earth's has important implications for Late Jurassic faulting
surface, which in turn will tend to cause post-seismic in the North Sea, since sediment input was rela-
subsidence. This load may be reduced by erosion of tively small and water depths in the half-graben were
the footwall. Similarly, the half-graben is a "hole", or substantial (Bertram and Milton, 1989).
negative load, at the earth's surface, and will tend to In the flexural model illustrated by Fig. 3, de-
cause post-seismic uplift. Deposition of sediment will formation is measured relative to the undeformed
tend to reduce this negative load. However, since the surface at a distance from the fault. In areas near
half-graben is much bigger than the uplifted footwall sea level it is often useful to measure (and predict)
(cf. Fig. 1), post-seismic movement around normal the deformation relative to sea level, as a fixed da-
faults is dominated by uplift. tum. It is then necessary to include the effects of
basin subsidence in addition to the more local move-
The resistance of the elastic layer to bending is
described by its flexural rigidity. The flexural rigidity, ments adjacent to the faults. The basin subsidence
D, is given by: is mainly determined by the distributed extension
throughout the lithosphere column (cf. McKenzie,
1978). Kusznir et al. (1988) and Kusznir and Egan
D = (1990) have developed a model of coupled simple
12 (1 - i/2)
where Te is the thickness of the elastic layer, E is
Young's modulus, and v is Poisson's ratio. However,
the earth does not have a well-defined elastic layer
with a sharp base, overlying a fluid. The elastic thick- Upper Crust X ^ |e Britt

ness of the lithosphere is, therefore, usually referred Lower Crust Q _ oh

M Ductile
to as an apparent or effective elastic thickness, cor-
responding to a perfect elastic layer with the same
response as the earth. Although the seismogenic Fig. 4. The "flexural cantilever" model of Kusznir et al. (1988).
layer is typically about 10-15 km thick (e.g., Chen Extension is accomplished by simple shear on a major fault(s)
and Molnar, 1983), the effective elastic thickness of in the upper crust, and by pure shear spread over a width W
in the lower crust and mantle. The isostatic loads generated by
continental lithosphere is generally estimated to be the extension are supported by the flexural rigidity of the upper
5 km or less (Barton and Wood, 1984; Watts, 1988; crust.
Footwa uplift during normal faulting implications for structural geometries in the North Sea 295

shear and pure shear, in which lithospheric exten- erosion down to sea level. The section with erosion
sion is achieved by major planar faults in the upper matches the observed profile very well, with 1500 m
crust and by distributed pure shear in the lower of erosion adjacent to the fault.
crust and mantle. This "flexural cantilever" model An important aspect of the model profile is the re-
is shown schematically in Fig. 4. Distributed exten- turn of the original surface below sea level east of the
sion in the lower crust and mantle is made equal to erosion zone. A potential depocentre is, therefore,
the fault-related extension in the upper crust. The predicted for some of the erosion products from the
flexural-isostatic response to upper crustal faulting, uplifted footwall. On the depth profile of Fig. 5, this
pure shear at depth, and sedimentary loading, is depocentre corresponds to the Draugen Field, where
calculated in the wavenumber domain. Upper Jurassic sands (Fr0ya Formation) form the
reservoir (Ellenor and Mozetic, 1986). These sands
Application of flexural/ isostatic models are probably the erosion products of the emergent
We have used the flexural cantilever approach high immediately to the west. Calibration of flexural
of Kusznir et al. (1988) to model both individual models against known geology (such as Fig. 5) then
structures and complete basin cross-sections in the enables us to use the models in a predictive way,
North Sea/mid-Norway area. Detailed discussion is where the syn-rift sands have yet to be confirmed by
given in Roberts and Yielding (1991) and Marsden drilling.
et al. (1990), and only brief examples are given here. The largest faults in the northern North Sea,
Figure 5 is a depth section across the western such as the Brae Fault and the boundary fault to
boundary fault of the Tftndelag Platform (mid- the East Shetland Platform, typically have severely
Norway shelf), in the vicinity of the Njord and eroded footwalls with post-rift sediments resting on
Draugen Fields (Ellenor and Mozetic, 1986). The basement (see fig. 3 of Roberts and Yielding, 1991).
Halten Tferrace lies immediately to the west. Seis- In these cases, no direct evidence remains of the
mic markers in the platform are tied to wells in the amount of footwall erosion. However, flexural mod-
Draugen Field and elsewhere. Within 10 km of the elling of these fault systems predicts that several
boundary fault erosion of the THassic-mid-Jurassic kilometres of section may have been eroded from
sequence has occurred. The erosion surface (top of the footwall. It would therefore be wrong to as-
the platform) shows little relief, but the underly- sume that the flat crest of such a footwall represents
ing sequence has been progressively uplifted as the a stable, rigid high; this crest cannot be projected
boundary fault is approached. This uplift has been into the basin as an undeformed regional datum for
accomplished by a series of normal faults, antithetic geometric "section balancing".
to the main structure. Using the regional dip of Marsden et al. (1990) have applied the flexural
the intra-THassic salt marker in the "undeformed" cantilever model to a complete basin cross-section
eastern part of the section, a "smoothed" footwall across the Viking Graben. A reasonably good fit
uplift profile can be drawn through the small fault to the observed structural geometry and sediment
blocks. Maximum uplift at the fault is then esti- thicknesses can be achieved by using observed fault
mated at approximately 1500 m. The age of the heaves and an effective elastic thickness of 3 km.
uplift is constrained as between mid-Jurassic and Uplift and erosion of fault-block crests throughout
mid-Cretaceous. The main period of faulting was the basin is predicted, in good agreement with de-
Late Jurassic, and thus we conclude that the footwall tailed well and seismic data. Similar estimates of
to the basin-margin fault zone became uplifted and footwall uplift on intra-basin fault blocks can also be
emergent during extension. achieved using a simpler, geometric approximation,
Figure 6 shows two model cross-sections (obtained as described in the next section.
using the flexural cantilever model) for the fault zone
shown in Fig. 5. Three kilometres of extension has Domino models
been used, on a fault with an initial dip of 50 (the
fault rotates to a shallow dip during extension). The The flexural cantilever model, discussed in the pre-
basin fill has been given an average density of only vious section, considers the flexural deformation of
2.0 g cm -3 , since the Upper Jurassic sediments do arrays of fault blocks within the framework of basin
not fill the half-graben and substantial water depth subsidence caused by lithospheric extension. How-
must have been present during rifting. An effective ever, even with low values for the effective elastic
elastic thickness of 3 km was used, to match the thickness, the wavelength of the flexural deformation
observed radius of reverse drag on the depth profile. is 15-20 km. Normal faults rarely occur in isolation,
The upper section in Fig. 6 was modelled with no and so each fault is generally within the reverse drag
erosion, whereas the lower section has complete radius of adjacent faults. The result of this superpo-
296 G. Yielding and A. Roberts

Depth Draugen Field (proj.)

Erosion Pinchout of i - Seismic onlap of Lower Cretaceous

Tertiary intoTriassic Upper Jurassic

Smoothed footwall uplift profile

m Lower Cretaceous
Projected reference datum
rn Upper Jurassic ^Marginal fault,
E3 Middle & Lower Jurassic predicted dip c.45
Trondelag Platform
mm 'Upper ' Triassic
No vertical exaggeration
'Intra' Triassic

Fig. 5. Depth section across the Tr0ndelag Platform/Halten Terrace basin margin (mid-Norway Shelf), from Roberts and Yielding
(1991). Note the 1500 m of erosion into the footwall of the main fault. The position of the Draugen Field is projected from
approximately 10 km off the profile.

a) No erosion

F = f o o t w a l l emergence F
= 750m
H = hangingwall subsidence H
= 2400m
r = reverse drag radius, above datum r
= 1 6km
e = e x t e n s i o n on marginal fault e
= 3km

b) Complete erosion
Eroded footwall

hangingwall subsidence
uplift of basin floor as a result of
footwall e r o s i o n
a p p r o x i m a t e m a x i m u m e r o s i o n of 1 500m
f o o t w a l l (= footwall uplift), F + u = F e

2 km V:H = 4:1 In both models, elastic thickness = 3km

initial fault dip = 50
| 25km initial crustal thickness = 35km

Fig. 6. Rexural-cantilever models of the basin-margin fault of Fig. 5. (a) is without erosion, (b) is with complete erosion to sea level.
Note the 1500 m of footwall erosion predicted by model (b). From Roberts and Yielding (1991).

sition of reverse drag fields is that each block has The appeal of the domino model is its extremely
a relatively uniform dip from one fault to the next simple geometry. Calculations of extension are,
(Fig. 7). This geometry has long been recognised therefore, very straightforward: the stretching fac-
from observations of normal-faulted terrains, and is tor is given by:
generally described by the "rotating domino" model
(e.g., Ransome et al., 1910, Wernicke and Burchfiel, sin#
=
1982, Jackson and McKenzie, 1983). In its simplest sin(0 - )
form, all the fault blocks are the same size, all the where is the original fault dip and is block tilt
faults have the same dip and all the blocks and faults (see Fig. 8). Moreover, the fault-block geometry is
rotate as the array is extended (see Fig. 8). very similar to the much more complicated approach
Footwa uplift during normal faulting implications for stmctural geometries in the North Sea 297

0 km 50
is taken as completely rigid within a fault block,
No Vertical Exaggeration
Airy (local) isostasy is assumed on a length scale
greater than one fault block. The equations relating
fault-block rotation to stretching, and subsidence to
stretching, are readily combined, so that the move-
ment of fault-block crests relative to sea level can be
Fig. 7. Flexural-cantilever model of array of planar faults. Note calculated. Jackson et al. (1988) derived a similar set
that the fault blocks within the basin adopt the characteristic of equations, but accounted for sediment loading in
"domino" form typical of extended terrains. Each fault block is a more explicit (and less approximate) way.
rotated by uplift at one bounding fault and subsidence at the
other.
Application of domino models
The northern North Sea is a good place to test
the domino model because of the abundance of
seismic data, and well data at footwall crests. Fault
blocks in the East Shetland Basin have been mod-
elled as rotating dominoes by Barr (1987a), Yielding
(1990) and White (1990), though with somewhat dif-
fering interpretations of the overall amount of Late
Fig. 8. The "rigid domino" model of continental extension.
All fault blocks and faults rotate simultaneously as extension
Jurassic extension. In the East Shetland Basin, it is
proceeds. The stretching factor = [sin#]/[sin(# - )], and the established that the basin floor was approximately at
fault throw F = ds'm^f. The average level of the basin floor sea level prior to the Late Jurassic stretching because
undergoes a tectonic subsidence S{ that is a function of , and the Middle Jurassic Brent Group contains coals, and
there is a further subsidence, , caused by sedimentary loading.
Footwall uplift U is given by U = F/2 - S\ - . See Jackson et blankets the basin (Brown, 1986, Bertram and Mil-
al. (1988) for further details. ton, 1989). Well data and depth profiles can be used
to calculate the amount of section lost from footwall
crests by erosion and degradation, as indicated in
of modelling the flexure associated with each fault. Fig. 9. This in turn can be used as an approximate
Provided that an array of major planar faults of estimate of the amount of footwall uplift above sea
roughly equal spacing exists, the domino model is level, i.e., relative to the "pre-rift" position of the
a good first-order approximation to the fault-block basin floor.
topography. The fault blocks are assumed to be In cases where the crestal erosion surface is rela-
perfectly rigid, so that the upper surface of each tively flat (Fig. 9a), it is reasonable to assume that
block is straight and undeformed. this represents erosion near sea level and, therefore,
A common criticism of the domino model is that a the eroded section equals the amount of uplift above
space problem is created at the keels of the rotating sea level. The crest of the Snorre Field has this ge-
blocks. However, it must be remembered that we ometry, with Lower Cretaceous limestones covering
are considering large normal faults that cut right eroded Triassic sandstones and siltstones (Hollander,
through the brittle (seismogenic) upper crust. The 1987). In most cases, however, the crest is more
transition from the brittle upper crust to the ductile complicated, as shown in Fig. 9b; the Brent Field is a
lower crust is probably temperature controlled. If published example with this type of profile (Bowen,
the keels of the blocks protrude into the underlying 1975; Livera and Gdula, 1990). At Brent there is an
ductile material, conduction of heat will cause the east-dipping surface between the present structural
temperature structure to re-equilibriate very quickly crest and the east-dipping main fault. This surface
(Jackson et al., 1988). The "brittle-ductile transition" is interpreted by Livera and Gdula (1990) as a sur-
will, therefore, not be significantly perturbed by block face of submarine degradation of the active fault
rotations. scarp, with considerable slumping of footwall ma-
The domino model describes the geometry of the terial along low-angle faults. The present structural
upper-crustal extension. Barr (1987a, b) was the first crest is generally overlain by a thin cover of marine
to incorporate this fault model into a framework of shales (Late Jurassic Kimmeridge Clay Formation).
lithospheric extension. He used the simple McKenzie The structural crest was, therefore, below sea level
(1978) model of uniform stretching to define the at the end of fault-block rotation (end Jurassic). A
overall tectonic subsidence of the basin floor. The maximum estimate of the amount of footwall uplift is
domino fault blocks are then assumed to be the up- given by the elevation diiTerence between the present
per layer of this uniform stretching, i.e., they share crest and the projected footwall cutoff at Top Mid-
the same factor. Thus, although the upper crust dle Jurassic level (Fig. 9b). Note that the depth of
298 G. Yielding and A. Roberts

Projected
footwall cutoff
the rock types at the block crest. At the Brent Field,
(Top M. Jurassic) submarine degradation and slumping of relatively
unconsolidated Middle and Lower Jurassic sands
and clays might well be expected at the fault scarp.
At Snorre, however, there was much more footwall
uplift, so that erosion had cut right down into the
TTriassic section. It is likely that the TViassic sand-
stones and siltstones were already more compacted
and cemented than the younger Jurassic sediments.
Projected Therefore, the main fault scarp would have been
-^" footwall cutoff
(b)
(Top M. Jurassic) more resistant to slumping, and erosion at the wave
Sea-level ..1 ^
base was the dominant process for removing footwall
material.
^~~~~Z^^ Collapse of
poorly-consolidated
fault scarp
Figure 10 compiles footwall erosion estimates from
many fields in the northern North Sea, taken mainly
from published profiles. Estimates were made as
Fig. 9. Schematic erosion profiles across North Sea fault blocks,
showing how footwall uplift, U, relative to sea level can be
outlined above, so that we are attempting to measure
calculated. (The depositional surface of the Middle Jurassic the amount of footwall uplift relative to sea level.
Brent Group was originally very near to sea level.) In (a), the These estimates of uplift are plotted against the
footwall crest represents an erosion surface. In (b), significant present size of the fault block (i.e., distance between
degradation of the submarine fault scarp has occurred, increasing
the "depth of erosion" into the footwall.
major faults, measured across strike). There is a clear
correlation between amount of footwall erosion and
fault-block size. Small blocks such as Thistle and
erosion immediately adjacent to the fault will over- Murchison show negligible crestal erosion, whereas
estimate the amount of uplift relative to sea level, large blocks such as Gullfaks and Snorre are severely
since this degradation surface was always below sea eroded (approximately 1 km of section lost).
level. This relationship corresponds closely to that ex-
The difference in profile between the Snorre (cf. pected from the domino model. A given extension
Fig. 9a) and Brent (Fig. 9b) Fields may be related to can be accomplished by a few widely spaced large

1.2H M.2

o 1.0H M.O T3

Q.

Q.
0.8 H 0.8 fj

T3

0.6 0.6
<D

o
Claymoi
" 0.4 h0.4 <=-
3 Statfj
/ Piper Ninian
N. A l w y n 0 s e b e r 9
0.2 H V. Troll 0.2

Heather
Thistle Murchison
-1 ~ i
10 20 30

Present width of fault block (km)

Fig. 10. Graph showing the "observed depth of erosion" at thirteen footwall fields in the North Sea, measured as shown by U in Fig. 9.
Based on depth profiles given by Spencer and Larsen (1990), Maher and Harker (1987) and Yielding (1990). The sloping lines show
predicted footwall uplift from the domino model, for two different sets of input parameters. Note that all the observed erosion profiles
can be accounted for by a small range of variation in initial fault dip and fault-block tilt.
FootwaU uplift during normal faulting implications for structural geometries in the North Sea 299

faults, or by a greater number of smaller faults Three-dimensional aspects of uplift and

(less throw). The larger fault-blocks are, therefore, subsidence
expected to be bounded by the larger faults and,
therefore, to have the greatest amount of footwall In the previous sections uplift and subsidence have
uplift and erosion. The sloping lines on Fig. 10 show been discussed in terms of sections perpendicular
two domino-model predictions of the relationship to the faulting, i.e., a two-dimensional approach. In
between block size and footwall uplift, given slightly reality, fault displacements vary along strike and,
different fault dips and block tilts. The upper line therefore, the associated uplift and subsidence will
is for an initial fault dip of 60 and a block tilt of also vary. The topography/bathymetry of the basin
8, the lower line is for an initial fault dip of 45 floor will be strongly controlled by the interaction of
and a block tilt of 6 (both correspond to a value the three-dimensional deformation fields around the
of about 1.1). These values are reasonable bounds faults.
on the observed structure for most of the fields Real faults have a finite length, and often end lat-
shown on Fig. 10. Footwall uplift was calculated erally at a point of zero displacement (e.g., Barnett et
using the expressions of Jackson et al. (1988), using al., 1987). The maximum displacement is at or near
a basin-fill density of 1.65 g cm"3. As argued by the centre of the fault, and will be associated with
Yielding (1990), this low value is reasonable, since the maximum amounts of footwall uplift and hang-
a significant part of the basin-fill was water (cf. ing wall subsidence (Jackson and McKenzie, 1983;
Bertram and Milton, 1989). The two calculated lines Gibson et al., 1989). Figure 11 shows schematically
bound the spread of observed amounts of erosion. the influence of a single large normal fault on a
Thus, the domino model appears to work well for single horizon. The decrease in footwall uplift from
predicting the amount of footwall uplift in North Sea the fault centre to the lateral tips gives rise to the
fields. characteristic shape of the structural closure that is
The advantage of the domino model is its simplic- observed on many footwall fields (see compilation by
ity, whilst still making useful quantitative predictions. Spencer and Larsen, 1990).
It is not applicable to isolated normal faults, or If the horizon shown in Fig. 11 is considered as
basin-margin faults. In its simple form, it specifically the basin floor, the tilt on that horizon will con-
excludes any description of the internal deforma- trol the local sediment transport direction. For large
tion of fault blocks, which may locally be significant. fault throws, material may be eroded from the area
Modifications of the domino model have been pub- of maximum footwall uplift. Transport of this ma-
lished which incorporate block deformation by in- terial may be predominantly down the dip-slope of
clined simple shear (Jackson and White, 1989; Faure the footwall, rather than directly over the fault-
and Chermette, 1989). However, a more realistic ap- scarp (Leeder and Gawthorpe, 1987). Such transport
proach will probably require quantitative assessment directions are shown schematically on Fig. 11. Devel-
of the population of smaller-scale faults that must opment of new, syn-rift, reservoir sediments requires
inevitably accompany the large block-bounding faults the presence of suitable material within the uplifted
(Childs et al., 1990). Such models are referred to footwall (e.g., new sands are unlikely to be generated
as "soft-domino" models (e.g., Walsh and Watterson, if the footwall rocks are predominantly shaly). The
1991). Assuming that the dominoes are rigid will Upper Jurassic sands of the Draugen Field (Fig. 5)
always underestimate the amount of extension. were probably derived from the adjacent footwall
Notwithstanding these limitations, the rigid- in this way. Roberts and Jackson (1991) document
domino model can be useful in undrilled areas to
assess the prospectivity of intra-basinal fault blocks.
Fault-block spacing is generally readily measured
on regional seismic profiles, and we can then use the
domino model to make predictions about the amount
of uplift or erosion on blocks of particular sizes. For
example, blocks above a particular threshold size
may have no chance of preserving a particular pre-
rift reservoir within structural closure at their crests.
In more equatorial climates, reef development near
footwall crests may be enhanced for particular up-
lift histories, e.g., those which remain emergent for Fig. 11. Cartoon illustrating how the amount of footwall uplift
varies along a single major fault. The arrows indicate the
the longest period (see Fig. 4 of Jackson et al., expected sediment transport away from the erosion zone at the
1988). footwall crest.
300 G. Yielding and A. Roberts

The usefulness of the ideas discussed in this pa-

per can be seen from Fig. 13, which summarises
the distribution of known Upper Jurassic ("syn-rift")
sands in the northern North Sea. (This map will un-
doubtedly be proved incomplete by further drilling
or release of presently confidential wells.) All of the
known sand development is controlled to a greater
or lesser degree by the Late Jurassic normal fault-
ing. The Viking Group sands on the Norwegian
Fig. 12. Cartoon illustrating an en echelon zone between two
Horda Platform were probably sourced by regional
major normal faults. The "relay ramp" or transfer zone between
the two faults necessarily has a lower footwall elevation than uplift of Norway, but their route into the basin ap-
the central parts of each fault. Therefore, it acts as a sediment pears to have been structurally controlled (Badley
transport route from the footwalls to the hanging wall basin. et al., 1988). In the north part of the East Shetland
Syn-rift sands will be preferentially deposited near these transfer
zones.
Basin, both the Munin sand member and the Mag-
nus Sandstone Member have been derived by erosion
of uplifted intra-basinal footwalls. The Munin sand
similar controls on sediment transport in the present member (Gradijan and Wiik, 1987) seen in the Stat-
fluvial systems of central Greece. fjord Nord Field is probably derived at least in part
However, there are cases in which the majority of from the eroded Snorre block immediately to the
the eroded footwall material is transported directly east. The Magnus Sandstone Member was derived
on to the hanging wall rather than down the dip- from the eroded footwall(s) to the northwest of
slope of the footwall as shown in Fig. 11. Where the the Magnus Field (Young, this volume). Minor de-
footwall rocks are only poorly consolidated, failure velopments of syn-rift sands occur adjacent to the
of the fault scarp may be the main control on the main boundary fault to the East Shetland Platform
removal of material from the footwall crest (see Fig. (e.g., Wheatley et al., 1987) and are concentrated at
9b). This is the case on the Brent Field (Livera and discontinuities (relay zones?) in that fault system.
Gdula, 1990) and many of the other major fields in The most prolific depocentres for Upper Juras-
the East Shetland Basin. In these examples, most of sic sands have been those adjacent to the Fladen
the reworked debris from the footwall is found on Ground Spur (Fig. 13). The South Viking Graben
the hanging wall side of the main fault. is essentially a single half-graben with a very large
A common feature of large normal faults is that bounding fault at its western side, with at least 7
they are rarely continuous along strike for more than km of displacement (Harris and Fowler, 1987, fig.
about 20 km (Jackson and White, 1989). This is 4). The Fladen Ground Spur is the footwall to
most readily documented for onshore fault systems this major fault. The well-known Brae Formation
for example in the Basin and Range Province (e.g., constitutes a complex set of submarine fans in the
Schwarz and Coppersmith, 1984; Bruhn et al., 1987). hanging wall of the main fault (e.g., Turner et al.,
Individual fault segments are often related in an en 1987, Harris and Fowler, 1987). Jackson and White
echelon fashion, with transfer zones or "relay ramps" (1989) have suggested that this fan development in
(Larsen, 1988) between them. Physical linkage of the the hanging wall may have been controlled by relay
two faults by a single transfer fault is not particularly zones, i.e., where major fault segments die out or
common, though is known in some cases. The im- change strike. Flexural/isostatic modelling of such
portance of the en echelon geometry is that it allows major faults indicates that several kilometres of ero-
sediment input into the hanging wall basin (see Fig. sion may have occurred in the immediate footwall
12). As the footwall uplift on each fault segment (see earlier section of this paper). The Brae fans are
decreases towards its tip in the relay zone, this step unlikely to account for all of this material, and we
in the fault pattern must inevitably be one of the would suggest that a large proportion of the sediment
lowest points along the footwall ridge. This structural was shed southwestwards from the Fladen Ground
saddle, therefore, attracts the hinterland drainage Spur, into the Witch Ground Graben (see Fig. 13).
or sediment transport developed on the dip-slope The deltaic Piper Formation blankets much of the
of the footwalls, and channels it into the hanging Witch Ground Graben, and like the Brae Formation
wall basin. Steps and discontinuities in fault patterns its development commenced in the Late Oxfordian.
have long been recognised as sediment entry points Stratigraphic cross-sections through the Piper For-
(e.g., Leeder and Gawthorpe, 1987), and Roberts mation (Boote and Gustav, 1987, figs. 7-9) indicate
and Jackson (1991) demonstrate this in detail in sediment transport from the northeast, i.e., from the
central Greece. Fladen Ground Spur. Later sands, e.g., the Volgian
Footwall uplift dunng normal faulting implications for structural geometries in the North Sea 301

Fig. 13. Map illustrating the main developments of "syn-rift" (Upper Jurassic) sands in the northern North Sea, based mainly on
Spencer and Larsen (1990). Note the importance of footwall erosion in providing sediment sources, and the control exerted by fault
geometry on the sediment route to the basins.

turbidite systems, show derivation from the south- the Upper Jurassic sands of the northern North Sea,
west (Claymore), as well as the northeast (Galley), as the depositional pattern of the Fulmar Formation
footwall uplift on the Halibut Horst developed. is closely controlled by the major fault systems (see
Further south in the Central Graben, the main Roberts et al., 1990, for a more detailed discus-
"syn-rift" reservoir is the Fulmar Formation. Like sion).
302 G. Yielding and A. Roberts

Conclusions Boote, D.R.D and Gustav, S.H., 1987. Evolving depositional sys-
tems within an active rift, Witch Ground Graben, North Sea.
(1) Uplift of the footwall is the rule during normal In: J. Brooks and K. Glennie (Editors), Petroleum Geology
faulting associated with crustal extension. of North West Europe. Graham and Trotman, London, pp.
(2) Modelling of faults as dislocations in an elastic 819-828.
medium is appropriate for "small" normal faults (the Bowen, J.M., 1975. The Brent Oil-field. In: A.W Woodland
(Editor), Petroleum and the Continental Shelf of Northwest
"production scale"), and for co-seismic (earthquake) Europe. Applied Science Publishers, London, pp. 353-360.
slip on "large" normal faults. For geological mod- Brace, W.F. and Kohlstedt, D.L., 1980. Limits on lithospheric
elling of the "large" basin-forming faults (the "ex- stress imposed by laboratory experiments. J. Geophys. Res.,
ploration scale"), flexural/isostatic models are more 85: 6248-6252.
appropriate. The domino model can be used where Brown, S. 1986. Jurassic. In: K.W. Glennie (Editor). Introduction
large faults are evenly spaced. to the Petroleum Geology of the North Sea. Blackwell
Scientific Publications, Oxford, 2nd ed., pp. 133-159.
(3) Footwall ("pre-rift") reservoirs generally show
Bruhn, R.L., Gibler, PR. and Parry, W.T., 1987. Rupture char-
a reverse drag profile, often modified by erosion or acteristics of normal faults: an example from the Wasatch
degradation at the crest. fault zone, Utah. In: M.P. Coward, J.F. Dewey and PL. Han-
(4) Eroded footwalls, both at basin margins and cock (Editors), Continental Extensional Tectonics. Geol. Soc.
within the basin, are an important sand source for London, Spec. Publ., 28: pp. 337-353.
"syn-rift" reservoirs. Chen, WP. and Molnar, P., 1983. Focal depths of intracontinental
(5) Sand transport routes during rifting are and intraplate earthquakes and their implications for the
thermal and mechanical properties of the lithosphere. J.
strongly controlled by the basin morphology, which
Geophy. Res., 88: 4183-4214.
in turn is controlled by footwall uplift and hanging Childs, C , Walsh, J.J. and Watterson, J., 1990. A method for
wall subsidence. estimation of the density of fault displacements below the
limit of seismic resolution in reservoir formations. In: A.T
Acknowledgements Butler et al. (Editors), North Sea Oil and Gas Reservoirs.
Graham and Trotman, London, pp. 309-318.
We are particularly grateful to Terkel Olsen and Cohen, S.C., 1984. Postseismic deformation due to subcrustal
Charles Jourdan for their interest in the applicability viscoelastic relaxation following dip-slip earthquakes. J. Geo-
phys. Res., 89: 4538-4544.
of these ideas. We also thank Dave Barr, James
Ellenor, D.W. and Mozetic, A., 1986. The Draugen oil discovery.
Jackson, Geof King, Nick Kusznir, John Walsh, Juan In: A.M. Spencer et al. (Editors), Habitat of Hydrocarbons
Watterson and Nicky White for many stimulating on the Norwegian Continental Shelf. Graham and Trotman,
discussions over recent years. London, pp. 313-316.
Faure, J.L. and. Chermette, J.C., 1989. Deformation of tilted
References blocks, consequences on block geometry and extension mea-
surements. Bull. Soc. Geol. Fr., 8: 461-476.
Badley, M.E., Price, J.D., Rambech Dahl, C. and Agdestein, T, Fowler, S. and McKenzie, D., 1989. Gravity studies of the
1988. The structural evolution of the northern Viking Graben Rockall and Exmouth Plateaux using SEASAT altimetry.
and its bearing upon extensional modes of basin formation. J. Basin Res., 2: 27-34.
Geol. Soc. London, 145: 455-72. Gibson, J.R., Walsh, J.J. and Watterson, J., 1989. Modelling of
Barnett, J.A.M., Mortimer, J., Rippon, J.H., Walsh, J J . and bed contours and cross-sections adjacent to planar normal
Watterson, J., 1987. Displacement geometry in the volume faults. J. Struct. Geol. 11: 317-328.
containing a single normal fault. Am. Assoc. Pet. Geol., Bull., Gradijan, S.J. and Wiik, M., 1987. Statfjord Nord. In: A.M.
71: 925-37. Spencer et al. (Editors), Geology of the Norwegian Oil and
Barr, D., 1987a. Lithospheric stretching, detached normal fault- Gas Fields. Graham and Trotman, London, pp. 341-350.
ing and footwall uplift. In: M.P. Coward, J.F. Dewey and PL. Harris, J.P. and Fowler, R.M., 1987. Enhanced prospectivity of
Hancock (Editors), Continental Extensional Tectonics. Geol. the Mid-Late Jurassic sediments of the South Viking Graben.
Soc. London, Spec. Publ., 28: 75-94. In: J. Brooks and K.W Glennie (Editors), Petroleum Geology
Barr, D., 1987b. Structural/stratigraphic models for extensional of North West Europe. Graham and Trotman, London, pp.
basins of half-graben type. J. Struct. Geol., 9: 491-500. 879-898.
Barton, P. and Wood, R., 1984. Tectonic evolution of the North Hellem, T, Kjemperud, A. and 0vreb0, O.K., 1986. The Troll
Sea basin: crustal stretching and subsidence. Geophys. J. R. Field: a geological/geophysical model established by the
Astron. Soc, 79: 987-1022. PL085 Group. In: A.M. Spencer et al. (Editors), Habi-
Bertram, G.T. and Milton, N., 1989. Reconstructing basin evo- tat of Hydrocarbons on the Norwegian Continental Shelf.
lution from sedimentary thickness; the importance of palaeo- Graham and Trotman, London, pp. 217-240.
bathymetric control, with reference to the North Sea. Basin Hollander, N.B., 1987. Snorre. In: A.M. Spencer et al. (Editors),
Res., 1: 247-257. Geology of Norwegian Oil and Gas Fields. Graham and
Birtles, R. 1986. The seismic flatspot and the discovery and Trotman, London, pp. 307-318.
delineation of the Troll Field. In: A.M. Spencer et al. (Edi- Jackson, J.A., Gagnepain, J. Houseman, G., King, G.C.P., Pa-
tors), Habitat of Hydrocarbons on the Norwegian Continental padimitriou, P., Soufleris, C. and Virieux, J., 1982. Seismicity,
Shelf. Graham and Trotman, London, pp. 207-216. normal faulting, and the geomorphological development of
Footwall uplift during normal faulting implications for structural geometries in the North Sea 303

the Gulf of Corinth (Greece): the Corinth earthquakes of McKenzie, D. 1978. Some remarks on the development of
February and March 1981. Earth Planet. Sei. Lett., 57: 377- sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32.
397. Myers, WB. and Hamilton, W, 1964. Deformation accompanying
Jackson, J. and McKenzie, D., 1983. The geometrical evolution the Hebgen Lake earthquake of August 17, 1959. U.S. Geol.
of normal fault systems. J. Struct. Geol., 5: 471-482. Surv. Prof. Paper, 435: 55-98.
Jackson, J.A., White, N.J., Garfunkel, Z. and Anderson, H., Ransome, F.L., Emmons, WH. and Garrey, G.H., 1910. Geology
1988. Relations between normal-fault geometry, tilting and and ore deposits of the Bullfrog district, Nevada. Bull. U.S.
vertical motions in extensional terrains, an example from the Geol. Surv., 407: 1-130.
southern Gulf of Suez. J. Struct. Geol., 10: 155-70. Reilinger, R., 1986. Evidence for postseismic viscoelastic relax-
Jackson, J.A. and White, N.J., 1989. Normal faulting in the ation following the 1959 M = 7.5 Hebgen Lake, Montana,
upper continental crust: observations from regions of active earthquake. J. Geophys. Res., 91: 9488-9494.
extension. J. Struct. Geol., 11: 15-36. Richter, C.F., 1958. Elementary Seismology. WH. Freeman, San
Jankhof, K., 1945. Changes in ground level produced by the Francisco, Calif.
earthquakes of April 14 to 18, 1928 in southern Bulgaria. Roberts, A.M. and Yielding, G., 1991. Deformation around
In: Tremblements de terre en Bulgarie, Nos. 29-31, Institut basin-margin faults in the North Sea/Norwegian rift. In:
M6t6orologique central de Bulgaria, Sofia, pp. 131-136 (in A.M. Roberts, G. Yielding and B. Freeman (Editors), The
Bulgarian). Geometry of Normal Faults. Geol. Soc. London, Spec. Publ.,
King, G.C.P. and Vita-Finzi, C , 1981. Active folding in the 56: 61-78.
Algerian earthquake of 10 October 1980. Nature, 292: 22-26. Roberts, A.M., Price, J. and Olsen, J.S., 1990. Late Jurassic half-
graben control on the siting and structure of hydrocarbon
King, G.C.P., Stein, R.S. and Rundle, J.B., 1988. The growth of
accumulations: UK/Norwegian Central Graben. In: R. Hard-
geological structures by repeated earthquakes, 1. Conceptual
man (Editor), Tectonic Events Controlling Britain's Oil and
framework. J. Geophys. Res., 93: 13307-13319.
Gas Reserves. Geol. Soc. London, Spec. Publ., 55: 229-257.
Kusznir, N.J. and Park, R.G., 1987. The extensional strength
Roberts, S. and Jackson, J., 1991. Active normal faulting in
of the continental lithosphere: its dependence on geothermal
Central Greece: an overview. In: A.M. Roberts, G. Yielding
gradient, and crustal composition and thickness. In: M.P.
and B. Freeman (Editors), The Geometry of Normal Faults.
Coward, J.F. Dewey and PL. Hancock (Editors), Continental
Geol. Soc. London, Spec. Publ., 56: 125-142.
Extensional Tectonics. Geol. Soc. London, Spec. Publ., 28:
Rundle, J.B., 1982. Viscoelastic-gravitational deformation by a
35-52.
rectangular thrust fault in a layered Earth. J. Geophys. Res.,
Kusznir, N.J., Marsden, G. and Egan, S., 1988. Fault block
87: 7787-7796.
rotation during continental lithosphere extension: a flexural
Schwartz, D.P. and Coppersmith, K.J., 1984. Fault behavior and
cantilever model. Geophys. J. 92: 546 (abstract).
characteristic earthquakes: examples from the Wasatch and
Kusznir, N.J., and Egan, S.S., 1990. Simple-shear and pure-
San Andreas Fault zones. J. Geophys. Res., 89: 5681-5698.
shear models of extensional sedimentary basin formation:
Spencer, A.M. and Larsen, V.B., 1990. Fault traps in the northern
application to the Jeanne d'Arc Basin, Grand Banks of
North Sea. In: R.P.F. Hardman (Editor), Tectonic Movements
Newfoundland. In: A J . Tankard and H.R. Balkwill (Editors),
Responsible for Britain's Oil and Gas Reserves. Geol. Soc.
Extensional Tectonics and Stratigraphy of the North Atlantic
London, Spec. Publ., 55: pp. 281-298.
Margins. Am. Assoc. Pet. Geol., Mem., 46: 305-322.
Stein, R.S. and Barrientos, S.E., 1985. Planar high-angle faulting
Larsen, PH., 1988. Relay structures in a Lower Permian in the basin and range: geodetic analysis of the 1983 Borah
basement-involved extension system, East Greenland. J. Peak, Idaho, earthquake. J. Geophys. Res., 90: 11355-11366.
Struct. Geol., 10: 3-8. Stein, R.S., King, G.C.P. and Rundle, J.B., 1988. The growth
Leeder, M.R. and Gawthorpe, R.L., 1987. Sedimentary models of geological structures by repeated earthquakes, 2. Field
for extensional tilt-block/half-graben basins. In: M.P. Coward, examples of continental dip-slip faults. J. Geophys. Res., 9:
J.F. Dewey and PL. Hancock (Editors), Continental Exten- 13319-31.
sional Tectonics. Geol. Soc. London, Spec. Publ., 28: 139- Turcotte, D.L. and Schubert, G., 1982. Geodynamics appli-
152. cations of continuum physics to geological problems. John
Livera, S. and Gdula, J.E., 1990. The Brent Oil Field. In: Wiley and Sons, New York, N.Y.
E.A. Beaumont and N.H. Foster (Editors), Atlas of Oil Turner, C.C., Cohen, J.M., Connell, E.R. and Cooper, D.M.,
and Gas Fields, Structural Traps, II. Traps Associated with 1987. A depositional model for the South Brae oilfield. In:
Tectonic Faulting. American Association of Petroleum Geolo- J. Brooks and K.W Glennie (Editors), Petroleum Geology
gists, Tblsa, Okla., pp. 21-63. of North West Europe. Graham and Trotman, London, pp.
Mher, C.E. and Harker, S.D., 1987. Claymore Oil Field. In: 853-64.
Brooks, J. and Glennie, J. (Editors), Petroleum Geology of Vita-Finzi, C. and King, G.C.P, 1985. The seismicity, geomor-
North West Europe. Graham and Trotman, London, pp. 835- phology and structural evolution of the Corinth area of
845. Greece. Philos. Trans. R. Soc. London, Ser. A, 314: 379-407.
Mansinha, L. and Smylie, D.E., 1971. The displacement fields of Wallace, R.E. 1984. Patterns and timing of late Quaternary
inclined faults. Bull. Seismol. Soc. Am., 61: 1433-1440. faulting in the Great Basin Province and relation to some
Marsden, G., Yielding, G., Roberts, A. and Kusznir, N.J., 1990. regional tectonic features. J. Geophys. Res., 89: 5763-5769.
Application of a flexural cantilever simple-shear/pure-shear Walsh, J.J. and Watterson, J., 1991. Geometric and kinematic
model of continental lithosphere extension to the formation coherence and scale effects in normal fault systems. In:
of the northern North Sea Basin. In: D J . Blundell and A.D. A.M. Roberts, G. Yielding and B. Freeman (Editors), The
Gibbs (Editors), Tectonic Evolution of the North Sea Rifts. Geometry of Normal Faults. Geol. Soc. London, Spec. Publ.,
Oxford University Press, Oxford, pp. 240-261. 56: 193-203.
304 G. Yielding and A. Roberts

Watts, A.B., 1988. Gravity anomalies, crustal structure and area. Bull. Seismol. Soc. Am., 47: 321-325.
flexure of the lithosphere at the Baltimore Canyon Trough. Yielding, G., 1990. Footwall uplift associated with Late Jurassic
Earth Planet. Sei. Lett., 89: 221-238. normal faulting in the northern North Sea. J. Geol. Soc.
Wernicke, B. and Burchfiel, B.C., 1982. Modes of extensional London, 147: 219-222.
tectonics. J. Struct. Geol., 4: 105-115. Yielding, G., Badley, M.E. and Freeman, B., 1991. Seismic
Wheatley, T.J., Biggins, D., Buckingham, J., and Holloway, N.H., reflections from normal faults in the northern North Sea. In:
1987. The geology and exploration of the Transitional Shelf, A.M. Roberts, G. Yielding and B. Freeman (Editors), The
an area to the west of the Viking Graben. In: J. Brooks and Geometry of Normal Faults. Geol. Soc. London, Spec. Publ.,
K.W. Glennie (Editors), Petroleum Geology of North West 56: 79-89.
Europe. Graham and Trotman, London, pp. 979-990. Young, R., 1992. Restoration of a regional profile across the
White, N., 1990. Does the uniform stretching model work in Magnus Field in the northern North Sea. In: R.M. Larsen, H.
the North Sea? In: D J . Blundell and A.D. Gibbs (Editors), Brekke, B.T Larsen and E. Talleraas (Editors), Structural and
Tectonic Evolution of the North Sea Rifts. Oxford University Tectonic Modelling and its Application to Petroleum Geology.
Press, Oxford, pp. 217-239. Norwegian Petroleum Society (NPF), Special Publication 1.
Whitten, C.A., 1957. Geodetic measurements in the Dixie Valley Elsevier, Amsterdam, pp. 221-229 (this volume).

G. YIELDING BadleyAshton and Associates Limited, Winceby House, Winceby, Homcastle, Lincolnshire, LN9 6PB, U.K.
A. ROBERTS BadleyAshton and Associates Limited, Winceby House, Winceby, Homcastle, Lincolnshire, LN9 6PB, U.K.
 305

Modelling secondary hydrocarbon migration in

Haltenbanken, Norway
L. Hermans, A.D. van Kuyk, F.K. Lehner and P.S. Featherstone

This paper reports the results of a map-based modelling study of hydrocarbon migration in the Haltenbanken area, offshore
Mid-Norway. The study represents the first field application of a previously published mathematical model that treats secondary
migration as a buoyancy-driven, separate-phase flow. Fluid viscosities exceeding that of formation water are required by the
theory, thus gas migration is not included. From pa la eo-structure maps and maps of hydrocarbon supply rate at different times in
the geological history, the model computes the evolution of hydrocarbon migration paths and accumulation patterns throughout
a given carrier bed. A general finding from numerous model tests is that system properties such as variations in carrier-bed
permeability and fluid mobility can profoundly affect accumulation patterns.
For the Haltenbanken study, hydrocarbon charge was assumed to derive from the Spekk Formation (Kimmeridge Clay
equivalent) only, expelling a single hydrocarbon phase a representative light oil downward into the carrier bed (the Garn
Formation), which was assumed to be homogeneous and continuous over the study area. A series of computer simulations of
the model was used to evaluate the response of the migration system to variations in the hydrocarbon supply rate and the
carrier-bed properties. Leakage through cap rocks and along faults was not considered and this, combined with the abundance
of generating source rock within the study area, resulted in solutions with most of the structures presently filled to spilling point.
While the model described in this paper clearly is an oversimplification of the actual situation, even in its present form it has
proven its usefulness for assessing the effect of essential migration system parameters such as supply rate, carrier bed and fluid
properties, and the accuracy of palaeo-structure maps, on migration and accumulation patterns.

Introduction by low and skewed hydrocarbon solubilities, limited

supply of compaction water and oil/source-rock cor-
Deterministic hydrocarbon-habitat modelling aims relations (McAuliffe, 1979; Jones, 1981; Tissot and
at predicting the quantity and distribution of hydro- Weite, 1984; Ungerer et al., 1984; England et a l ,
carbon occurrences through geological time and can, 1987).
therefore, play a major role in the decision of where In contrast with oil, the effect of dissolution might
to drill for oil or gas. In the sequence of processes be more important for hydrocarbon gases, notably
leading to the eventual entrapment of hydrocarbons, methane, because of the combination of the much
secondary migration plays a key role in the hydrocar- higher water solubility (of the order of 1000 ppm un-
bon distribution. der in-situ conditions: Dodson and Standing, 1944;
The literature on the mechanism(s) of secondary McAuliffe, 1979) and the increased diffusion coeffi-
oil migration is extensive, and many different pro- cient (Krooss and Leythaeuser, 1988). An order-of-
cesses have been proposed. In recent discussions, an magnitude calculation shows that saturation of a car-
hypothesis of separate-phase flow has generally been rier bed with gas by diffusion might take place in one
favoured (Hobson, 1954; McAuliffe, 1979; Neglia, million years. However, in view of the much higher
1979; Schowalter, 1979; Jones, 1981; Du Rouchet, transport velocity, separate-phase flow can become
1981; Tissot and Weite, 1984; Ungerer et al., 1984; dominant even in unsaturated aquifers. Moreover,
England et al., 1987; Lehner et al., 1987; Dembicki the usually vast volumes of gas generated far exceed
and Anderson, 1989). Other mechanisms, e.g., trans- the dissolution capacity of the average aquifer, again
port in solution as argued by Price (1976, 1981), suggesting separate-phase flow to be the prevailing
could operate in parallel and occasionally even in transport mechanism.
series, as is evident from the occurrence of diffusive Based on the assumption of segregated flow, a
losses of dissolved lighter components through capil- mathematical model was developed that describes
lary barriers (Krooss et al., 1988). However, these are secondary hydrocarbon migration in hydrostatic
more difficult to reconcile with constraints imposed aquifers. The background and underlying theory of

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 305-323. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
306 L. Hermans, A.D. van Kuyk, F.K. Lehner and P.S. Featherstone

this dynamic flow model were discussed comprehen- hydrostatic aquifers. Vertical migration across carrier
sively in a previous publication (Lehner et al., 1987). beds is assumed to take place instantaneously on
A brief outline is given below. The present paper the time scale set by the expulsion rates. Vertical
reports a field application of the numerical simula- migration through cap rocks and seals can be mod-
tor SECMIG, which is based on the model, to the elled as a quasi-steady flow, governed by the aquifer
Haltenbanken area, offshore Mid-Norway. potential drops between superimposed carrier beds,
The Haltenbanken area was chosen because it though this approach has not been included in this
contains oil and gas accumulations in an area of study.
generally good seismic data quality and coverage, The input required by the model's simulator,
although source-rock richness and distribution are SECMIG, consists of structure-contour maps at the
too poorly constrained for the area to serve as a base of the topseal, rate maps of hydrocarbon ex-
full calibration set. Nevertheless, the structural and pulsion (supply), and the properties of the carrier
lithostratigraphic setting, together with the known bed and hydrocarbon fluid for a geologically relevant
hydrocarbon occurrences and the existence of new period of time. The program output consists of cal-
prospects, make it a good location for exploring the culated hydrocarbon column heights below the seal.
capabilities of the technique. These may be plotted in different colour shades on a
structure-contour map of the seal at any time during
the simulation.
The migration model SECMIG
The simulator with which this study was performed
Modelling hydrocarbon migration is essentially a has a number of simplifications and restrictions:
three-dimensional problem, and this, combined with - in the derivation of the model the resistant force
the size and structural complexity of the systems of the water phase has been disregarded; hence,
to be modelled and the typically very high migra- transient flow will be modelled adequately only if hy-
tion velocities involved (metres/year) compared with drocarbon fluid viscosities significantly exceeds that
the time interval to be simulated (many millions of of the formation water;
years), would make an approach based on the solu- - a sharp transition between the oil and water
tion of the standard equations governing immiscible phases is assumed, ignoring capillary pressure effects;
two-phase flow in three dimensions computationally - the current version of the program is restricted
unfeasible. to single carrier-bed/impermeable cap rock systems;
A substantial simplification was found in exploit- - the aquifer is assumed to be hydrostatic;
ing the phenomenon of gravity segregation of oil - carrier-bed thickness is expected to exceed the
and water, which may be expected to occur in per- hydrocarbon fluid column height at every location
meable carrier beds (Hobson, 1954). By analogy during the simulation (unconfined flow);
with the analysis of the motion of the phreatic - faults are treated merely as vertical steps at
surface in unconfined groundwater flow (see, for ex- the base of the cap rock and no specific fault-zone
ample, Wooding and Chapman, 1966), lateral updip properties (i.e., permeability or capillary threshold
migration of a separate hydrocarbon phase under pressure) are allowed for;
seals is conceived in our model as a Boussinesq- - only a single hydrocarbon phase (oil) of fixed
type, transient, free surface flow (Boussinesq, 1904; composition is modelled;
Bear, 1972), with source and sink terms represent- - zero residual oil saturation is assumed within the
ing supply from source rocks and leakage through carrier bed;
cap rocks and faults, respectively. Thus, we assume - fluids are treated as incompressible.
that hydrocarbons, upon entering a carrier bed, will Since any migration system will contain capillary
form accumulations below a cap rock that are dis- barriers, we have implicitly assumed that these can
charged by a buoyancy-driven flow up-structure. This be specified in advance, based, for example, on litho-
enables a two-dimensional, map-view mathematical logical criteria. Estimates of residual oil in those
description of a three-dimensional time-dependent carrier-bed regions that have been hydrocarbon satu-
secondary migration system, the dependent variable rated at a certain stage in the migration history could
being the height of the hydrocarbon column below easily be incorporated in the present model. In addi-
the seal. This analysis implies a sharp distinction tion, hydrocarbon losses within the remainder of the
between vertical migration (i.e., cross-stratigraphy), carrier bed are expected to be generally low because
giving rise to the source and sink terms, and lateral of efficient focussing of the oil flow during vertical
migration. The latter is essentially strata-bound and migration. This assumption seems reasonable and
forms the part which is actually modelled, a seg- is supported by the recently published experimental
regated flow of a buoyant oil phase in individually results of Dembicki and Anderson (1989).
Modelling secondary hydrocarbon migration in Haltenbanken, Norway 307

Study area-regional setting 3 shows the local stratigraphy in cross-section. The

main source rock in the sequence is the Upper Juras-
The migration simulation study was performed for sic Spekk Formation (Kimmeridge Clay lithostrati-
a rectangular block in the Haltenbanken region, off- graphic equivalent), the lowermost section having the
shore Mid-Norway, covering an area of 82.5 x 72 km best-developed source rocks. It is overlain by sealing
situated between 630' and 800, and 6408' and Cretaceous mudstones and claystones that thicken
6445'N. Figure 1 shows a location map. The geol- progressively towards the deeper parts of the basin.
ogy is well described in the literature (Bukovics and The underlying Jurassic sandstones are expected to
Ziegler, 1985; Cohen and Dunn, 1986). The eastern form the main migration avenue for hydrocarbons
part of the study area is defined by the west flank generated and expelled from the Spekk Formation.
of the relatively stable and undeformed TY0ndelag The carrier beds in the Middle/Lower Jurassic (Garn
Platform, containing the Draugen (undersaturated and Tilje Formation) are extensive throughout the
oil), Mikkel (gas condensate + some oil) and, on region; however, the Upper Jurassic carrier beds
the Fr0ya High, the Njord (oil + gas) accumula- (Rogn Formation) are, in general, restricted (Cohen
tions (Pedersen, 1988). A major fault zone tending and Dunn, 1986). Downward hydrocarbon expulsion
to SW-NNE separates the Tfr0ndelag Platform from from the Spekk Formation into these carrier beds
the heavily structured and faulted Halten Terrace may be hindered by the locally intervening shaly
which covers the rest of the map and forms the main Melke Formation. The Melke Formation is absent
kitchen area. The major fault activity took place in on a number of structural highs but well developed
the Late Jurassic/Early Cretaceous period. Just be- in the lows, particularly towards the northwest within
yond the northern edge of the study area, oil, gas the study area. The very thick Lower Jurassic/THassic
and condensate have been discovered in South lyri- fluvio-deltaic Are Formation contains numerous coal
hans and North Tyrihans (Pedersen, 1988). Within seams, and is also considered a possible source rock
the study area, an uneconomic gas accumulation was (mainly for gas) in the area.
found by well 6407/4-1 in a very tight formation. All Since the present version of SECMIG can handle
fields produce from Jurassic strata. only one caprock/carrier-bed interface at a time, the
local stratigraphy had to be simplified in the model as
Stratigraphy and assumed model lithology indicated in Fig. 4, with the Spekk Formation acting
both as source rock and cap rock for the hydrocar-
Figure 2 shows an isometric depth display of the bons expelled downward into the underlying Juras-
main structurally mapped seismic horizons, and Fig. sic carrier bed. For the simulations, the seismically

go go o 10 o 1 2E o 6 70 o 8 E

Fig. 1. Location map of the Haltenbanken area, showing (a) the main structural elements and (b) the distribution of the hydrocarbon
accumulations. Figure modified from Cohen and Dunn (1986) and Hvoslef et al. (1988).
308 L. Hermans, A.D. van Kuyk, EK. Lehner and ES. Featherstone

SEA FLOOR

IASE PLIOCENE FORESETS

I
BASE TERTIARY

INTRA CRETACEOUS

BASE JURASSIC

10 km

Fig. 2. Isometric display of the main horizons in the Haltenbanken study area.

WNW ESE

HALTEN TERRACE

Njord Draugen

Om 0m

2000 J 2000

4000 h4000

6000 h6000

8000 - 8000

10000 J

10 km

Fig. 3. Stratigraphy of the Haltenbanken study area. The location of the cross-section is indicated in Fig. 5.
Modelling secondary hydrocarbon migration in Haltenbanken, Norway 309

(a) ACTUAL 4> (b) SECMIG

mapped
K L horizon

Spekk
U R gn source &
^ ^ ^ ^ ^ cap rock
Melke

Gam
o
'55 M Not
co
3 x / lie
~5

Ror

L ac /
Tilje

oo
' "' 'i'ii iiTi"h'JMiiiiihii'ii
Are
Tr U

Fig. 4. Stratigraphic column of the Jurassic in the Haltenbanken study area: (a) actual, (b) model stratigraphy used in SECMIG.

mappable base Cretaceous/Upper Jurassic marker SECMIG runs

was used as the sealing horizon. This was judged
sufficiently close to the actual source-rock/carrier- The input required for SECMIG consists of base-
bed interface, in view of the map's accuracy and the seal contour maps and hydrocarbon supply rate maps
near-parallel nature of the base Cretaceous and base throughout the simulated burial history, together
source-rock horizons. The present-day contour map with the physical properties of the carrier bed and the
of this marker horizon as used in the simulations oil. The value of the model's prediction, therefore,
is shown in Fig. 5. The Melke Formation has been relies heavily on the accuracy of the preceding basin
disregarded. Its effect on expelled hydrocarbon vol- reconstruction and the choice of parameters. A series
umes is provisionally accounted for by an expulsion of four computer runs of the model has been made
factor (see below). Supply from the Are Formation to evaluate the response of the migration system to
was not considered, since most of it is expected to input variations. The runs differed in the following
have passed the maturity window at the time of trap input characteristics:
formation. The cap rock is assumed to be imperme-
run I: best-fit constant sediment heat flow in
able over the mapped area. Faults are treated merely
space and time;
as steps in the cap rock, hence migration losses by
run II: same input as in I, but heat flow variable in
leakage along faults are not considered. The large
space and time;
regional faults on the edge of the Halten Terrace
run III: same input as in II, but with a gentler
have throws which are most likely greater than the
decline of carrier-bed porosity and perme-
actual carrier-bed thickness. In computer runs I, II
ability with depth;
and III of this study they were assumed to be per-
meable, enabling migration to take place from the run IV: same input as in III, but with a barrier to
downthrown into the upthrown block as if the carrier lateral migration introduced into the map
beds were in communication. In run IV they were (the fault zone west of Draugen).
assumed to be sealing, which was accomplished in In runs III and IV only the final 15 Ma have
the simulation by making the grid points in the fault been calculated with the specified parameter setting.
zone impermeable. To speed up the calculation, the initial 125 Ma of
these runs was modelled with a lower permeability,
while careful checks were made to ensure that this
procedure did not affect the final results.
310 L. Hermans, A.D. van Kuyk, EK Lehner and P.S. Featherstone

h64 30' N

1-64 08' N
6 30 E 10 km 8 00' E

Fig. 5. Present-day depth map of the base Cretaceous/Upper Jurassic marker horizon in the Haltenbanken study area, along which
secondary hydrocarbon migration was modelled. The dashed lines indicate the location of the cross-section of Fig. 3.

Calculation procedure variations in these parameters upon burial. Only the

hydrocarbon density was held constant. Although gas
Table 1 and Fig. 6 provide a scheme of the calcula- is also present in the area, only a single hydrocar-
tion procedure and the chosen parameter settings. bon phase (oil) could be modelled with the current
Well data and mapped seismic horizons were used version of the simulator. In the simulations a flux
to reconstruct the base-seal/source-rock burial his- boundary condition was used, thus allowing hydro-
tory, thus obtaining palaeo-depth maps. Single-point, carbon outflow from the map area, but excluding
palaeo-heat-flow values for the region, obtained from inflow.
best fits to present-day well bottom-hole temperature
data during burial reconstruction, were contoured
Input
over the map area and used as input for calculating
the palaeo-temperature at source-rock level. Having Base-seal burial and temperature reconstruc-
calculated the source-rock time/temperature history, tion
rate maps of hydrocarbon generation could be de- The burial reconstruction over the area was based
rived using the recently published kinetic model of on present-day depth maps of five horizons: Tbp
Klomp and Wright (1989) with laboratory-measured and Base Pliocene, Base Tertiary, Intra-Cretaceous
kinetic parameters appropriate to the Spekk Forma- and Tbp Jurassic, shown in the isometric display
tion (Kimmeridge clay type) source rock, and source- (Fig. 2). For run I (heat flow constant in space
rock properties that were averaged over the map and time) an average sediment heat flow of 55
area, Tb obtain the hydrocarbon-supply-rate history, mW/m2 was calculated and used as input to the
the result of the generation calculation should be palaeo-temperature calculations. More realistically,
input to an expulsion model. Since a suitable model in the other runs heat-flow variation in space and
is not available at present, an expulsion/generation time was included. In these runs a crustal-rifting
ratio was estimated. The carrier-bed and fluid pa- model modified from McKenzie (1978) was used for
rameters, forming the additional input, were made a the heat-flow calculations. The resulting palaeo and
function of depth and temperature to simulate in-situ present-day heat-flow maps (Fig. 7) show a steady
Modelling secondary hydrocarbon migration in Haltenbanken, Norway 311

TABLE 1
Calculation sequence with parameter settings for Haltenbanken study

(1) INPUT FOR BURIAL & TEMP. HISTORY RECONSTRUCTION OF SOURCE ROCK/SEAL HORIZON
present-day depth maps of relevant (digitised seismic) horizons (Fig. 2)
well BHT data within and around study area
map area 82.5x72 km2 (55x48 grid nodes) (Fig. 1)
(RUN1 75x75 km2 (50x50 nodes)
heat flow RUN 1 constant, 55 mW/m2, best-fit constant heat flow
RUNS 2-4 variable, based on a crustal-rifting model (Fig. 7)

i
burial and temperature history of Upper Jurassic Spekk source rock
from onset of deposition (140 Mabp) to present

(2) INPUT FOR HC GENERATION CALCULATION

time/temperature result from (1) and
first-order kinetic model with appropriate type-ll source-rock kinetic parameters
source rock thickness 100 m
init. volume fraction kerogen 0.1
init. atomic H/C ratio kerogen 1.2

I
history of hydrocarbon generation rate for Upper Jurassic source rock

(3) INPUT FOR HC SUPPLY ("EXPULSION.') CALCULATION

result from (2) and (multiplied by)
source rock expulsion/generation ratio 0.1

history of hydrocarbon expulsion rate for Upper Jurassic source rock

(4) INPUT FOR SECMIG

burial history seal/carrier
bed interface (-source rock) from (1) + result from (3) +
oil density (constant) 800 kg/m3
oil viscosity (variable) 7.67.10"7. exp (2310/temp) Pa.s (Fig. 6)
carrier-bed porosity (variable) RUNS 1-2 0.59. exp (-z/2340) (" ")
RUNS 3-4 0.42. exp (-z/4000) (" ")
carrier-bed permeability (variable) RUNS 1-2 exp ((porosity - 0.162)/0.020) mD ( " ")
RUNS3-4 exp ((porosity- 0.100)/0.022) mD (" ")

decline from a maximum around 120 Ma to the thicknesses based on well data against present-day
present-day situation. depth (Fig. 8) reveals no clear correlation. In view
of these uncertainties, the source rock was simply
Source-rock characterisation assumed to be continuous over the study area with a
The Spekk source-rock properties were taken to uniform thickness of 100 m, which is probably an un-
be constant over the area. Although lateral variations derestimate. An average richness of 4% wt. organic
obviously do occur, these were not mapped in view matter (10% vol.) was taken, based on geochemical
of the uncertainties involved and biased sampling. analyses of Spekk formation samples from within
The latter is the case particularly for the measured the study area and further to the north. The initial
thicknesses. Well thicknesses are in the range of 0 atomic H/C ratio of the immature source rock was
to 100 m, but since all data were obtained from set at 1.2, a value typical of a type II source rock.
structural highs, where the source rock is most sensi- For the expulsion/generation ratio we have taken
tive to weathering and erosion, and locally has been the very conservative value of 0.1 to compensate for
eroded completely, thicknesses greater than 100 m losses in the Melke Formation and in the Cretaceous
may be present in deeper basin regions. Estimates of Formations on top of the Spekk Formation.
the maximum Spekk Formation thickness within the
lows based on seismic interpretation vary consider- Hydrocarbon fluid (oil) properties
ably from "thinly developed" (Bukovics and Ziegler, The 40API undersaturated light oil encountered
1985), a few hundred metres (Ellenor and Mozetic, in Draugen was used as the simulation fluid. This was
1986), to a few kilometres (Heum et al., 1986). In the a compromise, since different types of light oil (and
case of uniformly increasing basin subsidence, for- gas) are found in the region. The oil density was
mation thickness might relate to present-day burial kept at a constant 800 kg/m3. The actual change
depth. However, plotting the measured source-rock in this parameter is limited, since during the up-
312 L. Hermans, A.D. van Kuyk, EK. Lehner and P.S. Featherstone

(a)

Draugen - type oil

RUNS I IV

| o , -

i 1
50 100 150
TEMPERATURE C

(b)

1 r~
0.2 0.3
POROSITY

1500-

(d)

E 3000 -

~I I
10"1 0 10 10'
PERMEABILITY mD

Fig. 6. (a-c) Oil viscosity and carrier-bed porosity and permeability trends used as input for the Haltenbanken SECMIG simulations.
Relationship (d) results from a combination of (b) and (c). Measured Garn Formation data are indicated by dashes. See Table 1 for
run input characteristics.

ward flow of hydrocarbons the effects of decreasing Carrier-bed properties

pressure (giving a density decrease) and tempera- The carrier bed was assumed to be homogeneous
ture (giving a density increase) partly compensate and continuous over the study area, except for the
for one another. The oil viscosity was modelled to fault zone at the eastern edge of the Halten Terrace
decrease with increasing temperature according to a in run IV, which was made impermeable to simulate
standard reservoir engineering relationship, which in a lateral migration barrier (Fig. 9).
SECMIG is approximated by the Andrade exponen- The present SECMIG version assumes an expo-
tial formula (Latil, 1980, pp. 146-149), with Draugen nential porosity/depth trend, although, for the Hal-
oil parameters: tenbanken data, linear compaction with depth might
1.3 x 10"3 Pa s (1.3 cP) at 37C have been more appropriate, as is apparent from Fig.
0.7 x 10"3 Pa s (0.7 cP) at 68C Draugen 6b and from data shown by Bj0rlykke et al. (1986).
reservoir temperature For the simulations two compaction curves were cho-
0.35 x 10"3Pa s (0.35 cP) at 104C. sen, both to fit a fractional porosity of around 0.30 at
a depth of 1650 m (Draugen): a lower porosity trend
Modelling secondary hydrocarbon migration in Haltenbanken, Norway 313

i
1
/ / \
140 Ma 120 Ma

1
75 Ma Present day
Fig. 7. Palaeo-heat-flow maps of the Haltenbanken study area based on a crustal-rifting model modified from McKenzie (1978), used
in SECMIG runs II-IV

to fit (0.07, 5000 m) for runs I and II, and a higher

porosity trend to fit (0.12, 5000 m) for runs III
and IV. In comparison with the local reservoir data,
somewhat low porosity estimates were used in the
simulations, particularly for the shallower (<3500 m)
depths. This choice was based on the idea that since
most of the reservoir rock samples analysed were
impregnated with hydrocarbons, which may have
impeded diagenetic porosity reduction, the "average"
carrier bed might indeed have a lower porosity.
A least-squares fit on porosity/permeability core
data plotted on a linear-log scale is commonly used
as a good first approximation for describing per-
meability trends. Indeed, there are many examples
of reservoirs for which good correlation is observed
(Archie, 1950; Fchtbauer, 1967). However, when
data over a wide permeability/porosity range have to
be taken into account, for example, a carrier bed dur-
ing its entire range of burial depths, a curved trend or
a set of depth-dependent logarithm of permeability-
versus-porosity trends (Bj0rlykke et al., 1986) might
be more appropriate, as suggested by field data

Fig. 8. Measured Spekk Formation source-rock thickness from

well data versus present-day depth of source-rock in Hal-
Spekk Fm. source rock thickness
tenbanken.
314 L. Hermans, A.D. van Kuyk, F.K. Lehner and P.S. Featherstone

I ' ! ' I
380000 390000 400000 410000 420000 430000 440000 450000

10 km

Fig. 9. Present-day depth map used for simulation run IV of the base Cretaceous/Upper Jurassic marker horizon in the Haltenbanken
study area, showing a lateral migration barrier (hatched zone) introduced along the Vingleya fault-zone in the southeastern part of the
study area.

(Tissotand Weite, 1984; Chilingar, 1963) and pre- factor in the analysis. Because a range of several
dicted by theory (Carman, 1937; Berg, 1975) (Fig. orders of magnitude is involved, a sensitivity analysis
10). The Garn Formation sample data plotted in of this parameter is generally recommended.
Fig. 6c tend to support this view, although combined
porosity/permeability measurements were available Results and discussion
only at shallow (Draugen; 1650 m) and deep lev-
els (below 3750 m), while according to Fig. 6b the The simulation results are displayed as rate curves
medium-depth samples probably would have tended and migration maps. Figure 11 shows the hydrocar-
more toward the centre in the porosity/permeability bon budget for the study area as obtained from the
plot. In view of these uncertainties, the simple linear- hydrocarbon generation calculation. Figures 12-15
log porosity/permeability relationship was still used show the SECMIG results of runs I-IV for a num-
in the SECMIG runs. For the present simulations it ber of past and present time steps. Figure 16 is an
was made to fit the values (0.30, 1000 mD) of Drau- enlargement of the present-day stage of run III (Fig.
gen and (0.07, 0.01 mD) in the deepest part of the 14d) with (re)migration paths indicated. The calcu-
study area in runs I and II, and the values (0.28, 4000 lated outflow of oil from the map area as a function
mD) and (0.12, 2.5 mD) in runs III and IV. To save of time is sketched in Fig. 17.
computer time, permeability was restricted to a max- In the following sections, five major effects will be
imum of 1 Darcy. Figure 10 offers a closer look at discussed.
the chosen trend in comparison with other porosity/
permeability data sets and correlations predicted by Oil migration velocity and implications
theory. An appreciation of the hydrocarbon transport
The permeability estimate remains an uncertain rate in the present SECMIG runs may be ob-
Modelling secondary hydrocarbon migration in Haltenbanken, Norway 315

Legend
This study runs l-ll
0.350
runs lll-IV
decreasing grain size
& sorting -x Haltenbanken Fm. trend after
3000 m Bjerlykke et al., 1986:
combination of set of 0 vs. K
trends (depth as parameter)
with 0 vs. depth trend; depths
O 0.250 H indicated
< diagenesis
<- ^ Available Haltenbanken Fm. data
Q
^ ^ (see Fig. 6c)
CO "> Trends for SST of different
O "" ' geological ages in NW Germany,
g 0.150 Tissot and Weite, 1984

Theoretical relationships:

^s Kozeny-Carman for constant

specific surface, Carman, 1937

0.050 s> Berg, 1975 for constant average

-"" grain diameter

both fitted to K = 1D, 0 = 0.3

0.000
0.010 0.100 1.000 10.000 100. 1000 10000 Note: these theoretical curves are
ABSOLUTE PERMEABILITY K MD shown merely to indicate trends;
they were derived and are valid
for high K only.

Fig. 10. Carrier-bed permeability/porosity trend used in this study, compared with other data sets and with correlations provided by
theory.

1 0.3 0.3
/

constant heat flow variable heat flow

run 1 run ll-IV
entire map entire map
0.2 I 0.2 I

D.
\
0.1 1 0.1
Si X 1 1
.ja/ 1 0
I
> 1
1
I
ro
3 1
E 1
3 1
O

0
90 70 28 present 90 70 28 present

Time Ma Time Ma

Fig. 11. Hydrocarbon budget of the Haltenbanken study area.

tained by considering the simple example of plane, bed permeability relative to the hydrocarbon phase,
steady-state up-structure fluid flow, with a stationary 0 = hydrocarbon fluid viscosity, h = vertical hy-
hydrocarbon/water interface parallel to the cap-rock drocarbon column height below the cap rock, a =
boundary. In this case application of Darcy's law, cap-rock dip angle.
combined with the assumption of a sharp oil/water In the steady state the hydrocarbon transport rate
interface and disregarding the resistant force of the equals the hydrocarbon supply rate (per unit width).
water phase (the water remaining stationary), yields Equation 1, therefore, immediately shows the rela-
the oilflowrate per unit width (dimension L3/LT): tionship between supply (expulsion) rate, hydrocar-
bon column height and the physical parameters for
<7migr =Apg( h sin a cos a (1) the conditions outlined here.
The value of ^rmigr corresponds to a mean fluid-
where Ap = water-hydrocarbon fluid density differ- particle transport velocity (dimension LIT):
ence, g = acceleration due to gravity, K0 = carrier-
 70 Ma 28 Ma 10 Ma

HC column height (m)

above 300.000
&
100.000 300.000
30.000 100.000
10.000 30.000
3.000 10.000
1.000 3.000
nn 0.300
i 1.000
0.100 0.300
LJ 0.030
HUH 0.100
0.010 0.030
0.003 0.010
1 below 0,003

a.
4 Ma Present

Fig. 12. Modelled secondary hydrocarbon migration in the Upper-Middle Jurassic in the Haltenbanken area, Norway. Run I: constant sediment heat flow. Physical properties are shown in Fig. 6.
I
 70 Ma 28 Ma 10 Ma

(d) Legend

F"<K* % HC column height (m)

H I above 300.000
* *
M i 100.000 300.000
/V
/ WKk 30.000 100.000
' ' H* H i 10.000 30.000
M B 3.000 10.000
M) ( 3.000
'""""" 0.300 1.000
0.100 0.300
0.030 0.100
0.010 0.030
. , . / / ' 0.003 0.010
below 0.003

4 Ma Present

Fig. 13. Modelled secondary hydrocarbon migration in the Upper-Middle Jurassic in the Haltenbanken area, Norway. Run II: same input as in I, except for heat-flow variable in space and time.
Physical properties are shown in Fig. 6.
 GO
00

10 Ma 4 Ma

t-
S
Legend

HC column height (m)

H above 300.000
fault
B 100.000 300.000
H [ | 30.000 100.000 EBB area of supply
H B 10.000 30.000
WBA 3.000 10.000 scale 1 : 1 000 000
1.000 3.000
0.300 1.000 10 km
CD 0.100 0.300 ft
0.030 0.100
0.010 0.030
0.003 0.010
below 0.003
ft.
ft*

2 Ma Present

Fig. 14. Modelled secondary hydrocarbon migration in the Upper-Middle Jurassic in the Haltenbanken area, Norway. Run III: same input as in II, but with a gentler decline of carrier-bed porosity
and permeability with depth. Physical properties are shown in Fig. 6.
Modelling secondary hydrocarbon migration in Haltenbanken, Norway 319

g
31
o o o o o o o o o o o c o
o o o o o o o o o c o - - o
O O O O O O O C O T - O O O
C O i - C J
O O O CO

P
0<

p
a
00
c

2 VO
P OD
1 3-
< E
a
^
d

2 c
8. <
cm
P
<u
<u
_c
&
cs o
o cOH
2 1?
B
c &
Pn
.8
'u

O
8 01)
u P
T3
>*Q4-1
J-l

o
o

8 *<a
C

5/5
e
o
< N
15 ,*3
g <i3P
%_
s
tr>
CU
C3
it
S
W) <u
E A
320 L. Hermans, A.D. van Kuyk, EK. Lehner and P.S. Featherstone

10 km

Fig. 16. Enlarged display of Fig. 14d (scale 1:700 000), showing the computed present-day hydrocarbon distribution in the Upper-Middle
Jurassic carrier system underneath the Spekk Formation source rock in Haltenbanken, Norway. The carrier-bed was assumed to be
continuous. Arrows indicate (re)migration pathways.

100

= 75

run II

=5= . 5 S ^

3 50 -J
E
^__ - ~" ^-~"~~~ 7'
>-^^C^ run I I I /
^* ^ ^ run IV
/<\
/ Vunl
O 25 /
/
/
/
/
I ni
90 70 50 30 10 present

Time Mabp
Fig. 17. Hydrocarbon outflow from the Haltenbanken study map area as a function of time. SECMIG runs I-IV.

ffmigr _ ApgK0sina As a result of the light oil and favourable perme-

(2)
(1 - Swc) h cos a 0 (1 - Swc) ability encountered, the linear oil-migration veloci-
ties in the Haltenbanken area are high in comparison
where 5WC = carrier-bed connate water saturation with the duration of other geological processes. Sub-
(assumed 0 in our simulations), and = fractional stitution of the Haltenbanken parameter values in
carrier-bed porosity. eqn. 2 reveals lateral migration velocities of the or-
Modelling secondary hydrocarbon migration in Haltenbanken, Norway 321

der of 0.1 m/a (runs I and II) and 1 m/a (runs III and However, the present-day generation rate is much
IV) at 4000 m depth, increasing to 100 m/a at 1600 higher in the constant heat-flow run. This results
m depth for a 4 cap-rock dip angle, the range being in typically higher (transient) migrating oil columns,
caused mainly by the decrease in carrier-bed per- particularly in low-permeability (deep) areas in which
meability with depth. Consequently, the hydrocarbon generation or (re)migration is still taking place, since
column heights as depicted in the SECMIG output in these areas the linear migration velocity is low (see
maps will adjust almost instantaneously in response eqn. 1). Owing to this transient effect in the constant
to changes in the input parameters (supply, geome- heat-flow run, less outflow has occurred and more
try, physical parameters). That is, transient times will oil is present within the map area in the present-
be short. Inherent in the high oil-migration velocity day situation. However, the logarithmic oil column
are the typically small hydrocarbon column heights height colour or grey-tone scale tends to exaggerate
encountered on the migration routes between the this difference in oil content visually.
accumulations, notably at shallow depth. The effect of generation/expulsion timing on the
present-day distribution of the hydrocarbons is less
Heat-flow history significant in these simulations than might be ex-
In the Haltenbanken area with constantly increas- pected, because of the rather uniform subsidence
ing burial with time (for an example burial graph see with little palaeo-tilting of the strata.
Fig. 18), present-day depths and temperatures will
dominate the present-day cumulative amounts of oil Permeability/depth trend
generated. However, the timing of the onset of oil In SECMIG the carrier-bed permeability/depth
generation will be affected by the heat-flow history. trend is obtained by combining porosity/depth and
The SECMIG output of run I (Fig. 12) compared with permeability/porosity relationships, calibrated with
run II (Fig. 13) and the area hydrocarbon budget in measured data (Fig. 6). The carrier-bed porosity range
Fig. 11 nicely show the impact of heat-flow history on is small and its effect upon migration is restricted. The
the timing of generation and its effect upon migra- permeability trend, however, has much more influ-
tion: as a result of the elevated palaeo-heat flow in ence, its range covering several orders of magnitude.
the variable heat-flow case (Fig. 7), both the onset and The drastic permeability reduction with depth trans-
major phase of generation are shifted backwards in lates into decreasing hydrocarbon migration velocities
time. This is reflected in earlier trap fill and, where and, hence, increasing migrating oil column heights
applicable, spill-over and remigration (Bora, Njord, at a given supply rate (Figs. 12-16). This effect is
structures in the northwest of the area). The some- much too pronounced to be compensated by the de-
what higher present-day cumulative hydrocarbon gen- creasing oil viscosity with temperature (depth).
eration calculated for the variable heat-flow case (see In runs III and IV a different porosity/depth trend
Fig. 11) will have added to the extent of spill-over and is used. The resulting difference in permeability,
re-migration in the most recent time stages. compared with the other two runs, rises to more
than one order of magnitude at a depth of 4000 m,
the result of which can be seen by comparing the
JJ CRETACEOUS TERTIARY | final stages of run II and run III. In this respect,
the relative timing of hydrocarbon generation and of
reduction in carrier-bed permeability play key roles
in retarding migration.

i.o i Map hydrocarbon outflow

The flux boundary condition used in SECMIG
E will only allow for the outflow of hydrocarbons from
I2-0 the map area and may, consequently, result in a
Q
0)
conservative estimate of the amount of hydrocarbons
retained, since supply from any kitchens outside
3.0 i
the map area is not taken into account. Whenever
possible, the boundary should, therefore, be chosen
such that it includes the entire kitchen area of the
150 100 50 Present prospects of interest, taking into account the effects
Time Ma of re-migration. In running the program, one may
Fig. 18. Burial history at base Cretaceous marker level typical record the outflow across any boundary segment as a
for the Haltenbanken study area. function of time.
322 L. Hennans, A.D. van Kuyk, F.K. Lehner and P.S. Featherstone

As shown in Fig. 17, in the Haltenbanken simu- critical for low-relief structures such as Draugen,
lations roughly half of the expelled oil has migrated where a few tens of metres difference in depth can
out of the study area by the end of the simulation. have drastic effects.
Run I shows by far the lowest total outflow (31% of Of particular interest are the results of run IV,
the oil expelled), as a result of the transient effect which suggest that a substantial proportion of the
described in the section on heat-flow history. Clearly, oil present in the Draugen structure originated
the combination of a low carrier-bed permeability through re-migration from Njord. Moreover, this re-
combined with a much higher recent oil supply rate migration may well be responsible for the observed
(invoked by the different heat-flow history in run I) distributions of lighter (gas) and heavier hydrocar-
is the main factor influencing outflow across the map bons over the Draugen and Njord fields, as generally
boundary in this series of simulations. In runs II-IV discussed by Gussow in 1954.
around 60% of the oil supplied has flowed out, with
only minor differences. Not surprisingly, the sealing Conclusions
fault zone in run IV inhibits migration towards the
right edge of the map, resulting in a somewhat lower The results of a first field-scale application of a
overall oil outflow than that in run III. secondary hydrocarbon migration simulator to an
82.5 x 72 km2 block in the Haltenbanken area, off-
Hydrocarbon accumulation and re-migration shore Mid-Norway, are reported. Limitations im-
The map in Fig. 16 shows most structures filled to posed by the current version of the simulator and
spilling point. In practice, added complications of gas, poor constraints on some of the geological and geo-
complex stratigraphy and leakage along faults, not in- chemical input parameters (source-rock data in the
cluded in the model, mean that the results have to deeper-buried regions, fault properties, carrier-bed
be treated with caution. The one structure modelled continuity) inherent in any exploration study com-
as being only partially filled is Draugen. This field is pelled the use of a grossly simplified model in this
known, however, to be filled to spilling point. Three study of the Haltenbanken area. The hydrocarbon
possible explanations for this discrepancy are evident: charge, a representative light oil, was assumed to de-
(a) the palaeo-structure maps are incorrect; in rive solely from the Spekk Formation (Kimmeridge
this context a partial structure opening of Draugen Clay equivalent). The oil was assumed to have been
during the simulations from 3 to 1.5 Ma B.P. is expelled downward into a single carrier bed (the
indicated (Figs. 14c, 15c); Garn Formation), which was assumed to be homo-
(b) the recent supply rate from the Draugen geneous and continuous over the entire area. Faults
kitchen area is underestimated; additional runs not were modelled as simple steps in the carrier bed.
shown in this article have ascertained that a forty- Neither leakage through cap rocks and along faults
fold increase in the recent supply rate would be nor losses due to residual oil were considered and
required to fill the structure if the palaeo-structure this, combined with the abundance of generating
maps were correct; source rock within the study area, resulted in solu-
(c) the migration is longer-range than assumed; tions with most of the structures presently filled to
the Draugen kitchen extends across our map bound- spilling point. Although Draugen, the shallowest trap
ary and there is.a hydrocarbon contribution from in the migration chain, is actually filled to spilling
outside the study area. point, our simulations do not show that this is the
Cause (a) definitely has influenced the present- case. This may mean that the estimates of expulsion
day result, suggesting in view of (b) that the partial efficiency or migration distance in this study are too
opening of Draugen did not actually occur. A higher conservative.
supply rate might, nevertheless, be realistic, since the It is realised that this picture would have been
present estimate of the expulsion/generation ratio of modified substantially if source-rock quality and dis-
0.1 to compensate for the presence of the Melke tribution maps had been available, and if allowance
Formation and losses in the overlying Cretaceous had been made for the complications of leakage,
sequences is very low, and the potential supply space- and time-dependent carrier-bed parameters
from outside the map area, notably the south, could and realistic hydrocarbon phase behaviour. Never-
be substantial. theless, in essential points the modelled present-day
An additional remark to be made in this respect is distribution of oil column heights and migration
that owing to limited interpreted seismic data avail- routes is found to agree with available field data.
able at the time the depth maps were prepared for While the model clearly is an oversimplification of
this study, the accuracy in the northeastern and east- the actual situation, even in its present form, it
ern part of the map is indeed low. This is particularly has proven its usefulness for assessing the effects
Modelling secondary hydrocarbon migration in Haltenbanken, Norway 323

of essential migration-system parameters (such as dictive tool in hydrocarbon exploration). In: A.M. Spencer
source-rock expulsion rate, carrier-bed permeability, (Editor), Habitat of Hydrocarbons on the Norwegian Con-
tinental Shelf. Norwegian Petroleum Society, Graham and
oil viscosity and accuracy of palaeo-structure maps) Trotman, London, pp. 259-274.
on migration and accumulation patterns. Hobson, G.D., 1954. Some Fundamentals of Petroleum Geology.
Oxford University Press, London, 139 pp.
References Hvoslef, S., Larter, S.R. and Leythaeuser, D., 1988. Aspects of
generation and migration of hydrocarbons from coal-bearing
Archie, G.E., 1950. Introduction to petrophysics of reservoir strata of the Hitra Formation, Haltenbanken area, offshore
rocks. Am. Assoc. Pet. Geol., Bull., 34: 943-961. Norway. In: L. Mattavelli and L. Novelli (Editors), Advances
Bear, J., 1972. Dynamics of Fluids in Porous Media, Elsevier, in Organic Geochemistry 1987. Org. Geochem., 13: 525-536.
New York, N.Y., 764 pp. Jones, R.W, 1981. Some mass balance and geological constraints
Berg, R.R., 1975. Capillary pressures in stratigraphic traps. Am. on migration mechanisms. Am. Assoc. Pet. Geol., Bull., 65:
Assoc. Pet. Geol., Bull., 59: 939-956. 103-122.
Bj0rlykke, K., Aagaard, P., Dypvik, H., Hastings, D.S. and Klomp, U.C. and Wright, P.A., 1989. A new method for the
Harper, A.S., 1986. Diagenesis and reservoir properties of measurement of kinetic parameters of hydrocarbon genera-
Jurassic sandstones from the Haltenbanken area, offshore mid tion from source rocks. In: B. Durand and F. Behar (Editors),
Norway. In: A.M. Spencer (Editor), Habitat of Hydrocarbons Advances in Organic Geochemistry 1989. Org. Geochem., 16:
on the Norwegian Continental Shelf. Norwegian Petroleum 49-60.
Society, Graham and Trotman, London, pp. 275-286. Krooss, B.M. and Leythaeuser, D., 1988. Experimental measure-
Boussinesq, J., 1904. Recherches theoretiques sur l'ocoulement ments of the diffusion parameters of light hydrocarbons in
des nappes d'eau infiltrees dans le sol et sur le debit des water-saturated sedimentary rocks II. Results and geo-
sources. J. Math. Pures Appl., 10: 5-78 and 363-394. chemical significance. Org. Geochem., 12: 91-108.
Bukovics, C. and Ziegler, P.A., 1985. Tectonic development of Krooss, B.M., Leythaeuser, D. and Schaefer, R.G., 1988. Light
the Mid-Norway continental margin. Mar. Pet. Geol., 2: 2-22. hydrocarbon diffusion in a caprock. Chem. Geol., 71: 65-76.
Carman, P.C., 1937. Fluid flow through granular beds. Inst. Latil, M., 1980. Enhanced Oil Recovery. Editions Technip, Paris,
Chem. Eng. Trans., 15: 150. 236 pp.
Chilingar, G.V., 1963. Relationship between porosity, perme- Lehner, F.K., Marsal, D., Hermans, L. and Van Kuyk, A.,
ability, and grain size distribution of sands and sandstones. 1987. A model of secondary hydrocarbon migration as a
In: L.M.J.U. van Straaten (Editor), Deltaic and Shallow buoyancy-driven separate-phase flow. In: B. Doligez (Editor),
Marine Deposits. Developments in Sedimentology, Vol. 1. Migration of Hydrocarbons in Sedimentary Basins. Editions
Proceedings of the International Sedimentology Congress, Technip, Paris, pp. 457-471. (Reprinted in: Rev. Inst. Fr.
Amsterdam-Antwerp. Elsevier, Amsterdam, pp. 71-75. Potrole, 1988, 43: 155-164.)
Cohen, M J . and Dunn, M.E., 1986. The hydrocarbon habitat of McAuliffe, C D . , 1979. Oil and gas migration chemical and
the Haltenbank-Traenabank area, offshore Mid-Norway. In: J. physical constraints. Am. Assoc. Pet. Geol., Bull., 63: 761-
Brooks and K.W. Glennie (Editors), Proceedings of the 3rd 781.
Conference on Petroleum Geology of NW Europe, Vol. 2. McKenzie, D., 1978. Some remarks on the development of
Graham and Trotman, London, pp. 1091-1104. sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32.
Dembicki, H., Jr. and Anderson, M.J., 1989. Secondary migration Neglia, S., 1979. Migration of fluids in sedimentary basins. Am.
of oil: experiments supporting efficient movement of separate, Assoc. Pet. Geol., Bull., 63: 573-597.
buoyant oil phase along limited conduits. Am. Assoc. Pet. Pedersen, O.E., 1988. Norwegian offshore fields reserves and
Geol., Bull., 73: 1018-1021. production. Scand. Oil Gas Mag., 16: 25-30.
Dodson, C.R. and Standing, M.B., 1944. Pressure volume tem- Price, L.C., 1976. Aqueous solubility of petroleum as applied to
perature and solubility relations for natural gas-water mix- its origin and primary migration, Am. Assoc. Pet. Geol., Bull.,
tures. Drill. Prod. Prac, API, pp. 173-179. 60: 213-244.
Du Rouchet, J.H., 1981. Stress fields, a key to oil migration. Price, L.C., 1981, Aqueous solubility of crude oil to 400C and
Am. Assoc. Pet. Geol., Bull., 65: 74-85. 2000 bars pressure in the presence of gas. J. Pet. Geol., 4:
Ellenor, D.W. and Mozetic, A., 1986. The Draugen oil discovery, 195-223.
In: A.M. Spencer (Editor), Habitat of Hydrocarbons on the Schowalter, T.T., 1979. Mechanics of secondary hydrocarbon
Norwegian Continental Shelf. Norwegian Petroleum Society, migration and entrapment. Am. Assoc. Pet. Geol., Bull., 63:
Graham and Trotman, London, pp. 313-316. 723-760.
England, W.A., Mackenzie, A.S., Mann, D.M. and Quigley, T.M., Tissot, B.P. and Weite, D.H., 1984. Petroleum Formation and
1987. The movement and entrapment of petroleum fluids in Occurrence. Springer-Verlag, Berlin, 2nd ed., 699 pp.
the subsurface. J. Geol. Soc. London, 144: 327-347. Ungerer, P., Bessis, F., Chenet, P.Y., Durand, B., Nogaret, E.,
Fchtbauer, H., 1967. Influence of different types of diagenesis Chiarelli, A., Oudin, J.L. and Perrin, J.F., 1984. Geological
on sandstone porosity. Proc. 7th World Pet. Congr., 2: 353- and geochemical models in oil exploration. In: G. Demaison
369. and R J . Murris (Editors), Petroleum Geochemistry and Basin
Gussow, W.C., 1954. Differential entrapment of oil and gas: Evaluation. Am. Assoc. Pet. Geol., Mem., 35: 53-77.
a fundamental principle, Am. Assoc. Pet. Geol., Bull., 38: Wooding, R.A. and Chapman, T.G., 1966. Ground water flow
816-853. over a sloping impermeable layer, 1. Application of the
Heum, O.R., Dalland, A. and Meisingset, K.K., 1986. Habitat of Dupuit-Forchheimer assumption. J. Geophys. Res., 71: 2895-
hydrocarbons at Haltenbanken (PVT-modelling as a pre- 2902.

L. HERMANS Koninklijkel Shell Exploratie en Produktie Laboratorium, Volmerlaan 6, 2288 GD Rijswijk, The Netherlands
A.D. van KUYK Koninklijkel Shell Exploratie en Produktie Laboratorium, Volmerlaan 6, 2288 GD Rijswijk, The Netherlands
FK. LEHNER Koninklijkel Shell Exploratie en Produktie Laboratorium, Volmerlaan 6, 2288 GD Rijswijk, The Netherlands
P.S. FEATHERSTONE Koninklijkel Shell Exploratie en Produktie Laboratorium, Volmerlaan 6, 2288 GD Rijswijk, The Netherlands
 325

Faulting processes and fault seal

R.J. Knipe

The understanding of fault seal generation is crucial to the assessment of the migration and trapping of hydrocarbons. The
effectiveness of the seal depends upon the porosity and permeability characteristics of the fault zone, which is controlled by the
microfabrics present. The range of processes which effect the microfabric evolution and the development of seals in fault zones
are reviewed. The examples presented illustrate the use of microstructural analysis for the characterisation of seal properties.
The importance of cyclic deformation in fault zones, where different deformation mechanisms operate during different parts of
these displacement events, and the role of these mechanisms in producing different types of seal is emphasized. Three broad
classes of seal may be recognised: (1) collapse seals, where the permeability/porosity of the fault zone is reduced by pore volume
decrease achieved by grain reorganisation produced by fracturing, grain deformation and dissolution and by grain boundary
sliding; (2) cement seals, where the reduced flow across a fault is achieved by the precipitation of phases in and adjacent to the
fault; (3) juxtaposition seah, generated be the juxtaposition of lithologies with different sealing capacities. All these seal types
can contribute to the sealing properties of a single fault and their distribution can be related to the larger fault zone geometry
and displacement pattern. Cementation patterns and histories in fault zones can be complex, but analysis of the deformation
microstructures and cement sequences in fault zones provides crucial data on the reservoir evolution. This data also helps
constrain the conditions, timing and composition of fluid flow events associated with different fault zones. Such analysis also
provides important information on the larger scale basin drainage patterns. The sealing capacity of different seals is calculated
and the stability of different types of seal during the subsequent geological history is discussed. The examples presented highlight
the need for future programmes to combine microstructural studies of fault seal with the analysis of the larger scale fault
geometry and the diagenetic evolution in fault zones and reservoirs. Such an integration is needed to understand and quantify
the interaction between reservoir development and faulting.

Introduction such as the pore geometry are known to be crucial

to sealing properties, but there are few studies (e.g.,
Faulting forms one of the main controls on the dis- Pittman, 1981) which have described and assessed
tribution, size and internal structure of hydrocarbon the porosity and permeability development of fault
reservoirs. The role of fault zones in providing both zones in detail.
a pathway for secondary hydrocarbon migration and This paper outlines the microprocesses associated
a seal to hydrocarbon accumulations has been recog- with seal evolution by reviewing the deformation pro-
nised for a long time (Hubbert, 1953; Smith, 1966 cesses which operate in faults and affect the sealing
and 1980). Despite the detailed studies conducted, properties of fault zones. This information is used
the assessment of the sealing properties of faults to assess the types of seal which can develop under
remains an important goal for hydrocarbon explo- different conditions, the sealing capacity of faults,
ration and production. Previous studies of fault seal the continuity of effective seals and the strength and
have concentrated on the properties of the reservoir stability of different seals. The final part of the paper
rocks and barrier lithologies juxtaposed by the fault- discusses the importance of linking information on
ing (Smith, 1966 and 1980; Berg, 1975; Schowalter, the micro-processes of faulting with other aspects
1979; Watts, 1987) or on the structural style or setting of reservoir evolution, particularly the larger scale
(Harding, 1974; Harding and Tbminas, 1988, 1989). geometrical development of fault arrays.
Very little attention has been paid to the processes
which produce fault rocks and the possible influence
Controls on fault seal
of these on sealing properties. This contrasts with
the amount of research aimed at assessing the role of Sealing may be defined as the reduction in the flow
fractures in producing high permeability reservoirs potential or permeability across a barrier provided by
(Nelson, 1985; Wilke et al., 1985; Charlaix et al., an adjacent medium juxtaposed either stratigraphi-
1987; Gueguen et al., 1987). Microstructural features cally or by faulting. For sealing of hydrocarbons

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 325-342. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
326 RJ. Knipe

to occur, the buoyancy forces associated with the The displacement pressure, dp, or the mean in-
difference between the density of water and hy- jection pressure is that which causes displacement of
drocarbons must be matched or exceeded by the the major part of the wetting fluid and can be defined
capillary pressure or displacement pressure of the as (Berg, 1975, eqn. 8):
seal (Hubbert, 1953; Smith, 1966 and 1980; Berg,
1975; Schowalter, 1979; Watts, 1987). The height of (21 cos )
dp = (2)
a hydrocarbon column which can be supported at
a sealing medium depends upon (1) the difference where r is a radius of curvature determined by the
between the displacement pressure characteristics of pore size of the rock.
the sealing medium and the reservoir, (2) the dif- Because the displacement pressure is dependent
ference between the density of water and the oil upon the pore geometry factor, r, it is also possible
under consideration, and (3) the hydrodynamic or to present an equation for the height of trapped
hydrostatic conditions (Fig. 1). A number of authors hydrocarbons in the hydrostatic situation in terms of
have provided analyses of the relationships between the pore geometry, e.g., Berg (1975, eqn. 6):
these variables and the sealing capacity or height of
the associated hydrocarbon column (Smith, 1966 and 2/ h (r t - 1 -r- 1 )
1980; Berg, 1975; Schowalter, 1979; Watts, 1987). In (3)
most cases the approach has been to measure the g(Pw-Ph)
capillary pressure of core material in the laboratory where rt is the pore throat radius in the seal and rp is
and to use this information in conjunction with data the average pore radius in the reservoir.
on the hydrocarbon and water properties to calcu- It is clear from these equations that the parame-
late the heights of hydrocarbon columns which may ters rt and rp are fundamental to the sealing process
be present at depth. A number of equations have and are controlled by the texture of the seal whether
been presented to aid the estimation of the trapped it is a stratigraphic unit or a fault zone. The fac-
column height (Z). Watts (1987), for example, based tors which influence the evolution of the grain size,
on the analysis of Smith (1966) used the following shape and sorting together with the characteristics
relationship for hydrostatic situations: of the cement phases present are fundamental to
the pore size, pore throat size and permeability of
Pdmh the sealing medium and, thus, to the sealing proper-
Z = (i)
(Pw-Ph)gImCOS<f) ties. Assessment of the diagenetic evolution of these
parameters during burial is common and measure-
where Pdm is the mercury/air entry (or displacement)
ments to quantify these parameters are achieved
pressure, h is the hydrocarbon-formation water in-
via laboratory measurement of undeformed sam-
terfacial tension, Pw and P h are the densities of the
ples (e.g., Wardlaw and Cassan, 1978; Bj0rlykke et
formation water and hydrocarbon phase, g is the
al, 1986). Analysis of the porosity evolution during
acceleration due to gravity, Im is the mercury-air in-
deformation associated with compaction have also
terfacial tension, and is the contact angle between
been conducted (e.g., Houseknecht, 1988; Wilson
mercury and air.
and McBride, 1988). There is a need to expand these
studies to include characterisation of fault zone ma-
terial in order to assess the evolution of fault seals
SEAL CONTROLS and to provide data which will allow estimation of the
sealing capacity of fault zones. Only via an integrated
analysis of "fault zone diagenesis" and of the oper-
Height of Trapped Hydrocarbon Column
ating deformation mechanisms in fault zones can the
seal evolution be fully incorporated into assessment
of the reservoir development.
Displacement Pressure Density Difference between
between Seal and Formation Water and
Reservoir. Hydrocarbon phase.
1 Faulting processes and fault seal

- Pore geometry. Chemistry of Fluid This section presents a review of the deformation
Phases. mechanisms possible in fault zones which are impor-
- Interfacial Tension. Temperature.
tant to sealing. The aim is to provide examples of
Pressure
- Wettability.
the range of microfabrics present in fault zones and
to discuss the range of processes which contribute
Fig. 1. Summary of controls of fault sealing. to the evolution of these fabrics. The information
Faulting processes and fault seal 327

presented is then used to assess the types of seal deformation mechanisms (Borradaile, 1981; Knipe
developed under different faulting conditions and 1986b; Moore et al., 1986; Jones and Preston, 1987;
to estimate the sealing capacity of faults. Particular Owen, 1987). Low confining pressures and high fluid
emphasis is placed on the evolution of the porosity, pressures favour the operation of these modes of
permeability and pore geometry during fault rock deformation. The permeability and porosity of the
evolution. aggregate can be transiently increased during the
The deformation mechanisms which can oper- deformation by the dilation which accompanies frac-
ate in fault zones depend upon the interaction of turing and frictional sliding.
the lithological characteristics (mineralogy, texture, Recent research has established the microstruc-
porosily, permeability) and the environmental condi- tural features which characterise the deformation
tions (stress level, confining pressure, fluid pressure, mechanisms listed above (reviewed in Groshong,
temperature, etc.). Deformation mechanisms have 1988; and Knipe, 1989). Although understanding of
been reviewed by Sibson (1977, 1986b), Groshong the details of each process involved in these mech-
(1988), Mitra (1988), and Knipe, (1986a, 1989). The anisms is still incomplete, identification of the type
mechanisms can be divided into three categories: of deformation involved is possible. Because the
(A) Diffusive mass transfer (DMT), where diffusion microstructural changes associated with each mech-
of material away from sites of high stress produces anism are different, the type of sealing arising from
deformation (Rutter, 1983; Gratier and Guiguet, the operation of each also differs. Establishing the
1986). The deformation is a three-stage process in- deformation mechanisms operating in different faults
volving dissolution, diffusion or transportation of is therefore critical to the assessment of the develop-
dissolved species, and precipitation. The mechanism ment of seals.
is favoured by small grain sizes, small grain contacts The accumulation of displacement in fault zones
and in situations where an efficient diffusion path in the upper crust ranges from the repeated tran-
medium is present. Where water acts as the diffusion sient, high displacement rate events, i.e., stick-slip
path the process is termed pressure solution (see behaviour (Frank, 1965; Brace and Byerlee, 1966;
Rutter, 1983; and Gratier and Guiguet, 1986 for re- Ohnaka, 1987), to continuous displacement or fault
views) and is important in compaction, cementation creep behaviour (Wesson, 1988). There is a complete
and lithification of sediments (e.g., Angevine and spectrum of fault displacement patterns between
Turcotte, 1983; Houseknecht, 1984,1988; Wilson and these two end members. Cyclic changes in defor-
McBride, 1988). mation mechanisms which accompany faulting are
(B) Crystal plasticity involves the deformation of therefore common and evidence for the cyclic alter-
crystal lattices by intragranular processes such as nation of the different mechanisms can be found in
dislocation motion and twinning. These mechanisms the fault rock products (Sibson 1980, 1986a; White
dominate the deformation of minerals under high and White, 1983; Mawer and Williams, 1985; Knipe,
grade conditions (see Barber, 1985; White, 1985; and 1986a, b, 1989; Underhill and Woodcock, 1987; Cox
Knipe, 1989 for reviews), but are also important to and Etheridge, 1989). Fig. 2 presents an example
the bending and kinking of phyllosilicates present of a deformation mechanism path associated with
in sediments and can play an important role in low different fault displacement events. The importance
temperature ductile fracture (see Lloyd and Knipe, of these cyclic events to the present discussion is
1992). that different deformation mechanism paths will pro-
(C) Frictional sliding, fracture processes and cat- duce different types of seal. The stability, strength
aclastic flow. In this case deformation involves a and effectiveness of the seal depends upon the de-
large group of processes ranging from fracture prop- formation mechanism paths experienced. In addition
agation, frictional sliding on newly created surfaces the deformation mechanism paths experienced will
and independent paniculate flow, where deforma- depend upon the lithological characteristics of the
tion is achieved by the frictional sliding of grains material undergoing faulting, and the conditions of
without grain breakage. Fracturing involves a large faulting. We can expect different deformation mech-
range of processes including brittle fracture and sub- anism histories at different depths and at different
critical fracture processes, possibly below the stress locations along individual faults because different
levels needed for brittle fracture (see Logan, 1979; strain-rate and stress histories are likely in the fol-
Kranz, 1983; Wang, 1986; Tbllis, 1986; Atkinson, lowing situations: (a) where different lithologies are
1987; Kozak, 1987; and Knipe, 1989 for reviews). in contact with the fault plane; (b) where different
Grain boundary sliding or independent particulate propagation processes operate; (c) where restricted
flow dominates the flow of unlithified material or fluid migration pathways cause differences in the
material which has been disaggregated by other cement distribution associated with faults; and (d)
328 RJ. Knipe

Cataclasite growth of new phases within and adjacent to fault

^development
zones. Asphalt or tar impregnations are included as
'*,;- ';" cement seals here.
Fracturing % / \
STRAIN (C) Juxtaposition seals. Seals produced by the jux-
RATE
taposition of two lithologies with different capillary
^-^ Disloc. / pressures. In this case the seal characteristics are
A / \ Motion /
largely independent of the faulting processes and
the fault only serves to juxtapose the two lithologies
/ \ DMT /
with either different pore geometries and/or fluid
pressures (see Hubbert, 1953; Downey, 1984).
Fault seals usually form the lateral seal to a hy-
TIME drocarbon accumulation where a cap (stratigraphic)
Fig. 2. A deformation mechanism path illustrating the variation seal is present above the reservoir. In addition, fault
in the deformation mechanisms which can dominate different
deformation events. Note also the change in the distribution of
seals are likely to be dynamic features where there
deformation mechanisms which can arise from changes in the is hydrocarbon flow in and leakage from the reser-
rock fabric which are induced by some deformation events; in voir (e.g., Harding and Tbminas, 1988). Juxtaposition
this case the DMT (diffusive mass transfer) field is expanded as seals are the most familiar type of fault seal (Smith,
a result of the grain size reduction which results from cataclasis.
1966, 1980, Berg, 1975, Downey, 1984; Allen, 1988).
However, there is an increasing recognition of the
importance of fault seals developed where the juxta-
posed lithologies are permeable (Weber et al., 1978;
where fault geometry causes different strain histories Smith, 1980; Harding and Tlxminas, 1989). The seal-
in and around the fault. Even in situations where the ing properties of a single fault may in part be due
displacement is continuous then important changes to all these factors if the range of conditions and
in the deformation mechanisms and the fault rock lithologies which are involved in the faulting are
microfabric can be associated with the final stages large. However, different types of fault rock and
of displacement or with post-displacement accom- therefore characteristic seals are likely to develop in
modation adjustments in the fault zone. The next different situations. To illustrate this, the next part of
sections of this paper outlines the different seals the paper discusses examples of the likely evolution
which may be produced by different deformation of pore geometry and permeability during faulting
mechanism histories. The influence of fault geometry in sandstones using examples of the microstructures
is discussed later in the paper. preserved in such zones. The information obtained
from the fault rock microstructure analysis has a two-
Seal types fold usage: (a) information on the grain size, shape
The last section emphasised the cyclic nature of and sorting together with the information on the
deformation processes which accompany faulting and pore geometry can be used to estimate the sealing
the likelihood that a combination of processes will capacity of the zone; and (b) the texture and min-
contribute to the final sealing properties of the fault eralogy of the phases present as clasts and cement
zone. This section reviews the different seals which phases provide important data on the conditions of
can develop by different deformation mechanism the fault rock evolution which can be related to the
paths by presenting some examples of faults with burial history and the diagenetic history of adjacent
different microstructures and different sealing char- lithologies. Both these aspects are discussed in more
acteristics. detail after the microstructures and the deformation
Fault seals may be grouped into three broad cate- mechanism histories in the chosen examples have
gories. been described.
(A) Porosity collapse seals. Changes in the perme- Figure 3 shows the microstructure of a cataclasite
ability, porosity and sealing potential within the fault derived from a sandstone. The original grain size
zone are induced by changes in the pore volume, (150-175 /im) is reduced by almost two orders of
shape and size by grain refinement and rearrange- magnitude in the fault zone by cataclasis to frag-
ment associated with frictional sliding and/or the ments with a grain size of 2-7 and the final tex-
cataclasis, dissolution, alteration and plastic defor- ture retains some (5-10%) porosity. The fault rock is
mation of fragments/grains within the fault zone. likely to have evolved by repeated deformation (frac-
Smear seals would fall into this category. turing) events. Transient, high permeability periods
(B) Cement seals. Changes in the permeabil- are likely to have been separated by low permeability
ity, porosity and sealing potential arising from the periods between the deformation events (Fig. 4). The
Faulting processes and fault seal 329

iJfPlbj;.

849 CflTfl 00011

Fig. 3. Secondary electron image of the microstructure of a cataclasite derived from a sandstone. The original grain size (150-175
) is reduced in the fault zone by a factor of 20-75 to approximately 2-7 . Note the well formed crystal faces present indicating
precipitation of overgrowths.

high permeability periods when the fault can act as a the material in the first few centimetres of displace-
fluid migration pathway are considered to represent ment. The rate of grain size reduction is important
periods when fluid pressure adjustments take place. to the sealing potential of the zone, because the
The longer the fluid pressure variations or the stress porosity of the aggregate will depend upon the grain
can be maintained at levels which allow large pore sorting in the fault zone. However, there are two
throat diameters to persist, the larger the migration reasons why the sorting characteristics of the fault
potential will be. Further discussion on this aspect zone may be secondary to the pore throat diameter
is provided in the section on seal continuity below. with respect to the sealing potential. Firstly, catacla-
At the end of each cataclastic deformation event the sites often develop by the progressive increase in the
permeability in the fault zone is reduced by a poros- proportion of fine grained cataclastic matrix, where
ity collapse associated with both the change in size of the matrix fragments are often of a restricted size
the pore diameter and pore throat diameter arising range at least one order of magnitude smaller than
from the grain size reduction and from the closing of the original grains. That is, the grain size evolution in
pores during the drop in fluid pressure. The idealised cataclasites is essentially a bimodal one in which the
history show in Fig. 4 shows only a small number of proportion of the fine grained component increases
the events which are likely to occur in fault zones with deformation. Frser (1953) has illustrated that
during their activity periods (see Watterson, 1986; the permeability in a bimodal mixing of two uniform
Sibson, 1989) and also presents an example in which grain size populations approaches that of the fine
the grain size reduction is progressive. Analysis of the grained component when 70-75% of the aggregate is
grain size reduction processes in a faulted quartzite still composed of the coarse grained component. The
by Lloyd and Knipe (1990) demonstrates that grain second reason why sorting may be secondary is the
size reduction can occur during the very early stages known small dependence of permeability on sorting
of displacement and produce a reduction in >75% of compared to the influence of sorting on porosity, i.e.,
330 RJ. Knipe

Fracture events. poorly sorted material may contain large pores, but
these are not interconnected.
The sealing properties in the example shown in
Figs. 3 and 4 will be primarily produced by the
collapse of the pore volume and of the pore throat
size during grain size reduction by cataclasis. The
sealing capacity of this type of fault rock will be
estimated in a later section.
The second example shown in Fig. 5 illustrates
Time
the effect of late stage modification (compaction) of
. Fracture events. a cataclastic microfabric by diffusive mass transfer
(DMT) processes. Such effects emphasise the sus-
ceptibility of the fine grained cataclasite to localised
grain dissolution. The change to a deformation dom-
inated by DMT may represent the last stages of
displacement accommodation in the zone when both
V_J
LJ the decreasing strain-rate history and the fine grain
size favour this deformation process. The result is the
production of a compacted microstructure where the
porosity is rapidly decreased by both dissolution and
Time
re-precipitation of cement. The relative proportion
Fig. 4. A possible history of the fault rock shown in Fig. 3 in of dissolution and precipitation can vary so that in
terms of permeability and pore throat evolution. The fracture
events are marked by transient high permeability events which
some cases material is removed but not precipitated
are separated by "rest" periods when the permeability and in the cataclasite. The decrease in the strain rate and
effective pore throat size is reduced. the operation of DMT in a fault zone may be

<^:&:M3mM

I:, IIP

:1;
11% :;i
- -f^mm^m I I

M|.
'<$&m
.** y^

Crtt CRTfl Zyn

Fig. 5. Secondary electron image of a compacted cataclasite. The compaction post-dated the grain size reduction in the fault zone and
represents the last stages of strain accommodation. Note the interlocking nature of the grains and the reduction in porosity compared
with Fig. 3.
Faulting processes and fault seal 331

Fig. 6. Backscattered electron image of a fault zone where many individual grains were not fractured during the deformation. The
deformation is achieved by grain disaggregation and frictional grain boundary sliding within the fault zone (orientated top left to
bottom right in the micrograph). Note the reduced porosity in the fault zone which is three or four grains wide.

related to either: (1) a "rest" period between tran- lution of early cements to produce secondary poros-
sient faulting events (stick-slip behaviour); (2) the ity may also be prone to this type of deformation. If
changing pattern of fault activity on a larger scale to this disaggregation takes place, the fault zone may
one in which during the fault array evolution a new experience a transient increase in porosity and per-
fault takes up the extension and previously formed meability during the faulting, the final post-faulting
faults become inactive; or (3) the end of tectonic microstructure may be similar to the pre-faulting
activity evolution in the area, when activity on all structure of grain arrangement. In the simplest case
faults decreases. In the first case the compacted cat- the sealing properties of the rocks should not be
aclasite is likely to be fractured by the subsequent affected by the faulting. However, as the example in
faulting events. The important point in all three cases Fig. 6 illustrates, porosity collapse by some fracturing
is that the sealing capacity of a fault can be enhanced and grain boundary sliding can accompany the de-
by the slow deformation which post-dates the rapid formation and induce changes in both porosity and
displacement events on the fault and may also be permeability. In addition, microstructural analysis of
enhanced by the continued burial of an inactive fault. fault zones developed in sandstones which are only
Figure 6 illustrates the microstructure of a differ- partially cemented and have suffered only limited
ent type of fault zone where many individual grains diagenesis reveal that seals can develop in the border
are not fractured during the deformation. The defor- zone to the fault. Figure 7 shows an example with
mation is achieved primarily by grain boundary frac- collapsed and deformed phyllosilicate aggregates in
turing, disaggregation and frictional grain boundary the pore throats adjacent to the fault zone. It is
sliding. Such fault zones are likely to occur in par- suggested here that fluid flow pulses associated with
tially or weakly lithified material and will be favoured faulting may cause collapse and redistribution of the
by deformation during high fluid pressures, low con- clay phases into the pore throats. This pore throat
fining pressures and at shallow depths of burial. It is clogging or "wash-in" seal can occur in the border
interesting to note that lithologies undergoing disso- zones of a fault zone where displacement or strain
332 RJ. Knipe

Fig. 7. TEM of aligned and bent phyllosilicates located in a pore throat adjacent to a fault zone. The phyllosilicates are adjacent to a
carbonate overgrowth (carb. overgth.) on a detrital grain.

associated with the fault is almost non-existent and is to the large scale channelling of mud diapirs along
one type of the adjustment to the fault zone borders fault zones. The disaggregation and redistribution of
which can contribute to the sealing properties of the clay particles during faulting can also lead to more
fault zone (Fig. 8). In this case the large changes
in permeability can be achieved without significant
changes in the porosity; only the pore geometry
changes by the redistribution of material. y
A more dramatic concentration of phyllosilicates
into a fault zone is illustrated in Fig. 9. In this
case detrital quartz grains in the host rock are frac-
tured and appear to be corroded in the fault zone.
The phyllosilicate concentration is primarily associ-
Redistribution of
ated with smearing, reorientation and deformation cement by Fluid
of detrital phyllosilicates. Porosity collapse in fault Flow.
zones developed in clay rich lithologies has also been
discussed by Knipe (1986c) and Moore et al. (1986). Fig. 8. "Wash-in seal": Sketches illustrating the development of a
It should be clear from the last two examples dis- collapsed phyllosilicate aggregate in a pore throat, (a) The initial
configuration of clay plates distributed around a pore between
cussed that a complete range of "wash-in" seals can detrital grains, (b) Final fabric of phyllosilicates blocking a pore
develop from those associated with the redistribu- after a deformation event where a fluid flow pressure wave has
tion of phyllosilicates on the scale of a few grains redistributed the grains.
Faulting processes and fault seal 333

Fig. 9. Backscattered electron image of a fault zone where clays are concentrated in the fault zone (top left to middle bottom of
micrograph). Note the high porosity outside the fault and the corroded quartz grains inside the fault zone.

extensive "smear" seals than is predicted by the sep- tions where rapid fluid influx and pressure drops can
aration of shale lithologies after displacement. This induce preferential precipitation.
is achieved by phyllosilicates from the shale unit be- The dilation magnitude, together with the fluid
ing carried along the fault zone by flow processes pressure and permeability changes during faulting
operating during the transient high permeability and events will influence the volume of cement produced.
disaggregation periods. For example, fault zones where both dilation and
TWo last examples of fault related seals are pre- fluid pressure changes are large can result in very
sented in Figs. 10 and 11. In these cases the pre- rapid precipitation and the production of a brec-
cipitation of cements along and adjacent to the cia composed of rock fragments floating in cement.
fault zone causes the changes in the pore geometry These have been termed implosion breccias by Sib-
and permeability which controls the sealing capacity son (1986) and related to the collapse of the wall
of the fault. The precipitation of cements can be rock into dilation sites generated by rapid slip events.
caused by changes in the fluid pressure and tem- On the other hand, situations in which the fault zone
perature experienced by fluids during the faulting dilation at the time of precipitation is small and
process. A detailed analysis of the fluid pressure the fragments remain in contact and provide a load-
cycles associated with faulting has been presented supporting framework, represent the other end of
by Sibson (1981, 1985, 1989) and the role of such the spectrum, where precipitation of cement creates
cycles in regional deformation is discussed by Cox overgrowths in the pores of the cataclasite. Whatever
and Etheridge (1989). Sibson (1981, 1985, 1989, the detailed mechanism of cement generation the re-
1990) presented models for the behaviour of faults duction in the porosity and permeability in the fault
as valves which control the channelling of fluids in zone can be large. For example, in sandstones initial
fault zones and pumps which cause expulsion of the porosities of 20% may be reduced to less than
accumulated fluids. Sibson also emphasises the role 1% and permeabilities can be decreased by more
of dilation sites created along the fault zones during than an order of magnitude (see also Gabrielsen and
the displacement events (e.g., dilational jogs) as loca- Koestler, 1986).
334 RJ. Knipe

Fig. 10. SEM of kaolinite grains filling pores within a fault zone.

Although the sealing processes in each of the ex- Seal timing

amples presented above is dominated by a small The timing of the seal development relative to hy-
number of factors it should be emphasised that drocarbon maturation and migration timing is crucial
individual faults may contain elements of a num- in the assessment of reservoir evolution. Integrat-
ber of seal types. For example, the development of ing the diagenetic history and burial histories of the
sealed fault zone borders by concentrated cemen- source and reservoir rocks with an assessment of
tation and/or by pore throat plugging may occur the faulting mechanisms can provide important con-
adjacent to a fault zone where cataclasis, dissolution straints on the development of reservoir properties.
and cement precipitation characterise the internal Figure 13 presents the results of an analysis which
fault zone structure (Fig. 12). However, it should links the timing of faulting with burial in a sandstone
also be clear that different deformation mechanisms reservoir. The figure is based upon the work con-
will dominate the deformation under different con- ducted by RJ. Knipe and G.E. Lloyd (1988, unpub-
ditions and produce seals with characteristics which lished). Each of the faults studied has a distinctive
reflect the deformation processes and conditions. deformation microstructure and cement evolution
Fault sealing at shallow depths, for example, where which can be linked to the burial history of the area.
the lithification is only partially complete and the Linking the reservoir diagenesis with the "fault zone
material strength is low, is more likely to be asso- diagenesis" by establishing the sequence of cement
ciated with precipitation of cements adjacent to the growth and deformation within and external to the
fault zone margins or by the injection of material fault zones present has provided information on: (a)
into the fault zone. At deeper levels with increased the timing of fault activity during burial and the
amounts of diagenesis the possibility of combined identification of faults active over long periods of
cataclasis, grain dissolution and precipitation pro- reservoir evolution and those active over short time
cesses contributing to the seal increases. periods; (b) the distribution of faulting (i.e., which
Faulting processes and fault seal 335

Fig. 11. TEM of a composite cement present in a cata clastic fault zone. The original quartz fragment (H) is enclosed in an inclusion
rich quartz cement (C). Phyllosilicates (illites) form the second cement infill between the quartz overgrowths.

faults were active in different parts of the basin); Sealing capacity

and (c) a link between the distribution of reservoir The data on the pore microstructure presented
cements and fault activity. in the examples described above are used in this
section to estimate the sealing capacity of faults asso-
ciated with different seal types. The sealing capacity
Fault Zone or the height of a hydrocarbon column which these
seals could support is calculated using the equation
Border Zone Seals.
[Cement and throat (eqn. 3 above) presented by Berg (1975). Values
seals]. for the pore size and pore throat diameters used
are based on those measured on mercury injection
testing, SEM, TEM micrographs and from published
data (e.g., Pittman, 1981). The values of the other
important variables, which will influence the sealing
capacity (i.e., the fluid and hydrocarbon density and
Cement Seals in Fault
the hydrocarbon interfacial tension), used in these
Zone. estimates enclose the range commonly encountered
Porosity Collapse
Seal. in reservoirs (and discussed in detail by Schowalter,
1979; and Watts, 1987). The results of the estimates
Fig. 12. Fault seal summary illustrating a composite seal where
domains within the fault have different characteristics and a
of seal capacity are shown in Fig. 14. The calcula-
border zone seal outside the movement zone is developed to tions illustrate the differences between the sealing
different thicknesses in different layers. capacities of the different types of seal discussed
336 R.J. Knipe

Seal continuity
TR ASSIC JURASSIC |
F4>\ 1
The distribution of seals along faults is known to
\V F 1 ' 2 be a fundamental control on the trapping potential of
Depth.
faults. There are a number of factors and processes
which will cause the type of seal to vary not only
[feet] F3,8Vv between different faults but also along individual
5,000 A faults. These controls include: (a) the distribution
of lithologies intersecting with the fault planes; (b)
the propagation and displacement mechanisms of
the faults, i.e., the deformation mechanism histories
1 1 N. s of different faults or fault segments; (c) the nature
Oil G e n e r a t i o n . ^ *v s^

10,000
of the fluid flow along the fault and; (d) the fault
Fig. 13. Summary of the analysis of faulting during burial of
plane geometry. These are discussed briefly below
a sandstone reservoir. Different faults (Fl-8) are active for and although the factors are discussed separately,
different parts of the burial history. The range of timing of fault the interrelationships and interdependence of these
activity with respect to burial and overlap with the oil generation factors are emphasised.
period will give rise to different sealing potentials for each type
of fault. (Unpublished study by R.J.Knipe and G.E. Lloyd.)
(a) The distribution of lithologies intersecting
with the fault planes. Each lithology which inter-
sects the fault plane will have different properties
above. The seals which are likely to produce the in terms of strength, reaction to the faulting process
most effective seals are those with the smallest pore and, thus, the type of seal likely to develop under
throat. the faulting conditions. The type of map described

1000
N. ? i i MK Smear,
^Cementation

Cataclasties
^ ^ V N.

100_

Oil
Reservoir
Column
Rocks.
Height. v Fine sst.

[m]. 10 -
N. Course |
Sst.

0.1
1 1

Pore throat radius.

[log cm]

Fig. 14. Plot of sealing capacity of different types of fault seal. The calculations are based on the equation of Berg (1975). The
calculations illustrate the oil column height when the values of interfacial tension range from 30 to 35 dynes/cm and the difference
between the water and oil densities from 0.1 to 0.385. The pore throat radius values are based upon the microstructural and physical
property measurements made by the author and data published in Pittman (1981), Berg (1975) and Frser (1935).
Faulting processes and fault seal 337

zones in which the energy driving the fault propaga-

tion is almost dissipated. This can result in a change
in the fracture process and the accommodation of
strain at a different rate and by different mecha-
nisms. In addition, the tip zones of a main faulting
event can also act as a location where aftershocks
are concentrated (see Cockerham and Eaton, 1987;
Yamashita and Knopoff, 1987; Sibson, 1989). These
factors can promote the concentration of fracturing
I I Shale/shale. ||| Shale/Sst. Q Sst/Silt. and cementation in tip zones.
Ill Sst/Sst. 3 Silt/silt. Fracturing or cataclastic fault rock development is
Fig. 15. Fault plane map showing the distribution of seal types known to be heterogeneous along faults (Brock and
on the fault plane which can arise from the intersection of Engelder, 1977; House and Gray, 1982; Hadizadeh
different lithologies with a fault. The intersections of different and Rutter, 1983; Blenkinsop and Rutter, 1986; Rut-
lithological units from the hanging wall (HW) and from the
ter et al., 1986; Sibson, 1986). This may arise from
footwall (FW) with the fault plane are illustrated. Also shown are
displacement magnitude contours, fine dashed line), including (1) the repeated creation of new tip zones during
the zero displacement, or tip line. Note the distribution of seal fault propagation, (2) the deformation of the wall
types and the zone in which the sandstone cataclasite (CS) and rocks, asperities and perturbations during displace-
the cataclasite derived from the silt overlap (SS).
ment along the fault zone or (3) different faulting
events which are restricted to segments of established
faults zones. Sibson (1986) has already recognised
by Allen (1988) to illustrate the separation and jux- the importance of the range of processes operating
taposition of lithologies along fault planes and to along fault zones in the production of different types
asses the distribution of juxtaposition seals can be of fault breccia. The result of this variation in the de-
modified in order to assess also the likely distribution formation mechanism histories along a fault zone is
of seal types along the fault planes. An example of a the potential to develop different seal characteristics
seal-type map is shown in Fig. 15. As illustrated, the along the fault.
fault plane can be divided into areas in which differ-
ent types of seal will dominate. In addition, there will (c) The nature of the fluid flow along faults.
be transition zones in which the mixing of seal types Fluid flow along fault zones is known to be both
is likely. For example, fault plane areas above the in- episodic and localised, even along individual faults
tersection of a mud/shale with the fault (Fig. 15) may (Sibson, 1981; Etheridge et al., 1984; Sibson et al.,
be susceptible to mud injection and the tip zone (loci 1988). The importance of the fluid pumping to seal
of zero displacement) of the fault may have a lower evolution is that fluids with different compositions
seal capacity because of incomplete cataclasis or a and temperatures can be focussed into the fault zone
different deformation mechanism associated with the and the drainage pattern or migration path followed
lower strain in this zone. These processes will result together with the exact PTx history of the fluid in
in a complex distribution of the displacement pres- the fault zone will influence the distribution of (1)
sures in the fault zone which will affect the sealing deformation, (2) the corrosion potential, and (3) the
capacity of the fault. The exact pattern of seal types precipitated cement.
and capacities will depend upon the distribution of It is also possible that the generation of cements
lithologies, the fault displacement pattern and the adjacent to faults can influence the fluid flow along
deformation mechanism histories experienced by dif- the fault. Strengthening of the border zones by pre-
ferent parts of the fault. Another important use of cipitation will reduce the potential of lateral leakage
the maps is the identification of low displacement of migrating fluids from the fault zone. If the ce-
pressure locations which will aid assessment of fluid mented zone extends for a distance from the fault
and hydrocarbon flow in and out of the fault zone. zone, which is larger than the width of the strain
accommodation zone associated with each increment
(b) The propagation and displacement mecha- of slip, then this process will result in the rapid
nisms of the faults. As discussed earlier in the pa- migration of fluid to the tip zone of a propagating
per the propagation and displacement mechanisms fault. This would aid the fault propagation at the
often involve the cyclic accumulation of strain. Dif- tip zone and allow migrating fluids to by-pass sealed
ferent parts of the propagation and displacement horizons. The size of the accommodation strain zone
cycles may be associated with different deforma- developed at the fault border will depend upon the
tion mechanism histories. For example, fault tips are exact faulting processes, the rock properties and the
338 RJ. Knipe

Fig. 16. Development of seal characteristics associated with fault ion. Cements associated with each propagation event result
in a border zone seal. See text for details.

fault zone geometry. The thickness of the cemented varying porosity/permeabilities encountered in the
zone will depend on the infiltration distance of the same fault (see Fig. 12).
fluid into the rock, the amount of solute present in
the fluid and the pressure history of the fluid during (d) Fault plane geometry. The exact geometry
infiltration. In situations where the rock permeability (e.g., curvature) of a fault plane can control the
is high and/or the pressure drop in the fluid associ- development of fault rock, the characteristics of the
ated with infiltration is slow then the thickness of a strain accommodation zone adjacent to the fault and
cemented zone generated during each faulting event can, therefore, influence the type of seal produced.
may be large. On the other hand thick cement seals Deviations from a planar feature can induce localised
may also develop sequentially, where rupture of an dilation or compression during deformations which
earlier border cement seal allows access to the un- range in size from grain or fragment sized asperities
cemented rock. This latter situation will be favoured or perturbations to large scale bends in the fault
by changes in the strain accommodation mechanisms orientation which occur on the scale of kilometres.
or patterns (see the section on "Fault plane geome- Dilation sites can act as areas of rapid fluid pres-
try" below). Small scale distributed fracturing in the sure drops during faulting which will lead to precipi-
tip zone (associated either with aftershocks or with tation (Sibson, 1989; Sibson et al., 1988) or may act
the decreasing energy available for fracturing as the as locations where fluidised gouge generated during
propagation rate of the fault dies out) may also result displacement is concentrated. Fibrous shear veins
in a concentration of cement growth at the tip zone. are a familiar small scale example of precipitation
These mechanisms provide ways of generating and along faults. These commonly have a patchy distri-
extending a border zone cement seal sequentially bution on exposed fault planes and are separated
during each increment of fault propagation and ex- by rock contacts where the modification of the rock
tending the fluid flow distance associated with each fabric may be low. Such faults are not effective seals,
faulting event (Fig. 16). The continuity of the seal as leaking may occur between the fibrous patches,
along the fault zone will depend upon the relative unless the displacement or dilation is large enough
extent of the cement zone generated and the propa- to allow development of fibres over the entire sur-
gation distance associated with each fault increment. face and/or contact areas are marked by fine-grained
Another important factor in this context will be that gouge. Sibson et al. (1988) have presented evidence
propagation of the fault into a permeable unit may that dilation can affect large areas associated with
arrest the fault propagation by causing a rapid fluid dilational jogs on fault planes. Although described
pressure drop and a focussing of fluid migration in the context of gold mineralisation, the process
along the unit. Sealing of the unit by cementation is also applicable to the present discussion of seals
would then aid continued propagation of the fault. and large cemented areas may develop in extensional
This last statement also emphasises the likelihood zones along faults.
of the border cement seals having different charac- An example of the affect of fault geometry on the
teristics and thicknesses in different lithologies with distribution of seal types is shown in Fig. 17. The
Faulting processes and fault seal 339

Seal strength and stability

The strength and stability of a fault seal after
its formation is crucial to the trapping potential of
^ ^ ^ \ ^ ^ ^ ^ - ^ ^ V ^ - F A U L T PLANE BEND. the fault. The susceptibility of the seal to failure
D. \ \ ^ ^ ^ ^ /\JtJ^'ll\j>^K Z eo of
n i n c r e adsfracturing.
e
will determine the probability of hydrocarbon leak-
age or non-trapping. Watts (1987) recognised the
importance of the strength of seals and proposed a
y^v. ^ ^ x
Fault plane contours classification of cap-rock seals based on their mode
DILATIONAL JOG.
of failure. Harding and Tliminas (1988) suggested
Zone of increased fracturing. that older faults are less likely to leak and noted the
Zone of enhanced cementation. importance of fault rejuvenation in promoting leak-
age. The decreased leakage associated with older
faults noted by Harding and Tminas may well relate
Fig. 17. Fault plane geometry and the distribution of seal types.
the sketch illustrates the importance of dilational jogs and fault to the reduction in the permeability and porosity of
bends on the distribution of deformation in the adjacent rocks. fault zones during the post-movement concentration
The shaded areas are zones in which enhanced fracturing and of diffusion mass transfer in fine grained fault rocks
cementation is possible.
as illustrated in example 2 described above.
The timing of seal evolution relative to the fault
activity and faulting processes on an individual fault
size or volume of the cemented zone developed at is important to the stability of the seal. The most
dilational jogs or fault bends is dependent on the stable seals will be those created late during faulting,
strain accommodation zone developed. The size of as these are less likely to be disturbed by subsequent
such accommodation zones will depend upon the deformation. Further strengthening of the seal would
stress levels, strain rates and strain magnitudes and be promoted if continued compaction and cemen-
the deformation mechanisms induced in the vicinity tation by diffusive mass transfer takes place in the
of the feature as well as on the size and geometrical fault zone after the main displacement on the fault.
characteristics of the bend or jog. The situation is The timing of fault seal development with respect
complicated by the likely superposition of repeated to the sequence of fault activity history on a larger
cycles of dilation and compression which can be (basin) scale will also affect the seal stability. The
induced in faults with complex geometries. Rocks most stable seals will be those which are unaffected
immediately adjacent to the fault zone can be af- by subsequent displacement of other members of the
fected by the large number of small perturbations in fault array. The fault array evolution is critical as the
the fault plane, while rocks at greater distances from distribution of fault activity at different times during
the fault plane may only be affected by the larger the basin or reservoir evolution will affect the seal
scale fault bends encountered during displacement. survival. Analysis of the deformation microstructures
The displacement rate and the curvature of a fault and the diagenetic history in and adjacent to dif-
bend or perturbation will affect the processes operat- ferent faults can help provide such information on
ing in the strain accommodation zone. Deviations in fault activity. In addition, such analyses can also aid
the fault plane of 5-10 in the movement direction the identification of those faults which have been
can induce important changes in the deformation reactivated during inversion.
mechanisms of the strain accommodation zone (see The different types of seal discussed in this paper
Knipe, 1985). Fracture arrays may evolve in the strain will have different strengths and different proba-
accommodation zone if the induced strain rates and bilities of being breached. Seals which border the
stress levels are high enough. These conditions will fault zones and arise from either cementation and/or
be favoured by perturbations or bends with a small pore throat plugging are likely to be more stable
radius of curvature or during rapid displacement features because they can be unaffected by later
events (see Knipe, 1985). Fracture arrays developed fault displacement events. Fault creep processes are
to negotiate parts of a fault zone may provide im- less likely to cause breaching of border zone seals
portant migration pathways for fluids and may also than faulting by transient, higher strain rate events.
induce the development of a wide sealed zone with The more extensive the fault border seals the more
a complex internal structure. In other words, areas likely they are to survive intact. However, if too
of the faults where bends or perturbations exist can extensive, they may then seriously restrict the de-
be considered as factory zones for fractures which sirable reservoir properties. As noted by Weber et
are then displaced along the fault during subsequent al. (1978) and Smith (1980), shaley smear seals will
movement. be more effective if they do not contain mixtures of
340 RJ. Knipe

sand. Cataclastic seals with extensive cements will be carbons GB are thanked for granting permission to
stronger and more resilient to later deformation than publish Fig. 3.
uncemented cataclastic fault zones.
As described above, the fault plane geometry, References
i.e., curvature changes along the displacement path
followed by the reservoir horizon, will affect the Allen, U.S., 1988. Model for hydrocarbon migration and entrap-
potential for seal breaching. Reservoir rocks which ment within faulted structures. Am. Assoc. Pet. Geol., Bull.,
73: 803-811.
develop early sealing properties at faults are more Angevine, C.L. and lurcotte, D.L., 1983. Porosity reduction by
likely to survive if changes in fault plane curvature pressure solution. A theoretical model for quartz arenites.
during subsequent displacement are small. Stability Geol. Soc. Am. Bull., 194: 1129-1134.
of the seal will be enhanced if adjacent faults in the Atkinson, B.K. (Editor), 1987. Fracture Mechanics of Rocks.
fault array, which are active after the fault under con- Academic Press, London.
Barber, DJ., 1985. Dislocations and microstructures. In: H.R.
sideration, do not induce reactivation or breaching of Wenk (Editor), Preferred Orientation in Deformed Metals
the previously sealed fault. For example, listric faults and Rocks An Introduction to Modern Texture Analysis.
are likely to induce more accommodation strains in Academic Press, London, pp. 149-182.
the adjacent rocks than movement on a set of pla- Berg, R.R., 1975. Capillary pressure in stratigraphic traps. Am.
nar domino faults. It is clear from the discussion Assoc. Pet. Geol., Bull., 59: 939-956.
Bj0rlykke, K., Aagaard, P., Dypvik, H., Hastings, D.S., Harper,
that fault seals can develop early during fault array
A.S., 1986. Diagenesis and reservoir properties of Jurassic
evolution and survive and it is equally clear that sandstones from the Haltenbanken area, offshore Norway.
assessment of such situations requires integration of In: A.M. Spencer (Editors), Habitat of Hydrocarbons on the
fault analysis on a range of scales. Norwegian Continental Shelf. Norwegian Petroleum Society,
Graham and Trotman, London, pp. 275-286.
Blenkinsop, T.G. and Rutter, E.H., 1986. Cataclastic deformation
Concluding statement of quartzite in the Moine Thrust Zone. J. Struct. Geol., 8:
669-682
Borradaile, G.J., 1981. Particulate flow and the generation of
The examples presented above indicate that the cleavage. Tectonophysics, 72: 306-321.
deformation mechanism histories in fault zones can Brace, W.F. and Byerlee, J.D., 1966. Stick-slip as a mechanisms
be both varied and complex and that microstruc- for earthquakes. Science, 153: 990-992
tural analysis can aid the assessment of the detailed Brock, W.G. and Engelder, J.T., 1977. Deformation associated
processes of sealing. with movement on the Muddy Mountain overthrust in the
Buffington window, SE Nevada. Geol. Soc. Am. Bull., 88:
The range of sealing processes which may affect 1667-1677.
particular faults is similar to the range of diage- Charlaix, E., Guyon, E. and Roux, S., 1987. Permeability of
netic changes which can affect reservoirs quality with a random array of fractures of widely varying apertures.
the added complication of the range of deformation Transport in Porous Media, 2: 31-43.
mechanisms which may operate. As with the diage- Cockerham, R.S. and Eaton, J.P., 1987. The earthquake cycle
and its aftershocks, April-September 1984. U.S. Geol. Surv.,
netic history of a reservoir, there is no simple method Bull., 1039: 15-28.
of predicting the exact sealing properties of a fault Cox, S.F. and Etheridge, M.A, 1989. Coupled grain scale
zone and each case will depend upon the timing, dilatancy and mass transfer during deformation at high fluid
conditions, duration and processes of faulting. pressure: examples from Mount Lyell, Tasmania. J. Struct.
Future studies of fractures and faults preserved Geol., 11: 47-62.
Downey, M.W., 1984. Evaluating seals for hydrocarbon accumu-
in core together with onshore studies are needed lations. Am. Assoc. Pet. Geol., Bull., 68: 1752-1763.
to build up a data-base on the fault behaviour of Etheridge, M.A., Wall, V.J., Cox, S.F. and Vernon, R.H., 1984.
different reservoir lithologies in different faulting sit- High fluid pressures during metamorphism and deformation:
uations. An integration of such fault zone deforma- implications for mass transport and deformation mechanisms.
tion and fault zone diagenesis studies with analysis J. Geophys. Res., 89: 4344-4358.
Frank, E C , 1965. On dilatancy in relation to seismic sources.
of reservoir diagenesis, fault geometries and fault
Rev. Geophys., 3: 485-503.
array evolution (timing, distribution, etc.) will aid the Frser, H J . , 1935. Experimental study of the porosity and
unravelling of the detailed role of faults in reservoir permeability of clasitic sediments. J. Geol., 43: 912-996.
development. Gabrielsen, R.H. and Koestler, A.G., 1986. Description and
structural implications of fractures in Jurassic sandstones of
the Troll Field, nothern North Sea. Nor. Geol. Tidsskr., 67:
Acknowledgements 371-381.
Gratier, J.P. and Guiguet, R., 1986. Experimental pressure
solution-deposition on quartz grains: the crucial effect of the
Discussions with G. Cowan, R. Frost, P. Gold-
nature of the fluid. J. Struct. Geol., 8: 845-856.
smith, G.E. Lloyd R.H. Sibson and I. Stewart on the Groshong, R.H., 1988. Low temperature deformation mecha-
processes of faulting are all acknowledged. Hydro- nisms and their interpretaion. Geol. Soc. Am. Bull., 100:
Faulting processes and fault seal 341

1329-1360. deep sea drilling project cores from forearcs. In: J.C. Moore
Gueguen, Y., Darot, M. and Reuschle, T., 1987. Permeability (Editor), Structural Fabric in Deep Sea Drilling Project Cores
evolution under stress. In: Doligez, B. (Editor), Migration of from Forearcs. Geol. Soc. Am. Mem., 166: 55-74.
Hydrocarbons in Sedimentary Basins. IFP Explor. Res. Conf., Nelson, R.A., 1985. Geologic Analysis of Naturally Fractured
45: 281-285. Reservoirs. Contributions in Petroleum Geology and Engi-
Hadizadeh, J. and Rutter, E.H., 1983. The low temperature neering, 1. Gulf Publishing Co., Houston, Texas, 320 pp.
brittle-ductile transition in a quartzite and the occurrence of Ohnaka, M., 1987. Nucleation and propagation processes of
cataclastic flow in nature. Geol. Rundsch., 72: 493-509. stick-slip failure and normal stress dependence of the physical
Harding, T.P., 1974. Petroleum traps associated with wrench parameterss of dynamic slip failure. Natural Disaster Sei., 9:
faults. Am. Assoc. Pet. Geol., Bull., 58: 290-304. 1-21.
Harding, T.P. and luminas, A.C., 1988. Interpretation of footwall Owen, G., 1987. deformation processes in unconsolidated sands.
(lowside) fault traps sealed by reverse faults and convergent In: M.E. Jones and R.M.F. Preston (Editors), Deformation of
wrench faults. Am. Assoc. Pet. Geol., Bull., 72: 738-757. Sediments and Sedimentary Rocks. Geol. Soc. London, Spec.
Harding, T.P. and lliminas, A.C., 1989. Structural interpretation Publ., 29: 11-24.
of hydrocarbon traps sealed by basement normal blocks and Pittman, E,D., 1981. Effect of fault-related granulation on
at stable flank of foredeep basins and at rift basins. Am. porosity and permeability of quartz sandstones, Simpson
Assoc. Pet. Geol., Bull., 73: 812-840. Group (Ordovician), Oklahoma. Am. Assoc. Pet. Geol., Bull.,
House, W.M. and Gray, D.R., 1982. Cataclasites along the 65: 2381-2387.
Saltville thrust, U.S.A., and their implications for thrust-sheet Rutter, E.H., 1983. Pressure solution in nature, theory and
emplacement. J. Struct. Geol., 4: 257-269. experiment. J. Geol. Soc. London, 40: 725-740
Houseknecht, D.W., 1984. Influence of grain size and tempera- Rutter, E.H., 1986. On the nomenclature of mode of failure
ture on intergranular pressure solution, quartz cementation, transitions in rocks. Tectonophysics, 122: 381-387.
and porosity in a quartzose sandstone. J. Sediment. Petrol., Rutter, E.H., Maddock, R.H., Hall, S.H. and White, S.H., 1986.
54: 348-361. Comparative microstructures of natural and experimentally
Houseknecht, D.W., 1988. Intergranular pressure solution in four produced clay-bearing fault gouges. Pure Appl. Geophys.,
quartzose sandstones. J. Sediment. Petrol., 58: 228-246. 124: 3-30.
Hubbert, M.K., 1953. Entrapment of petroleum under hydro- Rutter, E.H., Maddock, R.H., Hall, S.H. and White, S.H., 1986.
dynamic conditions. Am. Assoc. Pet. Geol., Bull., 37: 1954-
Comparative microstructures of natural and experimentally
2026.
produced clay-bearing fault gouges. Pure Appl. Geophys.,
Jones, M.E. and Preston, R.M.E, 1987. Deformation of Sed-
124: 3-30.
iments and Sedimentary Rocks. Geol. Soc. London, Spec.
Schowalter, T.T., 1979. Mechanisms of secondary hydrocarbon
Publ., 29.
migration and entrapment. Amer. Assoc. Pet. Geol., Bull., 63:
Knipe, R.J., 1985. Footwall geometry and the rheology of thrust
723-760.
sheets. J. Struct. Geol., 7: 1-10.
Sibson, R.H., 1977. Fault rocks and fault mechanisms. J. Geol.
Knipe, R.J., 1986a. Faulting mechanisms in slope sediments:
Soc. London, 133.
examples from Deep Sea Drilling Project cores. In: J.C.
Sibson, R.H., 1980. Transient discontinuities in ductile shear
Moore (Editor), Structural Fabric in Deep Sea Drilling
zones. J. Struct. Geol., 2: 165-171.
Project Cores from Forearcs. Geol. Soc. Am. Mem., 166:
Sibson, R.H., 1981. Fluid flow accompanyiong faulting: field
45-54.
evidence and models. In: D.W. Simpson and P.G. Richards.
Knipe, R.J., 1986b. Microstructural evolution of vein arrays
(Editors), Earthquake Prediction; An International Review.
preserved in Deep Sea Drilling Project Cores from the Japan
Am. Geophys. Union, M. Ewing Ser., 4: 593-603.
Trench, Leg 57. In: J.C. Moore (Editor), Structural Fabric in
Sibson, R.H., 1985. Stopping of earthquake ruputures at dilation
Deep Sea Drilling Project Cores from Forearcs. Geol. Soc.
fault jogs. Nature, 316: 248-251.
Am. Mem., 166: 75-88.
Knipe, R.J., 1986c. Deformation mechanism path diagrams for Sibson, R.H., 1986a. Brecciation processes in fault zones
sediments undergoing lithification. In: J.C. Moore (Editor), inferences from earthquake rupturing. Pure Appl. Geophys.,
Structural Fabric in Deep Sea Drilling Project Cores from 124: 159-175.
Forearcs. Geol. Soc. Am. Mem., 166: 151-160. Sibson, R.H., 1986b. Earthquakes and rock deformation in
Knipe, R.J., 1989. Deformation mechanisms recognition from crustal fault zones. Annu. Rev. Earth Planet. Sei., 14: 149-
natural tectonites. J. Struct. Geol., 11: 127-146. 175.
Kozak, J. (Editor) 1987. Physics of fracturing and siesmic energy Sibson, R.H., 1989. Earthquakes as a structural process. J.
release. Pure Appl. Geophys., 124: 973 pp. Struct. Geol., 11: 1-4
Kranz, R.L., 1983. Microcracks in rocks: a review. Tectono- Sibson, R.H., Robert, F. and Poulsen, K.H., 1988. High angle
physics, 101: 449-480. reverse faults, fluid pressure cycling and mesothermal gold-
Lloyd, G.E. and Knipe, R.J., 1992. Deformation mechanisms quartz deposits. Geology, 16: 551-555.
accommodating faulting of quartzite under upper crustal Sibson, R.H., 1990 Conditions of fault-valve behaviour. In:
conditions. J. Struct. Geol., 14: 127-143. Knipe, R.J. and Rutter, E.H. (Editors), Deformation Mech-
Logan, J.M., 1979. Britlle Phenomena. Rev. Reophys. Space anisms, Rheology and Tectonics. Geol. Soc. London, Spec.
Phys., 17. 1121-1132. Publ., 54: 15-28.
Mawer, C.K. and Williams, P.F., 1985. Crystalline rocks as Smith, D.A., 1966. Theoretical consideration of sealing and
possible palaeoseismicity indicators. Geology, 3: 100-102. non-sealinh faults. Am. Assoc. Pet. Geol., Bull., 50: 363-374.
Mitra, S., 1988. Effects of deformation mechanisms on reservoir Smith, D.A., 1980. Sealing and non-sealing faults in the
potential in Central Appalachian overthrust belt. Am. Assoc. Louisiana Gulf Coast Basin. Am. Assoc. Pet. Geol., Bull.,
Petrol. Geol., Bull., 72: 536-554. 64: 145-172.
Moore, J.C, Roeske, S., Lundberg, N., Schoonmaker, J., Cowan, Tullis, T.E. (Editor) 1986. Friction and Faulting. Pure Appl.
D., Gonzales, E. and Lucas, S., 1986. Scaley fabrics from the Geophys., 124: 608 pp.
342 RJ. Knipe

Underhill, J.R. and Woodcock, N.H., 1987. Faulting mechanisms trapping in Nigerian growth fault structures. Offshore Tech-
in high porosity sandstones; New Red Sandstones, Arran, nology Conf., Paper No. 3356. 2643-2652. (Reproduced in:
Scotland. In: M.E. Jones and R.M.E Preston (Editors), N.H. Foster and E.A. Beaumont (Editors), 1988. Traps and
Deformation of Sediments and Sedimentary Rocks. Geol. Seals. Treatise Pet. Geol. Reprint Ser., 6: 81-91.)
Soc. London, Spec. Publ., 29: 91-105. Wesson, R.L., 1988. Dynamics of fault creep. J. Geophys. Res.,
Wang, GH., 1986. Internal Structure of Fault Zones. Pure Appl. 93: 8929-8951.
Geophys., Spec. Issue, 124: 373. White, J.C. and White, S.H., 1983. Semi-brittle deformation
Wardlow, N.C. and Cassan, J.P., 1978. Estimation of recovery within the Alpine Fault Zone, New Zealand. J. Struct. Geol.,
efficiency by visual observation of pore systems in reservoir 5: 579-589.
rocks. Can. Pet. Geol., Bull., 26: 572-585. White, S.H., 1985. Defect structures in deformed minerals. In:
Wardlow, N.C. and Taylor, R.P., 1976. Merwy capillary pressure J.C. White (Editor), Applications of Electron Microscopy in
curves and the interpretation of pore structure and capillary the Earth Sciences. Mineral. Assoc. Can., Short Course, pp.
behaviour in reservoir rocks. Can. Soc. Pet. Geol., 24: 225- 121-150.
262. Wilke, S., Guyon, E. and de Marsily, G., 1985. Water penetration
Watterson, J., 1986. Fault dimensions, displacements and growth. through fractured rocks: test of a tridimensional percolation
Pure Appl. Geophys., 124: 365-373. description. Math. Geol., 17: 17-27.
Watts, N.L., 1987. Theoretical aspects of cap-rock and fault seals Wilson, J.C. and McBride, E.F, 1988. Compaction and porosity
for single and two phase hydrocarbon columns. Mar. Pet. evolution of Pliocene sandstones, Ventura Basin, California.
Geol., 4: 274-307. Am. Assoc. Pet. Geol., Bull., 72: 664-681.
Weber, K.J., Mandl, G., Pilaar, WE, Lehner, E, Precious, Yamashita, T. and Knopoff, L., 1987. Models of aftershock
R.G., 1978. The role of faults in hydrocarbon migration and occurrence. Geophys. J.R. Astron. Soc, 91: 13-26.

R J . KNIPE Department of Earth Sciences, The University of Leeds, Leeds, LS2 9JT, U.K.
 343

Controls on the development of fractured reservoirs in the

Monterey Formation of central California

R.G. Hickman and J.B. Dunham

The Miocene Monterey Formation of central California hosts important oil reserves in fractured cherts, dolomites and
siliceous shales. Commercial accumulations required favorable initial conditions of sedimentation followed by deep burial that
produced embrittling diagenetic changes prior to regional folding. Strata of the Monterey Formation are derived from three
main components: (1) fine-grained terrigenous sediment; (2) opaline diatom tests; and (3) calcareous tests of nannofossils. Strata
consisting of varying proportions of these components are interlayered on a centimeter-decimeter scale. The most favorable
reservoir strata accumulated in isolated structural lows, far from sources of terrigenous sediment. The anoxic character of these
basins prevented bioturbation that would have homogenized the sediments and eliminated the distinctive mechanical properties
of individual layers. Anoxic conditions also maximized preservation of organic material.
With increasing temperature due to burial, nearly pure opaline sediments were converted progressively to brittle opal-CT
(opal-CT is a mixture of cryptocrystalline cristobalite and tridymite) and quartz cherts, while opaline-rich sediments were
converted to semi-brittle siliceous shales. Clay-rich sediments remained ductile and incompetent. Diagenetic processes also
converted some pure calcareous sediments to dolomites.
Chert, dolomite and siliceous shale of the Monterey Formation are prone to fracturing not only because of their brittle
character, but also because of (1) strong ductility contrasts between them and interbedded ductile shales, (2) the thinly
interbedded nature of the various lithologies and (3) possibly high fluid pressures due to oil generation in interbedded organic
shales at the time of deformation. Fractures in the Monterey Formation include planar extension and shear fractures, breccias
developed along these fractures and along faults, and breccias associated with intense small-scale tectonic folding.
Extensive, interconnected subvertical extension fractures developed at an early stage of deformation. These fractures are
approximately perpendicular to fold axes. Together with normal faults with similar strikes, these fractures provided conduits for
hydrocarbons generated from interbedded organic-rich shales to migrate up the flanks of growing folds. As folding progressed,
less continuous extension fractures related to bending developed. These fractures strike parallel to the axes of the folds and
improved reservoir quality. The zones of highest porosity and permeability are developed in breccias produced by meter-scale
tight folding of cherts and dolomites that are interbedded with ductile shales. Because these breccias are stratigraphically
controlled, they can have substantial lateral extent and are important reservoirs.

Introduction surface exposures of the reservoir interval in close

proximity to major fields. Insights from this region
The Miocene Monterey Formation hosts impor-
into the controls of fracture development and ac-
tant oil reserves in the Santa Maria Basin of coastal
cumulation of hydrocarbons have applications to
central California, the offshore extension of the basin
exploration and development of fractured reservoirs
and the nearby Santa Barbara Channel (Fig. 1). Off-
elsewhere, particularly those occurring in strata of
shore exploration during the 1980's discovered re-
biogenic origin.
coverable reserves estimated at 0.34 billion metric
This study is based on work along the coastal out-
tonnes of oil and 60-80 billion cubic meters of natu-
crops between Point Conception and Point Peder-
ral gas in several closed anticlinal structures (Ogle et
nales (Fig. 1) and the study of associated subsurface
al., 1987). Recoverable reserves are mainly in chert
samples.
and dolomite reservoirs with very low intergranu-
lar porosities and permeabilities, and in siliceous
Tectonic setting
shales with moderate intergranular porosities, but
very low permeabilities. Commercial hydrocarbon Coastal California has had a complex Cenozoic
accumulations in all of these strata require extensive, tectonic history, mainly related to relative motions
interconnected fracture networks. of the American lithospheric plate with the Faral-
An understanding of the fractured reservoirs in lon and Pacific oceanic plates. Paleogene and early
these fields is aided by the occurrence of good Neogene basins of coastal California developed in

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 343-353. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
Structure of the Mid-Norway Heidrun Field and its regional implications 395

and its bearing upon extensional modes of basin formation. J. structure, Andaman Sea. In: A.W. Bally (Editor), Seismic
Geol. Soc. London, 145: 455-472. Expression of Structural Styles. Vol. 3, Am. Assoc. Pet. Geol.,
Barr, D., 1987. Lithospheric stretching, detached normal faulting Stud. Geol. Ser., 15, pp. 4.2-1-4.2-8.
and footwall uplift. In: M.P. Coward, J.F. Dewey and P.L. Harding, T.P., 1984. Graben hydrocarbon occurrences and struc-
Hancock, (Editors), Continental Extensional Tectonics. Geol. tural style. Am. Assoc. Pet. Geol., Bull., 3: 333-362.
Soc. London, Spec. Publ., 28: 75-94. Hay, J.T.C., 1978. Structural development in the Northern North
Brekke, H. and Riis, F., 1987. Tectonics and basin evolution Sea. J. Pet. Geol., 1: 65-77.
of the Norwegian shelf between 62N and 72N. Nor. Geol. lilies, J.H., 1981. Mechanism of graben formation. Tectono-
Tidsskr., 67: 295-322. physics, 73: 249-266.
Brewer, J.A. and Smythe, D.K., 1984. MOIST and the con- Jacobsen, V.W. and Van Veen, P., 1984. The Triassic offshore
tinuity of crustal reflector geometry along the Caledonian- Norway north of 62N. In: A.M. Spencer et al. (Editors),
Appalachian orogen. J. Geol. Soc. London, 141: 105-120. Petroleum Geology of the North European Margin. Norw.
Bukovics, C. and Ziegler, P.A., 1985. Tectonic development of Pet. Soc, Graham and Trotman, London, pp. 317-327.
the Mid-Norway continental margin. Mar. Pet. Geol., 2: 2-22. Johnson, R.L. and Dingwall, R.G., 1981. The Caledonides: their
Bukovics, C , Shaw, N.D., Cartier, E.G. and Ziegler, P.A., 1984. influence on the stratigraphy of the northwest European
Structure and development of the Mid-Norway continental continental shelf. In: L.W Illing and G.D. Hobson (Editors),
margin. In: A.M. Spencer et al. (Editors), Petroleum Geology Petroleum Geology of the Continental Shelf of North-West
of the North European Margin. Norw. Pet. Soc, Graham and Europe. Institute of Petroleum, London, pp. 85-97.
Trotman, London, pp. 407-423. Kirton, S.R. and Hitchen, K., 1987. Timing and style of crustal
Caselli, F., 1987. Oblique-slip tectonics Mid-Norway shelf. In: extension N of the Scottish mainland. In: M.P. Coward, J.F.
J. Brooks and K. Glennie (Editors), Petroleum Geology of Dewey and P.L. Hancock (Editors), Continental Extensional
North West Europe. Graham and Trotman, London, pp. Tectonics. Geol. Soc. London, Spec. Publ., 28: 501-510.
1049-1063. Krantz, R.W, 1988. Multiple fault sets and three-dimensional
Crowell, J.C., 1974. Origin of Late Cenozoic basins in Southern strain: theory and application. J. Struct. Geol., 10(3): 225-
California. In: W.R. Dickinson (Editor), Tectonics and Sed- 237.
imentation. Soc. Econ. Paleontol. Mineral., Spec. Publ., 22: Nelson, P.H.H. and Lamy, J.M. 1987. The M0re/West Shetlands
190-204. area: a review. In: J. Brooks and K. Glennie (Editors),
Dewey, J.F, 1969. Evolution of the Appalachian/Caledonian Petroleum Geology of North-West Europe. Graham and
Orogen. Nature, 222: 124-129. Trotman, London, pp. 775-784.
Dalland, A., Worsley, D. and Ofstad, K., 1988. A lithostrati- Price, I. and Rattey, R.P., 1984. Cretaceous tectonics off Mid-
graphical scheme for the Mesozoic and Cenozoic succession, Norway: Implications for the Rockall and Faeroe Shetland
offshore mid- and northern Norway. Norw. Pet. Dir. Bull., 65 troughs. J. Geol. Soc. London, 141: 985-992.
pp. Ramberg, LB. and Smithson, S.B., 1975. Gridded fault pattern
Duindam, P. and Van Hoorn, B., 1987. Structural evolution of in a late Cenozoic and Paleozoic continental rift. Geology, 3:
the west Shetland continental margin. In: J. Brooks and K. 201-205.
Glennie (Editors), Petroleum Geology of North West Europe. Ramberg, LB., Cook, FA. and Smithson, S.B., 1978. Structure
Graham and Trotman, London, pp. 765-773. of the Rio Grande rift in southern New Mexico and West
Eldholm, O., Sundvor, E., Myhre, A.M. and Faleide, J.I., 1984. Texas based on gravity interpretation. Geol. Soc. Am. Bull.,
Cenozoic evolution of the continental margin off Norway 89: 107-123.
and western Svalbard. In: A.M. Spencer et al. (Editors), Rawson, P.F and Riley, L.A., 1982. Latest Jurassic-early Cre-
Petroleum Geology of the North European Margin. Norw. taceous events and the late Cimmerian unconformity in the
Pet. Soc, Graham and Trotman, London, pp. 3-18. North Sea. Am. Assoc. Pet. Geol., Bull., 66: 2628-2648.
Gabrielsen, R.H. and Robinson, C , 1984. Tectonic inhomo- Reches, Z., 1978. Analysis of faulting in a three-dimensional
geneities of the Kristiansund-Bod0 Fault Complex, offshore strain field. Tectonophysics, 47: 109-129.
Mid-Norway. In: A.M. Spencer et al. (Editors), Petroleum Reches, Z., 1983. Faulting of rocks in three-dimensional strain
Geology of the North European Margin. Norw. Pet. Soc, fields II. Theoretical analysis. Tectonophysics, 95: 133-156.
Graham and Trotman, London, pp. 397-406. Surlyk, F , 1977. Mesozoic faulting in East Greenland. Geol.
Gabrielsen, R.H., Foerseth, R., Hamar, G. and R0nnevik, H., Mijnbouw, 56: 311-327.
1984. Nomenclature of the main structural features on the Surlyk, F., Piasecki, S., Rolle, F., Stemmerik, L., Thomsen, E.
Norwegian and Continental Shelf north of the 62nd parallel. and Wrang, P., 1984. The Permian base of east Greenland.
In: A M . Spencer et al. (Editors), Petroleum Geology of In: A.M. Spencer et al. (Editors), Petroleum Geology of
the North European Margin. Norw. Pet. Soc, Graham and the North European Margin. Norw. Pet. Soc, Graham and
Trotman, London, pp. 41-60 Trotman, London, pp. 303-315.
Halstead, P., 1975. Northern North Sea faulting. In: Juras- Ziegler, P.A., 1981. Evolution of sedimentary basins in North-
sic Northern North Sea Symposium, Article 10. Norsk West Europe. In: L.V. Illing and G.D. Hobson (Editors),
Petroleumsforening, Oslo. Petroleum Geology of the Continental Shelf of North-West
Harding, T.P., 1983. Divergent wrench fault and negative flower Europe. Institute of Petroleum, London, pp. 3-39.

WJ. SCHMIDT Conoco Norway Inc., Norway

Present address: Conoco, 600 N Dairy Ashfora\ Houston, TX 77079, U.S.A.
 397

Characterization of faulting and fracturing in Ekofisk Field

from seismic, core and log data

J. Dangerfield, I. Knight and H. Farrell

Early well test results and core descriptions indicated that the chalks of Ekofisk field are highly fractured and that the
fracturing controls reservoir permeability. The initial conventional seismic was obscured by the presence of overpressured gas
in the overlying shales and was considered to be of no use in interpreting faults. Improved seismic methods have enabled
significant faulting in the chalks to be mapped. The orientations of these faults correspond closely to the orientations of natural
fractures interpreted from oriented cores and formation microscanner borehole images. Regional faults and natural fractures
are oriented NNE/SSW in the northwestern quadrant of the field. Elsewhere faulting is less frequent and natural fractures
are predominantly oriented radially around the elliptical dome that forms the field. The direction of the resulting permeability
anisotropy is anticipated to be a function of the natural fracture orientation, which varies with location on the structure.

Introduction tion. Other methods showing faults and fractures and

Ekofisk is a mature oil and gas field, currently their orientations all fit with this interpretation. The
(August 1990) with 40 producing wells, 17 water in- Formation Microscanner logging technique showed
jectors and 5 gas injectors. It has been producing that strong fracturing was present over much of the
from chalks of the Tor and Ekofisk Formations since field with this orientation. Core data indicated the
1972 (Brewster et al., 1986), yet offers frequent sur- presence of both this NNE-SSW trend and a ra-
prises. The field is undergoing further development dial fracture trend. Anelastic strain data indicated
with a waterflood injection program. The need to that the present day orientation of the stress system
predict the waterflood behaviour has driven activity closely reflects the core data.
to define the fracture system. The results of a three-dimensional seismic pro-
gram shot in spring 1989, and of a three-dimensional
Seismic data over the crest of Ekofisk have been
borehole profile shot in summer 1989, plus further
poor because the presence of gas in the overlying
core and log data, will shortly be available to help
section distorts the view of the reservoir. The gas has
clarify the importance of the fracture sets as the
leaked from the reservoir over geological time and
waterflood is extended throughout the field.
is now concentrated in the Tertiary sediments above
the crest of the field. The gas reduces the velocity of
sound to less than that of water, resulting in seismic Seismic data
sections showing very large time depressions, multi-
pathing and apparent faults. The field appears to be The problems created by the gas are shown by
a collapsed anticline in seismic sections (Fig. 1). The ray tracing. Using a simple gas model and consider-
field had been producing for two and a half years ing shots from outside the gas, the problems of ray
before a well was drilled into the collapsed area. The bending, multi-pathing and apparent fault creation
reservoir came in about 800 ft higher than predicted, become obvious (Fig. 3). Placing a geophone down-
showing the presence of a very large, domal structure hole, and much closer to the reservoir, constrains the
(Fig. 2). By 1983, with 55 wells drilled on the field, acoustic wavefront to a local and largely undistorted
all the available data indicated that the structure was image (Fig. 4). Borehole profiling therefore provides
a simple, effectively unfaulted dome. significant advantages.
Borehole profiles and carefully positioned surface Surface seismic line 100 is a typical seismic line,
seismic acquired since 1984 have changed our in- running NE-SW over the crest of the field, show-
terpretation of the field; it is now believed to be ing time delays and loss of reflectors (Fig. 5). The
heavily faulted in a predominantly NNE-SSW direc- borehole profile line from the same place shows

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 397-407. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
398 J. Dangerfield, I. Knight and H. Farrell

SW NE
PC88-0b0bS

^-3.0
TOP EK0FISKS '"" - ^ ~

' ; jr^' jm^r ^0

' """w .,:"*--; ;-^"--7^-t_,.>.r--' '*."- ^:^,^-*^ ' i ^ . f v^aa^^iK: ^ _ >;, ^ ^ ^ , ' ^ ^ , ^ ^ < 1 ^ ^ ^ w v ~ > J ^ "'' ""~ ~"'

1Km.
-j

Fig. 1. Seismic line PC880606.

the major improvement possible with downhole geo- some large faults, particularly in the lower reservoir,
phones (Fig. 6). Although the borehole profile image the Tor Formation, but there were not enough lines
was also depressed in time by gas effects, it revealed a for a definitive interpretation (Fig. 8). The crestal
graben with 45 m of throw (Christie and Dangerfield, graben appeared to be the dominant structure, with
1984). The image location is marked by the line C-D a NNW-SSE orientation.
in Fig. 7. The fault on the west side of the graben was Borehole profiling is roughly twenty times as ex-
later encountered in the reservoir during drilling. It is pensive as surface seismic per kilometre of line;
believed to be part of a major crestal graben, such as another method of obtaining surface seismic data
commonly found above salt domes. A total of eight was therefore tested in 1987. Instead of shooting
wells were subsequently used to obtain the seventeen lines in a normal star pattern across the field, the
borehole profile lines shown in Fig. 7. The profiles seismic lines and the hydrophone cable were kept
showed major faults and rapid variation in character tangential to the edge of the gas cloud and just out-
between profiles. They confirmed that the field has side it (Fig. 9). This minimized sound travel through
Characterization of faulting and fracturing in Ekofisk Field from seismic, core and log data 399

Three-dimensional seismic methods shot to mini-

mize gas effects should permit the interpretation of
much smaller faults. A three-dimensional survey was
shot in the spring of 1989 using two separate shoot-
ing directions to obtain data as far as possible up
the flanks of the field (Fig. 13). To attempt to obtain
data from the gas-affected crest of the field a three-
dimensional walkaway borehole profile was shot in
the autumn of 1989 (Fig. 14) with 41 walkaway lines,
each recorded by an 8-geophone string downhole,
to produce 328 lines. The results are described in
AAPG Memoir 42 (Dangerfield, 1991).

Core data

Reservoir permeability is thought to be controlled

by fracturing because the permeabilities inferred
from reservoir performance and the analyses of well
tests are as high as 150 mD (Brown, 1987), which
is at least two orders of magnitude greater than
matrix permeabilities measured in cores (Thomas et
al, 1984). To confirm the influence of fracturing, as
well as to acquire data for reservoir management, a
core-based study of fracturing was undertaken.
Natural fracturing recognized in Ekofisk cores has
been classified according to four major fracture types
(Brewster et al., 1986). Tectonic, stylolite-associated,
and irregular fractures all enhance reservoir perme-
Fig. 2. Top Ekofisk depth map from 1979.
ability. Healed fractures may, at least locally, reduce
matrix permeability.
gas affected layers (Fig. 10). The lines imaged more Tectonic fractures form the bulk of the fracturing
of the crest than any previous data and were of excel- in the Ekofisk Formation. They are through-going,
lent quality. Faults in the reservoir were considered planar features, dipping from 65 to 80, that form
real, as opposed to due to gas, when the Paleocene well developed parallel sets. Where the fracture in-
reflector was unbroken (Fig. 11). The results made a tensity is high, sets of fractures conjugate to the dom-
considerable difference to our interpretation: unex- inant set are also seen (Fig. 15). Fracture spacing,
pected faults were imaged with up to 70 m of throw. measured perpendicular to the fracture surface, is
The data set was still too sparse for a definitive highly variable throughout the formations but spac-
interpretation but a strong NNE-SSW orientation ing rarely exceeds 100 cm at any point in the field. It
seemed likely (Fig. 12). is believed that the fractures form an interconnected
The new map provided immediate benefit to the network throughout most of the productive reservoir.
reservoir simulation model. The presence of major Stylolite-associated fractures (Nelson, 1981) occur
faults explained why pressures in the upper reser- primarily in the Tor Formation; they are only rarely
voir (Ekofisk Formation) and lower reservoir (Tor encountered in the Ekofisk Formation. Most large
Formation) are equal within certain restricted ar- stylolites have fractures associated with them. The
eas (Fig. 12). Elsewhere, the Tight Zone separating fractures appear to have been initiated at the sty-
the reservoirs keeps them in pressure isolation. The lolite seam and extend away from the stylolite into
reservoir simulation model had been unable to the the chalk matrix. The stylolite-associated fractures
match the observed pressures without "windows" at are typically parallel to the stylolite columns and
the locations of the faults. within 5-10 of vertical (Fig. 16). The fractures usu-
The new seismic lines challenged our view of the ally extend up to 10 cm from the stylolite column,
probable origin of the field as a salt dome. The large but may be as much as 20 cm in length. They usu-
faults appear to extend up from the basement where ally form sub-parallel, anastomosing networks; within
they may have originated as strike slip faults. each network the fractures are interconnected and

400 l Dangereld, I. Knight and H. Farrell

1 1 1 11 11 1 11 1 1 1111 1 11 1 1 111 1 1 1 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 1

- 11

I]

0 0.5 1 Km.
1. J

.
Fig. 3. Raypaths through the gas-affected area from a shot outside the gas.

Fig. 4. Raypaths for a borehole seismic survey.

Characterization of faulting and fracturing in Ekofisk Field from seismic, core and log data 401

SURFACE SEISMIC LINE

FSv-* ' n^t^- 28

-2.9

* *z?z&?'
-3.0

EKOFlSK-j

-3.1
TIGHT ZONE-!

3.2

IKm

Fig. 5. Surface seismic line 100.

SURFACE SEISMIC LINE

EKOFISK

TIGHT ZONE-

life ^^3*fe ^ * ^^fe & "*

IKm
WALKAWAY PROFILE

Fig. 6. Surface seismic line 100 with the walkaway borehole profile inserted.

form bedding parallel zones of enhanced permeabil- fracture types may be partially or completely sealed
ity. They do not form a pervasive network throughout by secondary minerals. Calcite is the most common
the formation unless they occur in combination with fracture filling mineral, although quartz, pyrite and
the tectonic fractures. clayfillingsare not uncommon.
Irregular fractures and healed fractures do not Tectonic fracture trends for the Ekofisk Formation
appear to enhance reservoir permeability. Any of the are shown on Fig. 17. In most of the wells (B19a,
 lf
402 / Dangetfield, I. Knight and H. Farrell

^ K-22

**
^
<6/
^

3 ^*^\ x M /***'*'.*^ TV\ \ A X

,, ,,, ,, V
1
1 sW \ V.V.'.V'' ' 'V ' W IS^r J/ I
P"^~^*^ >^ > V \*'*.\\\\'"' \\/( S^
^^yC^^^ \ U \**^. "*,*V,**\ \ y^
s *"->^ \ \ \**"^*** ^ K \
y' X^rV^ff - ."*'VA J T I V ^"

n^wi
s*
\ M ^ ^ ^ "
to

I
1 0.5KM
1 TERTIARY GAS \ \

1 TANGEN TIAL SEISMIC ACQUISITION V / \

0 | KM \\ ' \ 1
i i \\ \ |
Fig. 7. Borehole profiles.
Fig. 9. Tangentially oriented surface seismic.

Fig. 10. Advantages of shooting with the cable oriented tangen-

tial to the gas.

K13b, Cll, A15a and A6) a strong NNE-SSW frac-

ture trend is present. This NNE-SSW fracture trend
is parallel to the trend of the major faults that have
been mapped from seismic data in the northern area
Fig. 8. Top Tor 1988 interpretation.
of the field. It is assumed that the fractures with this
Characterization offaulting and fracturing in Ekofisk Field from seismic, core and log data 403

SEISMIC LINE PWEKO 87V

NW SE

tifflfflmmmm
PALEOCENE

EKOFISK

Fig. 11. Surface seismic line PWEK087V showing good data and clear faults beneath an unbroken Paleocene reflector.

WELLS WITH EQUAL PRESSURE

IN TOR AND EKOFISK RESERVOIRS

Fig. 13. Three-dimensional surface seismic grid.

Fig. 12. Top Tor seismic interpretation in 1989, also showing
wells with equal pressure in both upper and lower reservoir
formations. features differing only in scale, do not seem to
be genetically associated with the Ekofisk structure
orientation are genetically related to this faulting, i.e. itself; their location and orientation are not affected
they developed at the same time, and in response by structural position. In general the tectonic fracture
to the same stresses as the faults. Fractures and trends in the Tor Formation reflect those in the
faults with this trend, which are geometrically similar Ekofisk Formation.
404 J. Dangerfield, I. Knight and H. Farrell

Fig. 14. Three-dimensional borehole profile acquisition of 41 walkaway lines.

214 A~6 V V N V

Fig. 15. Tectonic fractures.

On the east flank in K4 and C4, the tectonic have been identified from seismic interpretation. In
fractures trend approximately east-west. This trend, contrast, in the northwest part of the field, which
which differs by a few degrees in each well, is is heavily faulted, the NNE-SSW trending regional
interpreted to be part of a radial fracture system fractures predominate, although radial fractures are
that is genetically related to the fold. These radial also present in some wells.
fractures are seen as a subordinate trend in the B19a Core observations indicate that where they occur
well and possibly weakly in the K13b well. Several in the same intervals of core, the average strike of
wells (B20b, A15a and b, and Cll) are in a structural stylolite-associated fractures parallels the strike of
position that would make radial fractures parallel the the tectonic fractures. The stylolite-associated frac-
NNE-SSW regional trend, therefore the fractures in tures are therefore the extension fractures and the
these wells could have either, or both, origins. tectonic fractures are the shear fractures associated
On the east flank of the field radial fractures ap- with the same stress system.
pear to be dominant. In this area, no major faults
Characterization offaulting and fracturing in Ekofisk Field from seismic, core and log data 405

Fig. 16. Stylolite-associated fractures.

a higher fracture intensity than the poorer quality,

pelagically deposited chalks, of which the Tight Zone
is the best example.
Attempts were made to use the extensive conven-
tional log data base by examining various combi-
nations of logs to see if they could identify partic-
ular chalk facies and corresponding fracture char-
acteristics. It was found to be impossible to define
a statistically reliable method of identifying chalk
lithofacies from suites of conventional wireline logs.
This essentially prevented the use of a potentially
huge data base for any form of quantitative fracture
modelling. Fracture identification and description in
uncored wells was restricted to recently developed
/
TECTONIC FRACTURES
EKOFISK FM.
logging tools such as the Formation Microscanner
Tool (FMS).
The FMS, developed in 1985, generates an electri-
cal resistivity image of the borehole. It is a four-pad
Fig. 17. Orientation of the tectonic fractures in the Ekofisk conductivity tool with two rows of eight electrodes on
Formation. each pad. Data from the older wells in Ekofisk was
obtained using an earlier version of this tool that had
resistivity electrodes on only two of the four pads. In
Formation microscanner data the final borehole image, variations in conductivity of
35-50% of the borehole wall are displayed. Lighter
One of the earliest observations made from cores tones represent low conductivity and darker tones
was that fracture type and distribution showed some high conductivity. Knowing the relative bearing of
association with chalk lithofacies. While tectonic the tool pads, it is possible to compute the direction
fractures are ubiquitous throughout the reservoir in- and magnitude of dip for any planar feature, such as
terval, the stylolite associated fractures were almost fractures or bedding, crossing the wellbore.
exclusively confined to the lower reservoir section, The FMS was first run in Ekofisk in 1986 and
the Tbr Formation. Fracture intensity also suggested has now been run in thirteen wells with hole devia-
a correlation with the stratigraphy; the reworked tions of up to 45. An excellent example of a frac-
chalks which form the best reservoir intervals have tured chalk section seen in an FMS log is shown in
406 J. Dangereld, I. Knight andH. Farrell

ORIENTATION
180 2?0 380 90 180

Fig. 19. Top Ekofisk depth map showing FMS fracture trends.

The FMS has provided a valuable insight into

fracture orientations, but the tool is selective in the
type of fractures it can identify. High-angle, open,
tectonic fractures are generally easily recognized as
a consequence of the relative geometries of the frac-
FORMATION MICROSCANNER LOG FROM ture plane to the wellbore. The short, sub-vertical
A FRACTURED CHALK INTERVAL stylolite associated fractures are practically uniden-
tifiable with the processing presently used on the
EKOFISK FIELD
FMS. Consequently, some wells interpreted to be
Fig. 18. An FMS log showing fractures.
poorly fractured from FMS data have proven to be
good producers (with high fracture intensity) on well
test analysis.
Fig. 18. This image shows the sinusoidal patterns of
Other applications for the FMS are also under
contrasting matrix and fracture conductivities which
investigation. The four-arm caliper on the tool pro-
correspond to high-angle, open, tectonic fractures.
vides useful data to assess borehole ellipticity, well-
Sealed fractures may be identified by the white (low-
bore break-out and hence present-day stress orien-
conductivity) sinusoids.
tation. This information may have application to the
Figure 19 shows the fracture orientations recorded
future design of well stimulation programs. Further
from FMS data for the Ekofisk Field. Fracture
tool applications should develop with further devel-
orientations are highly consistent with orientations
recorded from cores, and reflect the strong NNE- opment of image processing.
SSW fracture strike seen in both seismic and cores
over the northern half of the field. In the 2/4A-4A Anelastic strain data
well drilled on the southeast flank of the struc-
ture, fracture strike was seen to be northwest- For reservoir management it is important to know
southeast, agreeing with core and anelastic strain the relative apertures of fractures of different sets
data, and interpreted as part of the radial fracture and whether these apertures will change as the stress
pattern. state within the reservoir is altered by production
Characterization of faulting and fracturing in Ekofisk Field from seismic, core and log data 407

and associated compaction of the reservoir. In order venturers on Ekofisk Field for their permissions to
to examine the distribution of stress in the reservoir publish this paper. The co-venturers are: Petrofina
in situ stress orientations and magnitudes have been A.S., Norsk Agip A/S, Elf Aquitaine Norge A/S,
determined from measurements of anelastic strain Norsk Hydro A/S, Total Marine Norsk A/S, Statoil,
recovery of oriented core (Tbufel and Warpinski, and Phillips Petroleum Company. All opinions ex-
1984). pressed in this paper are those of the authors and
Data from seven wells indicate that the orientation do not necessarily reflect the opinion of Phillips
of the maximum horizontal stress is radial (Teufel, Petroleum Company or any of the co-venturers. In
1991). Thus the present day maximum horizontal situ stress measurements were performed and inter-
stress parallels the radial tectonic fractures. Fracture preted by L.W. Teufel of Sandia National Laborato-
width, or aperture, is one of the most important pa- ries through a USDOE/industry cooperative research
rameters controlling fracture permeability. Fracture program. Discussions with Larry Teufel, John Hens-
width is a function of fracture surface roughness and ley, Tom Moore, David Ellen, Rod Boade, Michel
the normal stress across the fracture. Fractures that D'Heur and Guy Fraselle contributed to develop-
are parallel to the maximum in situ stress will have ment of many of the ideas discussed in this paper.
larger apertures and should be more permeable than
those at a high angle to the maximum stress. This References
indicates that the permeability of individual fractures
in the NNE-SSW and the radial fracture systems Brewster, J., Dangerfield, J.A. and Farrell, H.E., 1986. The
geology and geophysics of the Ekofisk Field waterflood. Mar.
may be dissimilar.
Pet. Geol., 3: 139-170.
Brown, D.A., 1987. The flow of water and displacement of
Conclusion hydrocarbons in fractured chalk reservoirs. In: J.C. Golf and
B.P.J. Williams (Editors), Fluid Flow in Sedimentary Basins
Seismic imaging of Ekofisk Field, by tangentially and Aquifers. Geol. Soc. London, Spec. Publ., 34: 201-218.
oriented surface seismic and borehole profiling, in- Christie, P.A.F. and Dangerfield, J.A., 1987. Borehole seismic
dicates that there are many large-scale faults in the profiles in the Ekofisk Field. Geophysics 52: 1328-1345.
field and that they are predominantly oriented NNE- Dangerfield, J.A., 1991. In: A.R. Brown (Editor), Interpretation
of 3D Seismic Data. Am. Assoc. Pet. Geol., Mem., 42, 3rd
SSW ed.
Core and FMS Log studies indicate that small- Dangerfield, J.A. and Brown, D.A., 1987. The Ekofisk Field. In:
scale fractures with the same orientation are also J. Kleppe et al. (Editors), Proceedings of North Sea Oil and
present, but can be subdivided into different groups Gas Reservoirs Seminar, NTH, 1985. Graham and Trotman,
based upon their characteristics. The core-based London.
studies indicate that small-scale radial fractures are Nelson, R.A., 1981. Significance of fracture sets associated with
stylolite zones. Am. Assoc. Pet. Geol., Bull., 65: 2417-2425.
also present and do not appear to be related to
Teufel, L.W., 1991, Influence of lithology and geologic structure
large-scale faulting. The radial orientation of the in "in situ" stress: examples of stress heterogeneity in reser-
maximum horizontal in situ stress indicates that the voirs. In: L.W. Lake, H.B. Carroll and T.C. Wesson (Editors),
permeability of fractures may be a function of their Reservoir Characterization, II. Academic Press, New York,
orientation and location. The fracture network at N.Y., pp. 565-576.
Ekofisk is therefore extremely complex and difficult Teufel, L.W. and Warpinski, N.R., 1984. Determination of "in
to interpret when dealing with problems of reservoir situ" stress from anelastic strain recovery measurements
of oriented core: comparison to hydraulic fracture stress
management, such as production anomalies, water- measurements. Proc. 25th U.S. Nat. Rock Mech. Symp.,
flood performance, and sidetrack locations. Three- Evanston, 111., pp. 176-185.
dimensional seismic and a strong multidisciplinary Thomas, L.K., Dixon, T.N., Evans, C.E. and Vienot, M.E., 1984.
team approach are required in our future work. Ekofisk Waterflood Pilot. SPE Paper 13120.

Acknowledgements

The authors would like to thank the manage-

ment of Phillips Petroleum Company and their co-

J. DANGERFIELD Phillips Petroleum Company Norway, Utenlandsk A ksjesebkap, PO. Box 220, N-4056 Tananger, Norway
I. KNIGHT Statoil A/S, P.O. Box 300, 4001 Stavanger, Norway
Present address: Phillips Petroleum Company Norway, Utenlandsk Aksjeselskap, P.O. Box 220, N-4056 Tananger,
Norway
H. FARRELL Phillips Petroleum Company Norway, Utenlandsk Aksjeselskap, P.O. Box 220, N-4056 Tananger, Norway
 409

Ula Field: relationship between structure and hydrocarbon

distribution?
A. Brown, A.W. Mitchell, I.R. Nssen, I.J. Stewart and P.T. Svela

The Ula Field is located on the Norwegian Offshore Continental Shelf, approximately 260 km SW of Stavanger. During the
course of field development, technical perception of the reservoir has evolved from a simple, partially filled anticlinal structure
with a single oil-water contact (OWC), to a complex accumulation with a difference in OWC between the west and the east
flanks of the field of more than 280 m.
Data from development drilling indicates that the oil columns on both sides of the field are in excellent pressure
communication, thus indicating no effective compartmentalisation. The real petroleum distribution in the Ula Field was not
discovered until 2 years after production start-up. As a result, the well data collected to date are insufficient to allow the precise
definition of the mechanisms which controlled the original fluid distribution in the field.
Three models are presented to explain the different OWC values. In one model, the crestal NW-SE striking fault system
which bisects the field acts as a permeability barrier for a compaction driven aquifer influx from the west. This results in differing
aquifer pressures causing displacement of contacts across a fault permeability membrane.
An alternative model is that the same crestal fault system acted as a seal through geological time and the eastern and western
reservoirs of the field were charged with hydrocarbons independently. The fault seal has subsequently failed as a result of the
large pressure differential that has developed through oil production which has only been from the east flank.
The third model, requires the reservoir to be discontinuous beyond the west flank of the field; that the crestal fault system
is not sealing; and that petroleum charged the structure from the east, progressively filling the field crest. This process would
continue only until the bottom sealing limits of the field crest were reached. At this point water contained within the west flank
would no longer be displaced leaving a trapped west flank aquifer.
Presently, there is insufficient data to identify which of these three models offers the most likely explanation for the
phenomena observed on the field to date.

Introduction "a Jurassic hydrocarbon province located along the

faulted north-eastern margin of the oil-rich Cen-
This paper reviews the impact of development well
tral Graben" (Spencer et al., 1986). The reservoir
drilling upon the structural and reservoir model of
of the Ula Field is the Upper Jurassic sandstone
the Ula Field. In the development planning stage,
of the Ula Formation. Mudstones of the overlying
Ula Field was considered a simple, four-way dip-
closed salt-cored anticline with a single field-wide Tyne Group formed the cap-rock (Deegan and Scull,
oil-water contact (Home, 1987). Two years after 1977) with petroleum sourced from the "hot shale"
production start-up, however, drilling of the first of the Upper Jurassic Mandal Formation (Spencer
downflank injection wells on the east flank of the et al., 1986). Details of the lithostratigraphy (Fig.
field demonstrated that the reservoir configuration 2) are described within that proposed by Deegan
was much more complicated than had at first been and Scull (1977) and modified by Vollset and Dore
thought. It is now evident that structuring has played (1984).
an important role in the migration and trapping of
hydrocarbons in the Ula structure although the exact Discovery
relationship is still poorly understood.
BP farmed in to the 019A Licence in May 1976
in return for drilling one well, located and operated
Regional setting
by Conoco. Using this well, 7/12-2 (Fig. 3), the Ula
The Ula Field lies approximately 260 km SW field was discovered. The well penetrated 154 m true
of Stavanger and 65 km NW of the Ekofisk com- vertical thickness (TVT) of Upper Jurassic sandstone
plex (Fig. 1). It is situated within the Ula TYend, and reached total depth (TD) in the TYiassic. The

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 409-420. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
410 A. Brown, A. W. Mitchell, I.R. Nilssen, I.J. Stewart and P.T. Svela

50km

Fig. 1. Ula trend. Fields and discoveries.

Upper Jurassic sand had a net thickness of 148 m drilled (Fig. 3). Wells 7/12-3 and 7/12-4 were to test
with porosities ranging between 14 and 28% and possible fault and dip-controlled spill points as well
permeabilities ranging from a few milliDarcies (mD) as the quality and thickness variation of the Upper
to 2.6 Darcies (D). On testing the Jurassic sands, Jurassic sand. Well 7/12-3 encountered an entirely
40.2 API oil flowed at rates up to 7100 stb/d and water-bearing reservoir and it was decided to deviate
the well proved an Oil-Down-To (ODT) of 3507 m the well up-dip to establish oil-water contact. The
TVD s.s. sidetrack, 7/12-3A, proved a Water-Up-To (WUT) of
3509 m s.s. Thus, the data from 7/12-2 and 7/12-
Appraisal, early reservoir delineation and 3A when combined with structure mapping indicated
development scheme a field-wide oil-water contact at 3508 m s.s. Well
7/12-4 was then targeted to confirm the position of
Following the success of 7/12-2, BP took over the OWC and to test the reservoir sandstone quality
the operatorship of the PL019A Licence. A suite of and petroleum distribution on the south-east flank
two-dimensional seismic data was acquired and in of the field. The well proved only 84 m of Upper
1977, appraisal wells 7/12-3, 7/12-3A and 7/12-4 were Jurassic reservoir and tested oil down to a depth of
Ula Field: relationship between structure and hydrocarbon distnbution? 411

FORMA SIMPLIFIED
GROUF

T TION LITHOLOGY SEISMIC MARKER VELOCITY LOG

CL
OWC:3508mss
D
o
CC

Q
Z
< I
TERTIARY
<
Q
! !

CC
o
X TOP BALDER
a. BALDER
L:
SELL- '" 7/

LI ST A
MAUREEN ~~~r i r TOP EKOFISK A

i^ic
I I I

3
I I I
CL
D

3
o
CC

<I TOR TM.

CO


O "L I"J
HOD
U
< J^Ti-TL
BASE CHALK

2
CL
U
z a.
* D
CC O
CC
> c
0
BASE CRETACEOUS

1
r.lANDAI TOP FARSUND |
1 AR'aJND
G
TOP ULA C.I. = 100mss
a
CO Fig. 3. Top Ula reservoir. 1982 two-dimensional map.
CO
<
CL
^ o Ul A f .1

CO g BASE ULA
> BRYNr
r ,^
Structure
LL /

iAS.VJM
TOP TRIASSIC
Both the stratigraphic and three-dimensional seis-
CJ
CO
CJ mic data now indicate that the Ula structure is the
~~
CO
< ^ CL
RAK
consequence of structural inversion of an originally
CL O SMITH
CL BANK
TOP "SALT"
complex, extensional system. The structural grain
Z
LJ CL ' , "- 1 of the area was established by Late Permian time,
CL
<^ O + + + + + -(- +
4 4 - 4 - 4 4 4-4-
and this grain has had a fundamental role, being
cl 4 4 - 4 4-4-4-4-
4 4 4-444-4- re-utilised in younger structuring.
4 + + + + + +
+- 4- 4- + 4- + +

Fig. 2. Ula Field. Generalised stratigraphic section. Permian extension

The main Permian structural element at Ula is a
north-south orientated westerly heading fault over
3505 m TVD s.s. supporting the interpreted "OWC" which there is significant thickening of the post-
at 3508 m TVD s.s. rift Late Permian evaporite sequence (Fig. 4). The
The remaining unproven reserves lay in the north fault is a single surface over the major part of
of the field and were appraised by the well 7/12-6. the three-dimensional grid, although it splays to the
The well encountered the reservoir shallower than north and the south. This surface is interpreted
predicted and consequently the sand was fully oil as transtensional; the fault mosaic revealed by the
bearing. three-dimensional data at Base Zechstein shows no
The field development scheme was thus based simple pattern that may relate to simple dip-slip
upon a four-way dip-closed reservoir partially filled extension and the other minor faults at this level are
with afield-wideOWC at a depth of 3508 m s.s. The interpreted as oblique conjugate sets between the
development plan comprised crestal production wells main fault and a further important north-south fault
supported by down-dip water injection. that lies to the east of the three-dimensional grid.
To date, a three-dimensional seismic survey has Superimposed on the Base Zechstein structural
been acquired and seven appraisal wells together grain is a salt wall and sediment pod geometry
with thirteen development wells have been drilled with associated faults which evolved in Early Triassic
on the field. These data have led to a complete times. The broad trend of the salt wall is north-west
revision of our understanding of the reservoir, and to south-east.
the elements of this are discussed below. The salt wall effectively became stable in Early TYi-
assic times leaving elongate grounded basins which
412 A. Brown, A. W. Mitchell, I.R. Nilssen, I.J. Stewart and P.T. Svela

MAIN FAULTS
AT BASE CRETACEOUS

BASE CRETACEOUS

BASE ZECHSTEIN

Fig. 4. The Jurassic extensional system.

were largely infilled by largely onlapping Tftassic cover east of the Base Zechstein Fault towards the
sediments. Accommodation faults developed along south along the shallow faults (Fig. 4). An anti-
the margins of the salt walls. In the Late Tfriassic, thetic fault to the main fault was established and was
regional extension utilised the Permian fault system, linked by transfer systems to the east and the west.
and the faults cut up-section through the salt wall. The overall style is that of a pull-apart above the
This structural style was further developed in Late deeper transtensional fault. The main ages of exten-
Jurassic extension. sion were Early Kimmeridgian to Mid-Portlandian,
with a subsequent phase in the Late Ryazanian, after
Jurassic extension the deposition of the organic rich claystones of the
Jurassic extension is thought to have been con- Mandal Formation. The crestal fault system over the
trolled by transtension on the NNW-SSE Base Zech- Ula structure is considered to have originated in the
stein Fault, in response to regional dip-slip extension Early Kimmeridgian in response to the oblique pull-
occurring on northwest-southeast orientated Base off, although the major throw on these faults is of
Zechstein fault sets, south and west of the Ula area. Late Ryazanian age.
The sense of motion at Ula was sinistral and the It is the displacement associated with this crestal
effect was to "pull-off" Late Permian and younger fault system which has subsequently controlled the
Ula Field: relationship between structure and hydrocarbon distribution? 413

distribution of hydrocarbons encountered in the field Well averaged, overburden corrected core porosi-
today. ties vary between 10% and 23% with an average
porosity of 16.5%. Permeability ranges between less
Cretaceous and Tertiary inversion than 1 mD to a maximum 3.1 Darcies in well
The present-day Ula reservoir trap is the result 7/12-A-18. The field average core arithmetic per-
of structural inversion which developed in response meability is 200 mD.
to dextral strike-slip or transpressional motion along The observed reservoir homogeneity, combined
the base Zechstein Fault. As with many of the inver- with the good reservoir quality is not considered to
sion structures in the Central North Sea, contraction be the principal control of the petroleum distribu-
occurred at several discrete times. The principal ages tion.
of inversion are Early Cretaceous, Late Cretaceous
(Middle to Late Campanian) and Tertiary (Late Petroleum distribution in the Upper Jurassic
Ypresian and Middle Oligocene). The geometries
of the inversion structure are directly opposite to Combined appraisal and development well drilling
the extensional geometries and the same lineaments have now demonstrated that Ula Field does not have
have been used in contraction as in extension. a single field-wide oil-water contact at 3508 m s.s.
The structural history as it can be interpreted In fact, the hydrocarbon column on the east flank
from the three-dimensional data set is, therefore, of the field extends at least 280 m deeper than the
much more complex than originally thought. The hydrocarbon levels penetrated on the west flank.
trap developed not through halokinesis, but through Appraisal well 7/12-7 encountered an "effective
cover extension and contraction in response to lat- OWC" at a depth of 3788 m s.s. (Fig. 7). This is
eral movement along the deeper Base Zechstein the deepest productive oil encountered in the field,
Fault surfaces into which the cover faults linked. to date, and residual oil was logged to the TD
Initially the sense is transtensional in the Mesozoic, of 3822 m s.s. It is considered possible that this
although the system is reversed and acted in periodic "effective OWC" may possibly represent an ODT or
transpression after Early Valanginian times. producing OWC and that the field may originally
have been full to a structural spill point which is
currently mapped in the east at a depth of 3820 m s.s.
Reservoir characteristics of the
The oil limits on the west flank of the field were
Upper Jurassic
also revised after the drilling of development well
7/12-A-12 (Fig. 7). This well proved an effective
Extensive core coverage from both development
OWC at 3462 m s.s. Additionally, the reservoir was
and appraisal wells drilled after production start-up
found to be depleted to the same pressure as the
has permitted an accurate description of the Upper
crestal-producing wells indicating that, at the time of
Jurassic reservoir. It comprises marine sandstones
drilling, the crestal fault system was not sealing. No
(Bailey et al., 1981 and Home, 1987) which vary
in thickness from 84 m in well 7/12-4 to 166 m in production or injection wells are currently located on
well 7/12-3A (Fig. 5). Reservoir thickness is greatest the west flank of the field.
in the north-western area of the field, gradually
decreasing towards the flanks. The relationship between structure and
The sandstones are predominantly fine and petroleum distribution
medium grained and are relatively homogeneous.
Mudstones are absent from the producing reser- Since reservoir heterogeneity cannot explain such
voir section (Fig. 6). There is, however, variation in a fluid distribution, a complex relationship must
reservoir quality across the Ula field and this vari- exist between the structure of the Ula Field and
ation is controlled by an interplay between detrital the petroleum distribution. Consequently, any expla-
clay and quartz cement (Gluyas, 1990). Fine-grained nation of the proven oil distribution relies on an
argillaceous sandstones have a high proportion of accurate knowledge of the influence of faulting and
microporosity and relatively low quartz cement con- compartmentalisation that was not observed when
tent, but also low permeabilities. Clean sandstones the original field development plan was formulated.
on the other hand contain high proportions of quartz The current perception of the structural evolu-
cement but nonetheless moderately good porosities tion of the field, indicates that the trap as we see
and permeabilities are preserved. The proportion of it today did not begin to develop until the on-
quartz cement in the reservoir also increases signifi- set of structural inversion in Cretaceous times. The
cantly with depth. fault system which is currently recognised cutting the
414 A. Brown, A.W.Mitchell, I.R. Nihsen, I.J. Stewart and P.T. Svela

5708'

5706'

5704'
250

C.I. = 10metres
Fig. 5. Ula Field. Reservoir isopach. The map is based on combined well and seismic data.

reservoir has been inherited from earlier Jurassic ex- and downthrows to the southwest. The presently ob-
tension. The majority of faults are discontinuous, but served throw along the Ula Field section of the fault
there are three (Hidra, Crestal and Transfer) which is highly variable, but difficult to quantify accurately
are thought to control the hydrocarbon distribution because of a marked deterioration of seismic resolu-
proven by drilling (Fig. 8). These are described in tion in the fault zone. The Ula reservoir is estimated
turn below. to be 70-90 m thick on the hanging wall of the
fault. Reservoir presence on the footwall is uncon-
(A) The Hidra Fault Zone. This a major structural firmed, but is expected to be very thin or absent. The
feature in the Ula trend. This complex fault zone north-eastern limits of the Ula Field petroleum are
defines the North eastern limits of the Ula field controlled by this fault.
where it strikes in a northwest-southeast direction
Ula Field: relationship between structure and hydrocarbon distribution? 415

Fig. 6. Ula Field. Stratigraphy and reservoir zonation.

EOWC: 3462mss
3788mss

/R3
v \\ \ I ^ \ o \
\\\
\ \\
\\ \ 1 T

^^-7/12-7
\> \ \ V #
^>
CO \

\ \ \ \ V
Sw \ V ^ ^ x
\ V ) \ A \cN \ s
\ ^^ "W. \
>#
v

s^~'Ow
7/12-A-12 p V^ X / v
\ \
N

f \V.

C.I. = 100m
Fig. 7. Top Ula reservoir. 1988 depth map.

(B) The NW-SE trending Crestal Fault System. more, the south-eastern end of this fault has very
The throw across this system is variable. Along the little throw and heads in the opposite direction.
field crest, very little or no throw is seen. Further-
416 A. Brown, A. W. Mitchell, I.R. Nilssen, LJ. Stewart and RT. Svela

Fig. 8. Ula Field. Top reservoir depth structure.

(C) The NE-SW oriented Transfer Fault This is although none fully explains the observed data set.
located at the southern end of the Crestal Fault and The three models are discussed in turn below and in
has 30-50 metres of displacement. This is insufficient each the role of the Jurassic crestal fault system is of
to displace the reservoir sands yet, it separates water- paramount importance.
bearing Ula formation in the footwall from oil-
bearing oil in the hanging wall. This fault is regarded Sealing faults model
as sealing. This model assumes that the Central Fault Sys-
Pressure data obtained prior to production start- tem was, in this case, sealing during the course of
up are shown in Fig. 9. These data can be used reservoir filling. The reservoir would require charg-
to infer a common OWC on the field, but the ing separately from east and west, with the east flank
only real aquifer data available prior to production filled to structural spill point and the west flank only
start-up are from the west flank well 7/12-3A. It partially charged (Fig. 10).
was, therefore, not possible to determine the true The main problems associated with this model are
relationship between the east and the west flank that the Crestal Fault System, which has no throw
aquifers prior to the onset of field production. One over the central section, must have been completely
additional piece of information comes from well sealing during the period of reservoir charging and
7/12-A-12 completed on the west flank of the field. remain so up until the onset of production.
This well encountered depleted reservoir pressures Pressure data from the exploration and appraisal
similar to those on the east flank indicating that, wells do not show oil gradients which differ between
2 years after production start-up, the east and west the east and west flanks (Fig. 9). An oil-leg pres-
flanks of the field were in good communication. sure difference of approximately 200 psi would be
Three models are proposed using the available anticipated between east and west if the central fault
structural and well data. Each model is partially valid system had been sealing and the aquifers were in
Ula Field: relationship between structure and hydrocarbon distribution? 417

logical time, then the fault seal must have broken

-fr ULA OIL GRADIENT down after production start-up. However, there is no
if
O
ULA WATER GRADIENT
7/12-3
obvious change in the pressure history of the field to
% 7 12-3A indicate that the fault seal might have broken down.
7/12-4
7/12-6
Dynamic aquifer model
T 7/12-A-10(HP)
This model assumes that the difference in fluid
levels from west to east is a result of compaction
driven active aquifer influx from the west of the
field (Fig. 11). Here it is assumed that the Ula Field
)VJC n(3508)
OWC snfi} V ^ T
lies on an active pressure gradient where formation
waters are being driven through the Ula structure
as a result of compaction of sediments in the basin
to the west of the field. In this model, the Crestal
20PSI
Fault System could be either a partial permeability
barrier or completely non-sealing. In the former
case, this would allow fluids to be transferred from
the west to the east when a critical pressure had been
7250 reached. This hypothesis would also permit the field
PRESSURE (PSIG) to be hydrocarbon charged either from east or west.
Fig. 9. Ula composite RFT/FMT pressures. The fault and stratigraphic configuration of the Ula
Field is not consistent with a simple, reservoir-borne
contact, lb obtain similar oil gradients with such aquifer flow from the west to the east of the field,
greatly differing oil columns across a sealing fault since structural and stratigraphic mapping suggests
requires either that the oils have to be of a radically that the reservoir sand limits to the west of the field
different composition, which geochemical analysis are restricted.
proves that they are not, or that the aquifers are not Thus, aquifer drive through compaction of shales
in communication. lying in the basin to the west of the field is seen as the
With respect to the latter point, while an aquifer only alternative mechanism to support the dynamic
gradient has been obtained for the wells on the west aquifer hypothesis. However, published data suggest
flank of the field, no such gradient is available for that fluid flow rates from compaction of shales in the
pre-production wells drilled on the east flank. Hence, basin to the west of Ula are inadequate (Giles, 1987)
the original relationship between the east and west to produce the observed contacts, flowing at rates of
flank aquifers will remain unknown. less than 1 mm/year (Fig. 12).
However, this model fails to explain simply the
data obtained from development well 7/12-A-12. Confined water tank model
This well was drilled on the west flank of the In this model, the reservoir sand is regarded as
structure, 2 years after production commenced, and being continuous over the Ula Field, but the sand
proved that the west flank of the field was in direct is discontinuous beyond the west flank of the field
communication with the east. This indicates that if either through shaling out of the sand (Fig. 13) or
the Crestal Fault System were sealing through geo- through disconnection by down-dip faulting.

WEST FAULT SEAL WELL B EAST

Fig. 10. Fault seal breakdown.

418 A. Brown, A.W.Mitchell, I.R. Nilssen, I.J. Stewart and RT. Svela

WEST EAST

Fig. 11. Active aquifer.

CONVECTIONALLY DRIVEN FLOW

Convection in 100m
highly permeable sandstone

''
Hydrothermal Active
circulation through hydrothermal
ocean floor sediments system
ii i 1 I
I I I I I
I I I I I

COMPACTIONALLY DRIVEN
FLOW
^
Early compaction
of clays

AQUIFERS

Shallow unconfined
Aquifers
Producing Aquifers,
depths to 2000m tn
the Great Artesian Basin

-6
T -4

-3
r -2 2 3 4
-5
10 10 10 10 10 0.1 1. 10 10 10 10
Ground-water velocity
m/yr
Fig. 12. Rates of water flow in sedimentary basins.

It is assumed that the crestal fault system is not would continue until the bottom sealing limits of
sealing and that petroleum charged the structure the field crest were reached. At this point, water
from the east, progressively filling the field crest. contained within the west flank would no longer be
If the hypothesis were correct, then the process displaced, leaving a trapped aquifer to the west. In
Ula Field: relationship between structure and hydrocarbon distribution? 419

WEST EAST

Fig. 13. Confined "water tank".

this case, the pressure gradients of the oil column Because of the late discovery of the "real" fluid
would be similar and the aquifer pressures would distribution, the data set available for interpretation
lie on different gradients. Seismic data suggest that is insufficient to enable identification of the precise
the reservoir sand is not laterally extensive to the controlling mechanism. In fact, three models have
west, but provides insufficient resolution to indicate been proposed as possible explanations, although the
unequivocal sand pinchout. Thus, a true limit to the lack of confirmatory data precludes the identification
west of the reservoir is not seen. The Crestal Fault of the most likely.
System appears to be open both at the crest and The preferred model is that of sealing faults which
also at the down-dip. For this model to work, the have broken down as a result of the rapid develop-
crestal sand-against-sand fault contact is required to ment of a large pressure differential through field
be non-sealing and the down-dip sand-against-sand production. Only this model has the potential of ex-
contact to be sealing. This model also requires the plaining all the facts, but if correct, we must expect
south-eastern Transfer Fault to be sealing even with to encounter some new oil-water contacts in future
very limited throw. development drilling.

Conclusions Acknowledgements

The authors wish to thank the British Petroleum

Despite the three-dimensional survey and exten-
Company pic, Svenska Petroleum A/S., Statoil,
sive well data, the controls on the hydrocarbon dis-
Norske Conoco A/S and K/S A/S Pelican and Co.,
tribution in the Ula Field are not well understood.
for granting permission to publish this paper.
Accurate knowledge of the sedimentology of the ho-
mogeneous reservoir sands together with the factors
controlling their post-depositional reservoir quality References
indicate that the gross petroleum distribution is not
controlled by either primary or secondary reservoir Bailey, C.C., Price, I. and Spencer, A.M., 1981. The Ula Oil
facies. The three-dimensional dataset has revealed Field, Block 7/12, Norway. In: Norwegian Symposium on
Exploration, Norsk Petroleum-forening, Article 18, pp. 1-26.
that Ula Field has undergone a much more complex Deegan, C.E. and Scull, B.J., 1977. A Standard Lithostratigraphic
structural history than was previously thought and nomenclature for the Central and Northern North Sea.
it suggests that the structuring is primarily respon- Institute of Geological Sciences, London, Rep. No. 77/25.
sible for the observed fluid phenomena. Although a Giles, M.R., 1987. Mass transfer and problems of secondary
complex structure, Ula field does not appear to be porosity creation in deeply buried hydrocarbon reservoirs.
Mar. Pet. Geol., 4: 188-204.
excessively compartmentalised, yet there is clearly a Gluyas, J.G., Leonard, A.J., Oxtoby, N.H., 1990. Diagenesis and
major variation in the OWC from the western to the petroleum emplacement: the race for space Ula Trend,
eastern flanks. North Sea. In: International Association of Sedimentologists,
420 A. Brown, A. W. Mitchell, LR. Nibsen, I.J. Stewart and P.T. Svela

Utrecht; 13th Int. Congr., p. 193. Norway, In: A.M. Spencer et al. (Editors), The Habitat of
Home, P.C., 1987. Ula. In: A.M. Spencer et al. (Editors), Hydrocarbons in the Norwegian Continental Shelf. Graham
Geology of the Norwegian Oil and Gas Fields. Graham and and Trotman, London, pp. 111-127.
Trotman, London, pp. 143-151. Vollset, J. and Dora, A.G., 1984. A revised Triassic and Jurassic
Spencer, A.M., Home, P.C. and Wiik, V, 1986. Habitat of Lithostratigraphic Nomenclature for the Norwegian North
hydrocarbons in the Jurassic Ula Trend, Central Graben, Sea. Norw. Pet. Dir., Bull., 3.

A. BROWN BPNorway Limited U.A., P.O. Box 197, N-4033 Forus, Norway
A.W MITCHELL BP Norway Limited U.A., P.O. Box 197, N-4033 Forus, Norway
LR. NILSSEN BP Norway Limited U.A., P.O. Box 197, N-4033 Forus, Norway
I.J. STEWART BP Norway Limited U.A., P.O. Box 197, N-4033 Forus, Norway
Present address: BP Exploration Inc., P.O. Box 4587, Houston, TX 77210, U.S.A.
P.T. SVELA BP Norway Limited U.A., P.O. Box 197, N-4033 Forus, Norway
 421

A frontier basin analysis study based on limited data: the

South Mozambique Graben

J.E. Iliffe, I. Lerche and M. DeBuyl

The South Mozambique Graben is assessed for hydrocarbon potential by reconstructing its structural, burial and thermal
histories using a seismic line and an array of basin modelling techniques.
A depth converted cross-section is simultaneously restored palinspastically and backstripped at designated pseudowells for
each chronostratigraphic package.
The thermal history is predicted from extensional models which relate paleoheat flow to basement subsidence. Maturation
potential cross-sections are derived by applying a generation model to each of the pseudowells, assuming each layer to be a
potential source rock. The basin is then analyzed by considering the timing of hydrocarbon generation relative to the deposition
of potential source and seal rocks and the formation of trapping structures and migration pathways.
When a range of uncertainty is allowed for in the development of these models, the exploration geologist can make a better
judgement of the dynamic range of hydrocarbon potential of this frontier basin. Also, this type of study determines the type of
additional information required to reduce uncertainties significantly.
In this paper we illustrate how mathematical modelling can enhance the understanding of the hydrocarbon potential of a
basin given only limited data. Formerly, one-dimensional burial history models (e.g., Guidish et al., 1984; Cao, 1985) have
assumed that sediments must have been deposited, not deposited (hiatus) or eroded during each time phase; meanwhile, the
structural development has been largely ignored.
The primary goal is to demonstrate a basin analysis technique which unites the structural, depositional and thermal histories
of an area. Such an integrated approach may greatly contribute to an exploration geologist's understanding and assessment of
the hydrocarbon potential of a speculative basin.
The method consists of simultaneously removing depositional layers and reconstructing structural cross-sections in sequential
chronostratigraphic steps back in time.
Before the technique was applied to real field data from the South Mozambique Graben, the problems of backstripping wells
which penetrate faults were investigated by means of a series of tests on synthetic examples and this aspect is discussed first.

Backstripping wells which intersect faults: tests lithologic parameters from Sclater and Christie
on synthetic examples (1980).
After the first unit (A) was backstripped off each
The synthetic scenario designed for the initial well, the underlying layers were decompacted and a
experiments consisted of a simple fault block which new cross-section was drawn by correlating the units
had been rotated on planar growth faults in domino in each well (Fig. lb). This process was repeated
style (Fig. 1). The burial history of this system began for unit B (Fig. lc). A one-dimensional vertical back-
with the deposition of three pre-rift layers (i, 2 stripping restoration of a two-dimensional fault block
and 3y Fig. 1) each of which represents 5 million scenario has shortcomings, mainly in that the throw
years. These layers were followed by the deposition of the faults cannot be restored, since this would
of two syn-rift layers (A and B, Fig. 1), each one also require lateral movement of strata. Other anoma-
equivalent to 5 million years. These secondary layers lous features of one-dimensional restoration are that
were of a different lithology from the first three. the fault angles are steepened and originally planar
Pseudowells positioned across the structure were faults become curved. Also, faults are backstripped
backstripped using a one-dimensional isostatic burial to the surface and so appear to migrate towards the
history program similar to that of Guidish et al. downthrown side. Strata beneath the fault are also
(1984). The decompaction of each chronostrati- incorrectly restored.
graphic layer was accommodated by means of the This latter problem was solved by treating the fault
porosity/depth functions of Magara (1976) taking as if it were an unconformity at depth. Sediments are

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 421-437. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
422 I.E. Iliffe, I. Lerche and M. DeBuyl

(a) PRESENT
WELL NUMBER

(b) 5 (Ma)
~~~^ B /
\ 1

1 ^ 2 A r Shaley-sand
N 2 B= S a n d Fig. 2. (a) Input system for the detailed analysis. Note positions
1 z Shale of wells A and B. Layers 1-6 are pre-rift, A and B are syn-rift.
2= Sand
(b) Burial history of well A incorporating the "erosion" below
3 3- Shaley-sand
3 the fault solution to compensate for disturbances of bed seen in
^ \ /\y:V;i* "^:--. Fig. l c . (c) Burial history of well B, also using the subfaulting
r'~'^-'/''\ A -^I'cL^'-, * ^ ;^V Y "erosion" solution.
- _\ /-)
'~&^'-~>' ',^:- y\
5 (Ma) 5 (Ma)

(C) 10 (Ma)

2
\ / \
3
3 /L_
(c) 10 (Ma) 10 (Ma)

Fig. 1. (a) Basic structural cross-section showing the location

of pseudowells used in all of the experiments. Note that in
some experiments the fault angle and scale of the system vary.
(b) Situation after removing layer "A" corresponding to the
scenario 5 Ma. (c) After removal of both the syn-faulting layers Fig. 3. (a) Backstripped and reconstructed situation of Fig. 3a
A and B (10 Ma). Note layers 1, 2 and 3 are considerably at 5 Ma. (b) Simple geometric solution (structural, not back-
discordant. Note also the curvature of the fault trace and that stripped) of Fig. 3a at 5 Ma. (c) Backstripped and reconstructed
sediments beneath the fault are not properly restored. situation at 10 Ma. Note waviness of strata compared with
Fig. 3d. (d) Geometrical solution at 10 Ma.

essentially considered to be "eroded" from beneath

the fault, while deposition is occurs above the fault where the individual effects of lithologies, spatial
(Fig. 2). The values for depths to the strata are read scale and initial fault angles were investigated. Two
from the burial history plots. The cross-sections at extreme lithologic cases (homogeneous shale or ho-
time periods 5 and 10 Ma resulting from this proce- mogenous sandstone) along with two variations in
dure (Fig. 3) were contrasted against those derived the present day fault angle (40 and 60) and two
from a simple geometrical restoration. Comparison different system scales (1:50000 and 1:25000) were
of the throws of the faults at different stratal levels backstripped and reconstructed. Results showed that
for each solution (i.e., decompacted and undecom- shallow faults are more affected by compaction than
pacted faults) at the 5 Ma time step (Fig. 4) shows steeper faults, the angle of the fault has more varia-
that compaction reduces the throw of the faults and tion up the fault in shales than in sandstone, and the
this effect decreases with depth. variation up the fault is greater in the larger scale
A sensitivity study was undertaken for a sequence model (Fig. 5).
of experiments similar to the initial case (Fig. 1) Another pitfall of using one-dimensional models in
Afrontierbasin analysis study based on limited data: the South Mozambique Graben 423

extensional regimes is related to "contraction" of the

T H R O W OF F A U L T ( m e t e r s )
basin during backstripping. After each reconstruction
3 100 200 300 400 500 600 70
1 1 1 I step a well's burial history may be radically altered
(e.g., well 1, Fig. 6). This problem is alleviated by
keeping the position of the pseudowells fixed and
A-
"redrilling" each well after each layer of sediment
has been removed, and then rerunning the burial
I ^^^^^r**
CO
00
B-
\ 1 history. The composite burial history of the well in
UJ
> 1-
s 1
that one position is a better geologic representation
< 1 \ than a simple "one run" case. These constraint pro-
-1
2 2J /1 cedures may now be implemented on real geologic
o /
/
/ situations.
o /
o
_j 3 J /
X /
1
/ The South Mozambique Graben:
-J /
/ a case study
6 w Decompacted
1
1
Undecompacted One of the problems often encountered by the
6-1 IL
5 Ma PRE8ENT exploration geologist is the assessment of the hydro-
carbon potential of a speculative basin with limited
Fig. 4. Comparison of the variation in throw of the fault in data which is also often of poor quality. This study
the non-decompacted and the decompacted solution. Note that exemplifies such a situation and shows how basin
the throw is generally increased by decompaction although this modelling may improve evaluation based on the lim-
effect decreases with depth.
ited resources available.

SHALLOW (small scale) DEEP (large scale)

W

<
H
J
O

o
O

W

<
J
<

|=i:| SHALE RANGE IN ANGLE

OFFAULT
SANDSTONE
Fig. 5. Summary of results of experiments considering the effects of lithology, fault angle and depth on the backstripping of the input
system shown in Fig. 1.
424 J.E. Iliffe, I. Lerche and M. DeBuyl

STRUCTURAL HISTORY BURIAL HISTORY

i : J TIME (Mya) 1
\ i

SSI
J PRESENT

FAULT

\ L=== k
5 Mya 25 20 15 10 5

FAULT
FAULT

10 Mya 25 20 15 10
_ .. _H --_ - I ; - - . - : - . - : - : - : - : "- : - : - : - :
V ~~~~_

COMPOSITE STRUCTURAL/BURIAL HISTORY

20 15 10 5 0
^
ULT ^ . FAULT

Fig. 6. Variation in burial history at a location due to contraction during reconstruction. Note the changes in burial history between 5
Ma and 10 Ma for each well. Compare with the composite burial history at the bottom.

The available data for offshore South Mozambique tween the continental crust of Mozambique and the
consisted of a few seismic lines, two shallow wells and oceanic crust of the Mozambique channel which
the general literature. In this study the tools applied separates Madagascar from the African continent.
to these data are a one-dimensional Burial History Martin and Hartnady (1986) postulate a Mesozoic
Program (Guidish et al., 1984), a Tissot and Weite plate boundary parallel to the Lombobo Mountains
(1978) type generation model (Tissot and Espitalie, of the Kaapvaal Craton, suggesting that the South
1975), models of basin subsidence (Royden et al., Mozambique Graben basement is thinned continen-
1980; McKenzie, 1978) and some basic geological tal crust. The general consensus is that Madagascar
knowledge. has been transported southwards along a N-S trend-
ing transcurrent fault (Coffin and Rabinowitz, 1987)
Tectonic setting (Fig. 8). This movement has been dated on the basis
The seismic data revealed a "V" shaped rift basin of magnetic anomalies as being from 150 to 110 Ma
trending N-S to NNW-SSE (Fig. 7) bordered to the (Segoufin, 1978) or 160 to 125 Ma (Rabinowitz et al.,
east by a fault trending NNW-SSE, and to the west 1983). It is also possible that initial rifting began as
by a fault oriented more N-S. Series of tilted fault early as Karoo time (Permo-TViassic to Jurassic) or,
blocks dipping to the east on westward verging faults more probably, is associated with the major break-up
are visible within the rift. of Gondwana in Late TYiassic to Early Jurassic times
The tectonic setting of offshore Mozambique has (Mougenot et al., 1986).
not been presented coherently to date, mainly due The NNW-SSE to N-S fault trends observed
to lack of data. Indeed, there is some debate as to within the South Mozambique Graben originally may
the age of rifting and direction of stress fields (cf. have been part of this transcurrent fault system, since
Segoufin, 1978; Rabinowitz et al., 1983; Mougenot et flower structures are visible in deeper parts of E-W
al., 1986; Martin and Hartnady, 1986; Coffin and Ra- seismic profiles. The Graben has been described as a
binowitz, 1987; Lawver and Scotese, 1987), although pull-apart basin by Chatellier and Slevin (1988). The
Darracott (1974) recognized that the area lies be- transcurrent faults may have been reactivated later
A frontier basin analysis study based on limited data: the South Mozambique Graben 425

(a)
DEGREES LONGITUDE
i \ i f
LU
Q
Z)
h-
H
<
_l

00
LU
LU
CL

(k) DEGREES LONGITUDE

4T
33 34 35 fL 36

LU
Q
3

CO
LU
LU South
CC
O Mozambique
LU Graben
Q

100km

Fig. 7. Location and structural trends which cut Lower Cretaceous strata in the South Mozambique Graben.

in a purely extensional regional stress field in Late The oldest unit presumed present is a basement of
Cretaceous times (100 Ma). volcanic rocks which make up part of the Kaapvaal
Craton (DeBuyl and Flores, 1986). This unit is prob-
Seismic interpretation and depositional ably overlain by the Karoo (whose existence offshore
history has not been proven beyond doubt). These rocks
Interpretation of the seismic data was based on the consist of a sequence of continental deposits ranging
recognition of seismic units rather than following one in age from Permo-TViassic to Upper Jurassic. At the
or two reflectors regionally, which proved impossible base, the Ecca series exhibit coal measures, shale,
in the rift itself. Basic isopach and structural maps siltstone and sands. The overlying Beaufort series
were produced and a seismic section perpendicular contains sandstones and shales and, at the top the
to tectonic strike was then selected for subsequent Stormberg series, is represented by volcanoclastics,
analysis (-', Fig. 7). lavaflowsand sandstones (Kamen-Kaye, 1978; 1982).
While describing the general problem of sedimen- The Karoo unit may be present in the South
tation for the south Mozambique Graben it is useful Mozambique Graben, making up part of unit 6
to refer to the seismic character of the six chronos- (Fig. 9a). The reflectors in this unit are structurally
tratigraphic units recognized (Lafourcade, 1984). disturbed and stratigraphically discontinuous typi-
426 J.E. Iliffe, I. Lerche and M. DeBuyl

m jmmm AJ-.^Jjx:M
an Upper Cretaceous unconformity overlain by the
clayey silts, sands, sandy limestones and marls of the
RIO ZAMBZl\ \-;1;fcT.yMOZAMBIQUE Upper Domo or Grudja Formation. This unit (3)
R16 '5V-;> has parallel, sub-horizontal and continuous reflectors
which may indicate a basin type depositional setting
(Flores, 1973). It is apparent from the seismic section
(Fig. 9a) that this unit was either very thin, eroded or
not deposited on the flanks of the rift.
Above the Grudja Formation lie an oolitic al-
^20 gal limestone, marls and sands of the Eocene-age
platform, and a ramp sequence of the Cheringoma
Formation (unit 2).
The ramp, visible on the northeast side of the
seismic section (Fig. 9a) contains slump structures.
The reflectors are slightly inclined, discontinuous and
non-parallel on the ramp and faint or absent in the
basinal areas.
Overlying these sediments are the deltaic sedi-
ments of Miocene to Recent age (unit 1). Reflectors
are gently dipping, thin and sub-parallel.

Backstripping and reconstruction

The seismic profile (Fig. 9a) was converted to a
true depth section (Fig. 9b) using pseudowells 1 to
DEGREES LONGITUDE 12 which were located at shot points where stacked,
Fig. 8. Main structural trends in the Mozambique Region. Note interval velocities had been calculated. These veloc-
the transcurrent fault in the area. ities were then used to calculate the thickness of
each layer. The accuracy of these thickness estimates
depends on the accuracy of the velocities, which was
lying them as a syn-rift deposit. The age of this not provided.
unit is speculative, since it has never been drilled. After obtaining a "true" section, (Fig. 9a) the
Reports from nearby countries with better data sug- lengths of each pick were compared and the actual
gest that the Karoo was deposited during Gondwana amount of extension as a ratio of the total extension
break-up, in Late TYiassic to Early Jurassic times was calculated for each time line, to produce an
(Mougenot et al., 1986). Accordingly this Karoo unit extension rate history (Fig. 10a).
is taken to be 195 Ma at its base and 140 Ma at its The structural interpretation was constrained by
top. This Jurassic syn-rift deposition is not reported length balancing an imaginary reflector below and
onshore, probably because most of the Gondwana parallel to the deepest seismic reflector at the base
break-up in the region now lies offshore, and the of unit 6 (Dahlstrom, 1969).
onshore areas of this time were probably erosional The data needed for the one-dimensional isostatic
highs manifested in the rock record as unconformi- burial history calculations are: depths to formation
ties, as suggested by Forster (1975). tops, absolute ages at these tops, layer lithologies,
Following deposition of unit 6, a marine transgres- paleobathymetry of each layer, density of rock matrix
sion occurred, beginning in Early Cretaceous times and porosity/depth relationships for various litholo-
(Albian?) as recorded by the sands and shales of the gies (Guidish et al., 1984).
Maputo Formation. Reflectors in unit 5 are discon- Several assumptions are therefore required to de-
tinuous and structurally disturbed. velop the analysis. The lithology of unit 6 is presumed
The Maputo Formation is overlain by the Lower to be sands related to Jurassic syn-rift deposition.
Domo Shales Formation from Cenomanian to TUro- Owing to the lack of downhole sonic and density
nian in age. There is a bright acoustic doublet visible information, the lithological porosity/depth parame-
at the top and bottom of the corresponding unit (4). ters were defaulted to those available in Sclater and
Internally the unit is generally seismically transpar- Christie (1980). Layers of thickness 1000 meters or
ent. Where visible, reflectors are discontinuous. more were taken to be monolithological. This is ob-
The doublet at the top of unit 4 may be the viously in error; in general, however, such a working
mid-Turonian Domo sands which are truncated by hypothesis provides an evaluation in keeping with
A frontier basin analysis study based on limited data: the South Mozambique Graben 427

Scale
^2 Km

PRESENT DAY LENGTH OF TIME LINES

1 40

15

Fig. 9. (a) Interpreted seismic section A-A' trending east-west, (b) Depth converted geologic cross-section with a generalized
stratigraphic column.

the resolution of the bulk of the data (Cao, 1988). paper (Fig. lie). The removal of throw on these
More critical to the model is the estimation of abso- faults slightly contracted the section such that those
lute ages. The paleobathymetry estimation is based wells that crossed a fault had to be "redrilled".
mainly on the geologic interpretation and sparse This process of backstripping, restoring, rebalancing,
literature. redrawing and redrilling was repeated until the top
All twelve pseudowells were backstripped one unit, of the Karoo reflector appeared as a horizontal line
after which the cross-section was redrawn (Fig. 11a). at the top of the model.
Since no faulting was evident after 55 Ma, no rebal- Each cross-section (Fig. lla-e) was then analyzed
ancing was necessary after unit 1 was removed. for depth to detachment and length of the uppermost
When the second unit was backstripped (76 Ma to bed using the method described by Gibbs (1983).
55 Ma), the throw on some faults had to be adjusted The final lengths used in the depth to detachment
according to the procedure described initially in this calculations were between the border faults in the
428 J.E. Iliffe, I. Lerche and M. DeBuyl

(a)
A G E (Ma) faulting, sympathetic to the southern emplacement
100
of Madagascar. The occurrence of flower structures
in the deeper parts of the rift, the resemblance the
"V" shaped graben has to a shear zone (Tchalenko,
1970) and wrench fault settings (Harding, 1974),
and the geologic data presented by Martin and Hart-
nady (1986) are suggestive of early strike-slip faulting
manifestation of the north-south separation of the
African and Antarctic plates (Simpson et al., 1979;
Martin and Hartnady, 1986). Strain was therefore
out of the plane of the East-West cross-sections (Fig.
lla-e). It is unclear what the Stromberg volcanics
onshore might represent tectonically. The vigorous
extension of the rift in Late Cretaceous times re-
sulted in the blocks being tilted in a more systematic
EXTENSION RATE normal domino fashion than in the previous event.
This event broadly coincides with the docking of
Madagascar in its present position relative to Africa,
to the onset of rifting between Madagascar and the
Antarctic (Segoufin and Patriat, 1980), to the ini-
AGE (Ma) tial separation of Antarctica and Australia (Veevers,
(b)
1986) and to postulated changes in poles of rotation
of South America relative to Africa (Rabinowitz and
LaBreque, 1979; Martin et al., 1982). It is, thus, ap-
2 parent that a major plate reorganization occurred.
I 10-
o Mougenot et al. (1986), working independently to
the north, also concluded that the region underwent
an initial period of strike-slip faulting followed by a
Late Cretaceous phase of extension and volcanism.
Such a two-phase history was first proposed by Flores
CHANGE OF DEPTH TO DETACHMENT
(1970). It may be significant that Upper Cretaceous
volcanics also occur at this time (DeBuyl and Flores,
Fig. 10. (a) The extension rate shown for each time period of 1986).
the backstripping. (b) Variation of the depth to detachment for
each time period of the backstripping.
The depth to detachment in this area may rep-
resent the brittle/ductile transition zone. If so, then
measured fluctuations in the depth to detachment
present day true section, i.e., within the zone of may be an indication of the heat flow history of the
deformation. region. Continuing this reasoning, the region may
have experienced its highest heat flow during late
Results and discussion of structural and Cretaceous times. This deduction is somewhat sub-
depositional reconstructions stantiated by the occurrence of Cretaceous sills and
feeder dikes in the area (DeBuyl and Flores, 1986).
When considering the variation of extension rate The fastest extension rate measured in the basin,
with time it was found that there was a slow acceler- some 5 cm per 1000 years, is slow when compared
ation of extension from 195 Ma to 76 Ma, the acme to spreading rates of between 1 and 17 cm per
of rifting being between 96 and 76 Ma (Fig. 10a). It year for present day opening oceans. This basin,
was also noted that the depth to detachment seems therefore, does not appear to have been very active
to have shallowed from 195 Ma up to the period and an extensional model to describe the heat flow
between 96 and 76 Ma and dropped gently since then and subsidence history must be applied with some
(Fig. 10b). It is significant that, even though the data caution.
are sparse and of poor quality, events of high ac-
cumulation, rapid subsidence and quickest extension
Thermal history
are almost coincident at 96 to 76 Ma.
The slow extension rates from 195 to 96 Ma may Having assessed the structural and burial histories
be attributed to early, subtle, probably strike-slip of the graben from the seismic line, some approx-
A frontier basin analysis study based on limited data: the South Mozambique Graben 429

(a)
55 (Ma)
12 11 10
~" ~"'~
^
<^\

\/
\ 11 / /
\ // 7 1
1
/

(d)

140 (Ma)
12 n

(e)
1 9 5 (Ma)
12 11 10

Fig. 11. (a) Cross-section just prior to Eocene deposition 55 Ma. (b) Cross-section at 76 Ma. (c) Cross-section at 96 Ma.
(d) Cross-section at 140 Ma. (e) Cross-section at 195 Ma.

imation to the thermal history of the basin must the information to obtain estimates of paleoheat flow
be made in order to apply hydrocarbon generation (Lerche et al., 1984). In lieu of any available thermal
models in the area to complete the basin analysis. information the nearest heat flow values of 1.2
In regions with data that include downhole measure- and 1.29 HFU come from the Mozambique channel
ments of thermal indicators, it is possible to invert (Anderson et al., 1977) the heat flow history of
430 J.E. Hiffe, I. Lerche and M. DeBuyl

the South Mozambique Graben is estimated from

the extensional model of McKenzie (1978) and the
dike intrusion model of Royden et al. (1980).
The geologic observations of several dikes or sills
in the region implies volcanic activity, which favors
the use of the Royden et al. (1980) dike intrusion
model. The McKenzie model does not involve any
intrusive effects and is, therefore, probably less ap-
propriate. However, in order to bracket the dynamic
range of possible thermal histories, both the Royden
et al. (1980) and McKenzie (1978) models were used.
Another constraint to consider when refining an ex-
tension model is the heat flow history derived from
depth to detachment calculations. Where the depth
to detachment is known, and with the detachment
assumed to be the brittle/ductile mid-crustal phase
boundary of greenschist to amphibolite metamorphic
facies, a crude estimate of heat flow through time
can be given using the equation: 109 96 76 65 55
TIME (Ma)

Fig. 12. (a) Observed basement subsidence curve from well 6

showing two subsidence events, (b) Determining the range of
extension factors from inspection of best fit between observed
where Q is heat flow at surface, K is thermal conduc- curves and those predicted by the extensional models.
tivity and Gtz is geothermal gradient. The variation
of paleoheat flow will essentially match the shape of this study and reported independently by Mougenot
the depth to detachment, e.g., the shallowest depth et al. (1986).
to detachment of 7 km yields the highest heat flow In practice, the fitting of the observed to predicted
value of between 2.8 and 3.6 heat flux units (HFU), basement subsidence curves has a degree of subjec-
using a thermal conductivity of 5 mcal/cm/C/s, and tivity. Therefore, the fit was bracketed to provide
a temperature at the depth which the phase change some idea of minimum and maximum error (Fig.
occurs of between 400 and 500C. This implies a 12b). For each well, and for each of the two models,
present day heatflowvalue of around 1.6 to 2.1 HFU there was an upper and lower limit of the possi-
which is higher than that observed in the Mozam- ble factors. Each of these upper and lower factors
bique Channel (Anderson et al., 1977). was then used in the paleoheat flow equation of
The McKenzie (1978) and Royden et al. (1980) the model under consideration. The observed pat-
models essentially predict the heat flow at a partic- tern of basement subsidence shows two events (Fig.
ular time after rifting in a basin, given certain pa- 12a), but only the second has any direct supporting
rameters of that basin. Both models relate basement field evidence. In order to accommodate all possi-
subsidence (tectonic subsidence) to the amount of ble variations, three general heat flow histories were
extension (McKenzie, 1978) or dike intrusion (Roy- considered (Fig. 13): (1) a single thermal heat pulse
den et al, 1980) and to the heat flow. It is, therefore, event at the first rifting (195 Ma); (2) a constant
possible to take the basement subsidence curves for heat flow of 1 HFU until 96 Ma at which time there
each of the pseudowells, which are output from was instantaneous rifting and an associated thermal
the burial history program (Fig. 12a), and find the event; and (3) an initial thermal event at 195 Ma
best fit predicted basement subsidence curve of the followed by a later event at 96 Ma (instantaneous
model that corresponds to a specific factor (Fig. 12b). rifting).
Inspection of the basement subsidence curves war-
The Tissot and Weite (1978) type generation
ranted the division of the structural history into an
model was applied to each well for all three pos-
initial rifting stage (195-109 Ma) and a later event
sible cases and a range of extension factors for each
at 96 Ma (Fig. 12). These two events correspond to
of the events. The input of a heat flow curve Q(t) at
known tectonic episodes of a strike-slip, pull-apart
time t was in the form of the equation:
basin setting, coincident with the southward em-
placement of Madagascar in the early Jurassic, and Q(t)=Q0cxp(at)
extension and volcanism in the Late Cretaceous, re-
spectively, as described from seismic and well data in where a is a variable which determines the slope and
A frontier basin analysis study based on limited data: the South Mozambique Graben 431

are within the oil window. Those layers may then

be studied more closely for their actual source po-
tential. The generation model predicts how many
CASE 1
milligrams of oil are produced per gram kerogen for
each layer at each well site. This value was then plot-
ted in the center of the formation for each well and
contoured (Figs. 14-17) to give maturation potential
cross-sections.
Such maturation potential sections were generated
for each of the three cases of heat flow and for
the upper and lower limits of each extension factor
for both the McKenzie and Royden et al. models of
extension. In this way, possible variations are taken
into account and some degree of confidence may be
placed on the results.
If there was only one rifting event at 195 Ma,
3
the Royden et al. (1980) model predicts that the
Q
basin is extremely oil prone (Fig. 14c, d) (if a source
2

(HFU)
CASE 3 rock exists!). On the other hand, if there was only
H
one thermal event of minimal magnitude at 96 Ma,
the McKenzie (1978) model suggests that most of
the basin is undermature, although still with some
96 potential in the east, near wells 4 and 5 (Fig. 15a, b).
Fig. 13. Three case histories: (a) initial rifting and thermal The results of the two rifting events case implies
cooling at 195 Ma; (b) a single pulse of rifting and heating at 96 that the later smaller heating event at 96 Ma has
Ma; (c) two rifting and thermal events at 195 Ma and at 96 Ma.
more effect on maturation of the sediment than
the rifting event of greater magnitude but occurring
curvature of the line, and Q0 is the present day heat earlier at 195 Ma (Fig. 16a-d). During the initial
flow. The extension factors taken from the models thermal event the sediments had either not yet been
were used to formulate an a value for the heat flow deposited or were still relatively near the surface
behavior with time in the model. with lower temperatures. During the later event the
For the first two cases this technique posed no same sediments were at a greater depth and so
problems, as long as the present day heat flow was experienced higher temperatures.
assumed known. A conservative estimate of 1.0 HFU The timing of generation shown for each case us-
was used for the present day heatflowwhen data was ing well 5 (Fig. 17) also suggests that the late thermal
lacking. event is more important for the hydrocarbon poten-
In the third case, of two thermal events, it was tial of this basin (Fig. 17b, c), since a later generation
necessary to calculate what the maximum heat flow facilitates sealing and trapping to a greater extent. It
would be at the inception of the second rifting, is, therefore, of importance that this event coincides
given the a value taken from the extension factors with a raising of the brittle/ductile boundary and the
predicted from the observed basement subsidence documented lava flows in the region.
curves for the first event. The maximum heat flow The key is not how high the heat flow occurred,
of the second event can be pinned by backtracking but rather how hot the sediments became and how
the equation from the present day to 96 Ma with a long they remained hot.
known a value as shown below:
Basin prognosis
R2 = oexp[+a(T p d -T R 2 )]
To evaluate a basin for hydrocarbon potential we
where QR2 is maximum heat flow at the second should answer five questions:
rifting event, Q0 is present day heat flow (1.0 HFU), (1) Does the basin have potential trapping struc-
r pd is time since initial rifting till present day and TR2 tures? In the case of the South Mozambique Graben
is time since initial rifting till second rifting event. the answer is strongly affirmative.
The Tissot and Weite (1978) hydrocarbon gener- (2) Does the basin have reservoirs? Most likely the
ation model is applied to all the strata. From the syn-rift unit, 6, may provide reservoir quality sands
results of this approach we can ascertain which layers and gravels in the form of fans at the sides of blocks.
432 IE. Iliffe, I. Lerche andM. DeBuyl

1 000
DEPTH
2000
(meters)
3000

milligrams oil/
gram kerogen

1 000
DEPTH
2000
(meters)

milligrams oil/
gram kerogen

Fig. 14. (a) Royden et al. (1980) model, lower limit, case 1. (b) Royden et al. model, upper limit, case 1. (c) McKenzie model lower
limit case 1. (d) McKenzie model, upper limit, case 1.

Kihle (1983) suggests that reservoirs may also be shales are potential seals. In most cases the point
present in the Maputo, Domo sands, Grudja sands in time at which generation from the lowermost po-
and Cherinyoma limestones. tential sources occurs is after the deposition of these
(3) Are sealing conditions present? The Lower shales (Fig. 17a, b), which would further enhance
Cretaceous and certainly the Upper Cretaceous potential.
A frontier basin analysis study based on limited data: the South Mozambique Graben 433

1 000
DEPTH
2000
(meters)
3000

milligrams oil/
gram kerogen

(c)
12 11 10 9 8
1 000
DEPTH
2000
(meters)
3000

25
mi,l
fi ? k m '9rams oil/
gram kerogen

Fig. 15. (a) Royden et al. model, lower limit, case 2. (b) Royden et al. model, upper limit, case 2. (c) McKenzie model, lower limit,
case 2. (d) McKenzie model, upper limit, case 2.

(4) Is there a source rock? This is a question Domo shales as being euxinic in nature, which would
which may only be answered definitively with a well. increase source potential, and Hartnady (1990) sug-
However, the fact that copious shales are present gests that the Limpopo drainage system is part of
(of both terrigenous and marine provenances) en- a large prograding delta, which also improves the
hances the possibility of a high quality organic source hydrocarbon potential of the region. Marine Jurassic
shale being present. Flores (1973) refers to the lower rocks have been reported further north and have
434 J.E. Iliffe, I. Lerche and M. DeBuyl

(a)
12 11 10 9
1 000
DEPTH
2000
(meters)
3000
25
milligrams oil
50 gram kerogen
2 km

1 000
DEPTH
2000
(meters)
3000

Fig. 16. (a) Royden et al. model, lower limit, case 3. (b) Royden et al. model, upper limit, case 3. (c) McKenzie model, lower limit,
case 3. (d) McKenzie model, upper limit, case 3.

been suggested by Kent et al. (1971) to be present in been produced from Lower Cretaceous or Jurassic
the offshore rifts. potential source rocks.
(5) Has the source rock adequately matured? The advantage of the technique is in the assimila-
Based on the limits of these analyses, which likely tion of all aspects of the basin to tie down the relative
encompass the real situation, hydrocarbons will have timing of structural, depositional and thermal events.
 l|
(a)
I

(b) (C)

CASE
IT
Q
I
Q CASE II
Q CASE III
HFU

HFU HFU

195 100
TIME ( M a ) r-
195 150 100 50 195 100 50
TIME ( M a ) TIME (Ma)
R
OIL GENERATION OIL GENERATION

(milligrams o i l / g r a m k e r o g e n ) (milligrams oil/gram kerogen)

1
OIL GENERATION
I ^'
milligrams oil/gram kerogen)

mg/g
mg/g
S
5

^,,/^
195 100 100 100
TIME ( M a ) TIME ( M a ) TIME (Ma)

Fig. 17. (a) Oil generation rate of the Royden et al. model, upper limit, case 1, well 5. (b) Oil generation rate of the Royden et al. model, upper limit, case 2, well 5. (c) Oil generation rate of the
Royden et al. model, upper limit, case 3, well 5.
436 J.E. Iliffe, I. Lerche and M. DeBuyl

It is important to determine whether maturation of history has recently been reported by Mougenot et
source occurred before or after deposition of a seal- al. (1986) from the area just to the north.
ing layer, or whether traps were formed and sealed (2) Extension rates estimated from variation in
before maturation. The South Mozambique Graben strata lengths show an acceleration of extension at
has potential source rocks most likely maturing after about 96 Ma, which also coincides with the point at
sealing rock (Upper Cretaceous shales) and traps are which the depth to detachment was calculated to be
in place (Fig. 17a, b). at its shallowest.
The biggest problem encountered with an oil- (3) Of the three different heat flow histories cal-
probable region, such as the South Mozambique culated from the subsidence models of Royden et
Graben, is potential recovery. The porosity at depths al. (1980) and McKenzie (1978), the Royden et al.
of between 3000 and 4000 meters is not too promis- model appears to be more favorable from the mat-
ing unless overpressured conditions are experienced. uration standpoint. Also, the later of the two rifting
Up dip stratigraphic traps may be more prolific. events modelled has more effect on source rock
The questions which cannot be answered are those maturation.
of hydrocarbon migration and accumulation, al-
though we can show that generation occurred mostly Hydrocarbon potential
around the time of deposition of the Upper Creta- (1) Structural and stratigraphic traps are present
ceous shale and generally after faults had become due to the extensive faulting.
inactive. Perhaps the migration potential could be as- (2) Reservoirs may be present in syn-rift sands or
sessed using a two-dimensionalfluidflow/compaction Cretaceous rocks.
model such as that of Nakayama (1987), but the lack (3) The Upper Cretaceous shales would suffice as
of more refined data renders such an endeavor less an effective seal.
valuable at the present stage of understanding of the (4) A source rock which can reach maturity for
South Mozambique Graben. hydrocarbons likely lies somewhere within the lower
Cretaceous or Jurassic strata. The probability of
Conclusions such a source rock being present is enhanced by
the marine/terrigenous shale depositional setting for
Theoretical testing these strata.
(1) When wells which intersect faults are one- (5) Regardless of the model used or heat flow
dimensionally backstripped, strata below the faults history configuration, the basin is expected to have
are improperly restored. This was solved by assuming mature hydrocarbons in the eastern part at a depth
that structural "erosion" takes place below the fault. of 3000-4000 meters.
(2) The changing burial history in a backstripped (6) The ideal situation of: relative timing of source
well, due to contraction of the cross-section during deposition followed by faulting, then seal deposition
reconstruction, was accommodated by "redrilling", followed by maturation of the source, seems to be
and recalculating the burial history for each time the case in the South Mozambique Graben.
step.
(3) Sensitivity analysis on backstripping planar
faulted extensional scenarios revealed that shallow Acknowledgements
faults are more affected by compaction than steeper
faults, the angle of the fault has more variation up The work reported here was supported by the In-
the fault in shales than in sandstones, and as the dustrial Associates of the Basin Analysis Group at
scale of the system gets larger so does the apparent USC. Data were generously provided by the West-
variation. ern Geophysical Co. of America. Many thanks to
Donna Black for typing and to Elain Hadaddin, Ann
Tectonic and thermal implications Watkins and Mike Garbee for drafting services. The
(1) The structural history of the South Mozam- paper benefited considerably from the comments of
bique Graben has two tectonic episodes recognizable the anonymous reviewers, to whom we extend our
on the seismic reflection profiles and subsidence thanks.
histories of the pseudowells. The first event is of
strike-slip nature, possibly sympathetic to the em-
References
placement of Madagascar. The second event is a
more normal extension manifested on the seismic Anderson, R.N., Langseth, M. and Sclater, J.G., 1977. The
section as a sequence of domino-type fault blocks mechanisms of heat transfer through the floor of the Indian
in the Cretaceous section. This type of structural Ocean. J. Geophys. Res., 82: 3391-3409.
A frontier basin analysis study based on limited data: the South Mozambique Graben 437

Cao, S., 1985. A quantitative dynamic model for basin analysis Lawver, L.A. and Scotese, C.R., 1987. A revised reconstruction
and its application to the Northern North Sea Basin. M.S. of Gondwana land. In: G.D. McKenzie (Editor), Gond-
Thesis, University of South Carolina, Columbia, S.C. wana Six: Structure, Tectonics and Geophysics. Am. Geophys.
Cao, S., 1988. Sensitivity analysis of 1-D dynamical model Union, Geophys. Monogr., 40: 17-83.
for basin analysis. Ph.D. Dissertation, University of South Lerche, I., Yarzab, R.F. and Kendall, C.G.St.C, 1984. Determi-
Carolina, Columnia, S.C. nation of paleoheatflux from vitrinite reflectance data. Am.
Chatellier, J.-Y. and Slevin, A., 1988. Review of African petro- Assoc. Pet. Geol., Bull., 68: 1704-1717.
leum and gas deposits. J. Afr. Earth Sei., 7: 561-578. Magara, K., 1976. Water expulsion from elastic sediments during
Coffin, M.E and Rabinowitz, P.D., 1987. Reconstruction of compaction-direction and volumes. Am. Assoc. Pet. Geol.,
Madagascar and Africa: evidence from the Davie Fracture Bull., 60: 543-553.
Zone and Western Somali Basin. J. Geophys. Res., 92: 9385- Martin, A.K., Goodland, S.W., Hartnady, C.J.H. and Du Plessis,
9406. A., 1982. Cretaceous palaeopositions of the Falkland Plateau
Dahlstrom, C.D.A., 1969. Balanced cross sections. Can. J. Earth relative to southern Africa using Mesozoic seafloor spreading
Sei., 6: 743-757. anomalies. Geophys. J. R. Astron. Soc, 71: 567-579.
Darracott, B., 1974. On the crustal structure and evolution of Martin, A K . and Hartnady, C.J.H., 1986. Plate tectonic devel-
southeastern Africa and the adjacent Indian Ocean. Earth opment of the South West Indian Ocean: a revised recon-
Planet. Sei. Lett., 24: 282-290. struction of East Antarctica and Africa. J. Geophys. Res., 91:
DeBuyl, M. and Flores, G., 1986. The Southern Mozambique 4767-4786.
Basin: the most promising hydrocarbon province offshore McKenzie, D., 1978. Some remarks on the development of
East Africa. In: M.T. Halbouty (Editor), Future Petroleum sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32.
Provinces of the World. Am. Assoc. Pet. Geol., Mem., 40: Mougenot, D., Virlogeux, P., Vanney, J.-R. and Malod, J., 1986.
399-425. La Marge continentale au Nord du Mozambique: resultats
Flores, G., 1970. Suggested origin of the Mozambique Channel. pr61iminaires de la campagne md40/macamo. Bull. Soc. G6ol.
Trans. Geol. Soc. S. Afr., 73: 1-16. Fr., 8: 419-422.
Flores, G., 1973. The Cretaceous and Tertiary sedimentary Nakayama, K., 1987. A dynamic, two dimensional, fluid flow
basins of Mozambique and Zululand: In: G. Giant (Editor), model for basin analysis. Ph.D. Dissertation, University of
Sedimentary Basins of the African Coasts, Part 2: South and South Carolina, Columbia, S.C.
East Coasts. Association of the African Geological Survey, Rabinowitz, P.D. and LaBreque, J.L., 1979. The Mesozoic South
Paris, pp. 81-111. Atlantic Ocean and evolution of its continental margins. J.
Forster, R., 1975. The geological history of the sedimentary basin Geophys. Res., 84: 5973-6002.
of southern Mozambique, and some aspects of the origin Rabinowitz, P.D., Coffin, M.E and Falvey, D., 1983. The separa-
of the Mozambique Channel. Palaeogeogr., Palaeoclimatol. tion of Madagascar and Africa. Science, 220: 67-69.
Palaeoecol., 17: 267-287. Royden, L., Sclater, J.G. and Von Herzen, R.P., 1980. Continen-
Gibbs, A.D., 1983. Balanced cross-sections construction from the tal Margin subsidence and heat flow: important parameters in
seismic sections in areas of extensional tectonics. J. Struct. formation of petroleum hydrocarbons. Am. Assoc. Pet. Geol.,
Geol., 5: 153-160. Bull., 64: 173-187.
Guidish, T.M., Kendall, C.G.St.C, Lerche, I., Toth, D J . and Sclater, J.G. and Christie, P.A.F., 1980. Continental stretching:
Yarzab, R.F. 1984. Basin evaluation using burial history an explanation of the post-mid-Cretaceous subsidence of the
calculations: an overview. Am. Assoc. Pet. Geol., Bull., 69: central North Sea basin. J. Geophys. Res., 85: 3711-3739.
92-105. Segoufin, J., 1978. Anomalies mosozoiques dans le bassin de
Harding, T.P., 1974. Petroleum traps associated with wrench Mozambique. C.R. Acad. Sei. Paris, 287: 109-112.
faults. Am. Assoc. Pet. Geol., Bull., 58: 1290-1304. Segoufin, J. and Patriat, P., 1980. Existence d'anomalies
Hartnady, C.J.H., 1990. Gondwana break-up, Limpopo drainage- Mesozoiques dans le bassin de Sanalie: implications pour
system evolution and the Kalahari Basin-Mozambique Delta les relations Afrique-Antarctique-Madagascar. C. R. Acad.
connection. J. Afr. Sei., 7. Sei. Paris, S6r. B, 291: 85-88.
Kamen-Kaye, M., 1978. Permian to Tertiary faunas and palaeo- Simpson, E.S.W, Sclater, J.G., Parsons, B., Norton, I. and
geography: Somalia, Kenya, Tanzania, Mozambique, Mada- Meinke, L., 1979. Mesozoic Magnetic Lineations in the
gascar, South Africa. J. Pet. Geol., 1: 79-101. Mozambique Basin. Earth Planet. Sei. Lett., 43: 260-264.
Kamen-Kaye, M., 1982. Mozambique-Madagascar Geosyncline, Tchalenko, S., 1970. Similarities between shear zones of different
I: deposition and architecture. J. Pet. Geol., 5: 3-30. magnitudes. Geol. Soc. Am. Bull. 81: 1625-1640.
Kent, P.E., Hunt, J.A. and Johnstone, D.W., 1971. The geology Tissot, B. and Espitalie, J., 1975. Levolution thermique de la
and geophysics of coastal Tanzania. Inst. Geol. Sei., London, matiere organique des sediments: applications d'une simula-
Geophys. Paper, 6: 101. tion mathematique. Rev. Inst. Fr. Petrole, 30: 743-777.
Kihle, R., 1983. Recent surveys outline new potential for offshore Tissot, B. and Weite, D.M., 1978. Petroleum Formation and
Mozambique. Oil Gas J., Feb. 28: 126-134. Occurrence. Springer-Verlag, New York, N.Y., pp. 500-521.
Lafourcade, P., 1984. Etude geologique et geophysique de la Veevers, J.J., 1986. Breakup of Australia and Antarctica esti-
Marge Continentale du sud Mozambique (17S 28S). Ph.D. mated as mid-Cretaceous (95 5 Ma) from magnetic and
Dissertation, Laboratoire de Geodynamique Sous-Marine, seismic data at the continental margin. Earth Planet. Sei.
Villefranche-sur-Mer, University Pierre et Marie Curie, Paris. Lett., 77: 91-99.

J.E. ILIFFE Department of Geological Sciences, University of South Carolina, Columbia, SC 29208, US.A.
I. LERCHE Department of Geological Sciences University of South Carolina, Columbia, SC 29208, U.S.A.
M. DeBUYL Western Geophysical, 10001 Richmond Avenue, Houston, TX 77252, U.S.A.
 439

The Danish North Sea: a basin prognosis

R.O. Thomsen, I. Lerche and J.A. Korstgrd

Applying one- and two-dimensional fluid-flow/compaction models to the NW part of the Danish Central Graben, the
geohistory, thermal history, source capacity and oil generation were examined. The burial history shows that the NW Danish
Central Graben developed through several stages of subsidence a Late Jurassic differential stage, a Late Cretaceous/Early
Tertiary uniform stage and a Late Miocene/Quaternary uniform stage. The thermal history is assessed by inversion of measured
vitrinite reflectance with depth and all wells give a consistent cooler past; resolution of the available data on the thermal history
is discussed. Pseudo-wells have been "drilled" on seismic sections in areas without well data. Synthetic vitrinite reflectance data
have been generated for the pseudo-wells using the modelled thermal history. The sensitivity of the maturity and thermal history
modelling to changes in parameters such as erosion events, errors in the present day heat flow and paleo-overpressure evolution
is discussed. The modelled maturation history of the Late Jurassic shale shows that the shale reached maturity some 5-50 Ma.
Modelling of the hydrocarbon generation indicates that the peak generation took place some time between 10 Ma and the
present day. Possible migration paths are determined from modelled excess fluid pressure and areas of possible accumulation of
hydrocarbons are indicated. The potentials of the areas are evaluated and an area along the eastern boundary fault between the
Tail End Graben and the Ringk0bing-Fyn High is suggested as a prospective new target for further exploration.

Introduction and also on hydrocarbon migration and on possible

accumulation sites.
The task of petroleum exploration is becoming
Both a one-dimensional and a two-dimensional
progressively more difficult in a time where the more
model have been applied to the Danish North Sea.
obvious structural reservoirs have been discovered.
The one-dimensional model was used as a first ap-
Exploration targets will progressively become either
proximation, with the two-dimensional model ap-
more subtle stratigraphic reservoirs in well explored
plied to an area which appeared promising through
basins or reservoirs in frontier basins in areas where
the one-dimensional study. For details about the
exploration is both costly and logistically difficult.
one-dimensional model the reader is referred to Cao
Within the range of uncertainties in the basic geo-
(1985), Cao et al. (1986) and Cao and Lerche (1987).
logic assumptions computer aided models provide us
For details of the two-dimensional model the reader
with the ability to perform an integrated quantitative
evaluation of the hydrocarbon potential of an area. is referred to Nakayama and Lerche (1987) and Wei
and Lerche (1988).
Although dynamical fluid-flow models for use in
quantitative basin analysis have been available for a
number of years (Weite and Yukler, 1981; Ungerer Geologic setting
et al., 1984; Cao, 1985; Cao et al., 1986; Cao and
Lerche, 1987; Nakayama and Lerche, 1987; Wei and The Central Graben developed as part of the
Lerche, 1988) these models are only slowly being North Sea system of rifts and is in itself a com-
integrated in the exploration process. plex multi-stage rift system comprised of a series
Many advantages lie, however, in the use of dy- of highs, grabens and half grabens. The division of
namicalfluid-flowmodels. In particular such models the area into structural elements has been adapted
provide a tool not only for predicting the hydrocar- from Vejbaek and Andersen (1987) (Fig. 1). Evolu-
bon potential of a basin, but also for testing different tion of an early fracture system is believed to have
hypotheses concerning the development of the basin originated in the Early Permian (Andersen et al.,
based on implications from the available data. In this 1982; Glennie, 1984). The area, however, lacks both
way a confidence range can be put on the assessed well and seismic control of the pre-Permian basi-
thermal history andfluid-flowconditions, on the pre- nal regime. Thus, it is not possible at present to
dicted hydrocarbon generation timing and locations, determine whether or not the system was active at

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 439-455. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
440 R.O. Thomsen, I. Lerche and JA. Korstgrd

Pseudo well
Well

PSW-2 \ \ 4>

* % psw 6

-JN5& ^
l \
'&c> orro-t <\ \
PSW-6 ^ J , ^
PSW-4 ^ \ . ^iJ <& psw-e
0
PSW.12 X o-t V, o
GRENSEN
NNOSE <^"
X^ PSW-14N

km
\
Fig. 1. The main structures and the commonly used subdivision into structural elements in the northern part of the Danish Central
Graben. Also shown is the distribution of wells (solid dots) and pseudo-wells (open circles) used in this study. The structural outline is
based on Vejbaek and Andersen (1987).

this time (Gowers and Saeb0e, 1985). It would seem seismic resolution, combined with erosion during
that the opening of the North Sea Grabens did not the Early Cimmerian and mid-Cimmerian tectonic
take place any sooner than late Permian or Early events, very little can be said about the original
THassic (Ziegler, 1982). The presence of Rotliegen- sediment distribution in large parts of the Central
des volcanics in the Danish wells B-l, P-l, Q-l and Graben (Andersen et al., 1982).
W-l, however, could indicate an early rifting stage At the end of the Early Jurassic the mid-
prior to deposition of the Zechstein evaporites. The Cimmerian tectonic phase induced renewed block
present day outline of the Zechstein sequence re- faulting in the Northwest European graben systems
veals deposition in a narrow graben trending NNW (Ziegler, 1982). Tectonic uplift in the central North
through the Tkil End Graben-S0gne Basin and in the Sea area caused erosion of Early Jurassic sediments
Feda Graben (Gowers and Saeb0e, 1985). The domi- and these are now found only in local basins (An-
nant lithologies in the Rotliegendes are volcanics and dersen et al., 1982). During the Middle and Late
redbeds, while the Zechstein transgression resulted Jurassic the structural framework was dominated by
in marginal shallow water carbonates and thick evap- N-S trending basement-attached faults and by NW-
orite sequences (Andersen et al., 1982). SE trending right lateral strike-slip faults (M0ller,
The beginning of the break-up of the Atlantic 200- 1986). This new trend left the S0gne Basin inactive,
180 Ma ago resulted in a new structural framework in while the major part of the Tkil End Graben subsided
northwestern Europe, modifying the Permian struc- as one large rotational fault block along its boundary
tural pattern in the North Sea, with the Viking and to the Ringk0bing-Fyn High. The Late Cimmerian
Central Graben as new major structural elements tectonic event caused a widespread diachronous un-
(Fisher, 1984). The TYiassic was marked by a regional conformity which in large parts of the area consti-
regression and return to a continental depositional tutes the boundary between the Late Jurassic and
regime. Due to scarcity of both well data and good Early Cretaceous sediments.
The Danish North Sea: a basin prognosis 441

The Early Cretaceous setting is a continuation of

that of the Jurassic and the depocenters appear to Sand

be related to those of the Late Jurassic (Gowers

and Saeb0e, 1985). TVansgressional sedimentation is
evident from on-lap sequences on seismic sections, Shale
although internal unconformities occur close to the
fault zone at the western margin of the Tkil End
Graben, suggesting tectonic activity in this area dur-
ing the Early Cretaceous (Andersen et al., 1982).
Tensional and left-lateral trans-tensional movements Shale/marl
occurred along the boundary faults of the Gertrud
and Arne-Elin Grabens giving rise to normal faulting
Chalk
(Vejbask and Andersen, 1987).
The Late Cretaceous ushered in a marked change
from predominantly siliciclastic sedimentation to de- Hiatus
position of chalk. Technically this change marks the
transition from rift-dominated subsidence to pas-
sive regional subsidence (Vejbaek and Andersen,
1987). Compressional and wrench-induced inversion Shale-silty
of previously tensional sub-basins dominated the
Late Cretaceous-Early Tertiary with the main in-
version occurring during Turonian-Santonian, mid-
Maastrichtian and post-Danian-Early Tertiary times
(Vejbaek and Andersen, 1987).
Regional subsidence continued and, during the
Tertiary and Quaternary, the basin subsided evenly Sand-silty

and rapidly. This subsidence pattern was disturbed

only by minor tectonic events and, locally, by haloki-
netic movements (Andersen et al, 1982). The Early
Tertiaiy sedimentation was mainly clay dominated,
while sand and silt intercalations increase towards
the Late Tertiary (Andersen et al., 1982). A general
stratigraphic column with formation names is shown
in Fig. 2.

Salt
The data base

A large amount of data is available in the open

domain from publicly released wells and the types of
data used in this study are summarized in Table 1.
Although there is an uneven spatial distribution of
wells with data, most of the structural elements are
Volcanics
represented by well data (Fig. 1). To improve the
data distribution and to include all structural ele-
ments, the seismic survey RTD-81 (GECO, 1981) has
been used for "drilling" pseudo-wells with synthetic
data. The distribution of wells and pseudo-wells used Fig. 2. Generalized stratigraphic column (after Michelsen, 1982
is shown in Fig. 1. In addition, line RTD-81-23 has and Michelsen et al., 1987).
been selected for a detailed two-dimensional simu-
lation (Fig. 3). Dating of the individual reflectors is
based on a correlation of biostratigraphy and seismic pre-Tertiary consistent with the works of Vejbask and
markers for selected wells in the northwestern part Andersen (1987) and M0ller (1986).
of the Danish Central Graben (Stouge, 1988; Stouge For simplicity, an average interval velocity for each
and Toxwenius, 1988). Furthermore, an effort has sequence was chosen for depth conversion of the
been made to make the seismic interpretation of the seismic sections. Reflector ages, lithologies derived
Arnt-Elln Graben

Fig. 3. Interpreted seismic section (RTD-81-23). SF = sea floor; TMM = Top Middle Miocene; TO = Top
Oligocene; TE = Top Eocene; TCH = Top Chalk; LCU = Late Cimmerian Unconformity; TLJ = Top Late
Jurassic; TEC = Top Early Cretaceous; ILJ = Internal Late Jurassic; BMT - Basement.
The Danish North Sea: a basin prognosis 443

TABLE 1

Well data used in this s>tudy (for location of wells see Fig. 1)

Stratigra- Wireline Composite Well data Tempera- Poros- Vitrinite Permea- Geo-
phical log log log summary ture ity reflect. bility chemistry
Lulu-1 _ X X X X _ X _ X
Mona-1 - - X - - - X - X
Sten-1 - X X - - - X - X
P-l X X - X X X - - -
Q-l X X - X X X X X X
Otto-1 - X X X - - - - -
T-l X X - X X X - X -
1-1 X X - X X X X X X
B-l X X - X X X X - X
Bo-1 - X - X X - X - X
H-l - - - - X X - - -
Adda-1 - - - - X X - - -
Adda-2 - - - - - - - - -
W-l X X X - X X X - X
E-l X X - - - - X - X

from well ties and the interval velocities used in the the proportions of different kerogen types assessed
depth conversion are shown in Table 2. The present by the geochemical study (R. Thomsen, 1989) for the
day temperature-depth relations at the pseudo-wells Upper Jurassic shales (Table 3). Elsewhere, values of
have been derived from a previous temperature map- 50/50 type II/III kerogen were assigned.
ping of the Danish Central Though (Klint Jensen,
1983).
The geochemical character of the Late Jurassic
shales was examined from pyrolysis data (E. Thom-
U. Kimm. CI. FM
sen, 1988) with the objective of discriminating the
1-1 0 L - - -
proportions of kerogen types I, II and III in the
shales. Consistent with earlier studies (Lindgreen et
al., 1982; E. Thomsen, 1989; R. Thomsen, 1989), the
Kimmeridge clay equivalent (Farsund Formation)
shows good to excellent source capacity, and data fa-
vor type II kerogen with some type III kerogen (Fig.
4) (Lindgreen et al., 1982; Goff, 1983; E. Thomsen et
al., 1983; Conford, 1984; Waples, 1985; E. Thomsen,
1989; R. Thomsen, 1989). The underlying Oxfordian
clay has less source capacity than the Kimmeridge
clay equivalent and is gas prone (Conford, 1984). As 3
input for the hydrocarbon generation model we used I
o 50
"X
H
TABLE 2 o
Reflector ages, lithologies derived from well ties and interval
o
velocities used in the depth conversion

Reflector Age Lithology Interval velocity

(Ma) (m/s)
1 0 water 1480
2 1.65 sand 2127
3 10.2 shale 1920
4 26.5 shale 2032
5 36 shale 2173
6 60.2 shale 2560 'I'I'Ir
100 200
7 92-96 chalk 3891 mg C0 2 /gTOC
8 128-131 shale 2525
9 136-145 shale 2908
10 sandy shale 3385 Fig. 4. Modified van Krevelen diagram for the Upper Jurassic in
7-175
well 1-1 showing mainly type II kerogen.
444 R.O. Thomsen, I. Lerche and JA. Korstgrd

TABLE 3 The overall structural pattern changed to a more

Proportions of different kerogen types for the Upper Jurassic NW-SE trend.
shales
During the early Late Cretaceous the subsidence
Type I Type II Type III pattern became more uniform in character. The
0.0 0.4 0.6 major subsidence occurred south of the Inge High
0.0 0.3 0.7 across the Heno Plateau and the Outer Rough Basin
0.1 0.3 0.6
where subsidence rates exceeding 35 m/m.y. are
0.0 0.5 0.5
0.0 1.0 0.0 typical. In the Heno Plateau, Gertrud and Feda
0.0 0.2 0.8 Grabens, subsidence rates reached 17-24 m/m.y.,
0.0 1.0 0.0 with the main subsidence where these structural el-
0.0 1.0 0.0
0.0 0.8 0.2
ements meet. Most activity along the Jurassic faults
had ceased at this stage, but renewed inverted move-
Source: R.Thomsen (1989).
ments along the boundary faults of the Arne-Elin
Graben are inferred by a difference between subsi-
Burial history-geohistory dence rates in the Arne-Elin Graben and the sur-
rounding areas (Fig. 5a, b).
Using the one-dimensional model, the basin sub- This subsidence pattern continued with only minor
sidence was modelled and expressed in terms of changes throughout the Maastrichtian and Danian.
basement subsidence rates in m/m.y. The results are Rapid subsidence with rates of 35-38 m/m.y. contin-
presented as contour maps of the basement subsi- ued across the Outer Rough Basin, the Inge High,
dence rate at different times in the past. This kind of and the Heno Plateau, and in the former Gertrud
mapping enables us to interpret the modelled subsi- Graben. In the former Arne-Elin Graben the rate of
dence in terms of tectonic activity of the basement. 15 m/m.y. outlines the NNW-SSE structural trend
The contour outline of Late Jurassic basement of inversion tectonics, dividing the area into eastern
subsidence rates (Fig. 5a, b) is characteristic of strong and western areas of major subsidence (Fig. 5a, b).
differential subsidence with closed contours confined The post-Danian Tertiary is characterized by rapid
within distinct areas. Rates exceeding 30 m/m.y. oc- regional subsidence with rates of 30-45 m/m.y. to-
curred in the Tail End and Arne-Elin Grabens; in talling 900-1300 m, interrupted only by a minor
the northern Gertrud Graben a total basement sub- quiet period from Early Miocene to Late Miocene.
sidence of about 1100 m occurred at rates of over 40 The most rapid subsidence took place from post-
m/m.y. The S0gne Basin and the eastern part of the Middle Miocene to Present, where the basement
Gertrud Graben both subsided at lower rates than dropped more than 1100 m. During the Pleistocene
the Tail End Graben. The Inge High was at this stage subsidence rates extended from 55 m/m.y to near
already established as a structural high, while the 300 m/m.y.
adjacent Feda Graben, Outer Rough Basin and the
Heno plateau underwent subsidence at intermediate
Thermal history
rates (15 m/m.y.). The Gert Ridge subsided only
slowly (4-10 m/m.y.). The Late Jurassic structural The thermal history is assessed differently in the
pattern was mainly NNW-SSE with minor NW-SE two models, in the one-dimensional model by in-
striking faults. Movement was along the boundary version (Lerche, 1988a, 1988b) of measured vitrinite
fault between the S0gne Basin-TMl End Graben and reflectance with depth for several wells (Table 1)
the area east of Gertrud Graben, normal faulting using an exponential heat flux relation:
occurred along the Inge High boundary faults, and
movement along the faults marking the outline of Q(t) = Q0 exp(/?i) (1)
the Gert Ridge (Fig. 5a, b).
A significant change in the subsidence pattern where Q(t) is the basement heat flow and QQ is the
is seen during the Early Cretaceous stage. Low to present day heat flow, t is time and is a constant
intermediate basement subsidence rates (5-15 m/ determined by minimizing the difference between
m.y.) are dominant throughout most of the area. predicted and measured vitrinite reflectance with
Maximum subsidence rates of 25 m/m.y. are seen in depth (Lerche et al., 1984). In the two-dimensional
the Gertrud Graben, and 20 m/m.y. in the central model the heat flow history is fixed as an input
part of the Tail End Graben and in the southern part parameter.
of the Arne-Elin Graben (Fig. 5a, b). The Gertrud The present day heat flow, 0 , is derived from
Graben developed as a half graben with decreasing downhole temperature logs (real wells), and by con-
subsidence towards the southern part of the graben. tour interpolation (pseudo-wells). A contour map of
 b
1.

Co
S'

Fig. 5. (a) Subsidence patterns based on basement subsidence rates (in m/m.y.) for different time periods: Late Jurassic, Early Cretaceous, Early Late Cretaceous and Late Cretaceous-Danian.
 -P*

O Pseudo well

Well

^ I n c r e a s i n g Subs.

/ A c t i v e Fault

b
1

00
Fig. 5 (continued), (b) Fault activity for the same time periods as in (a).
The Danish North Sea: a basin prognosis 447

The modelled heat flow histories for the nine

wells with R0 data show consistently cooler pasts
with heat flows approaching zero between 35 and
80 Ma ago (Fig. 7). This fact implies that vitrinite
reflectance has no resolution of the early heat flow
history. The reason is the dramatic Tertiary drop
of the basin causing rapid deep burial of the pre-
Tfertiary sediments. A consequence is rapid heating
of the sediments, which is so dominant that the
vitrinite reflectance data has lost resolution of any
earlier heating. This behavior also indicates that any
significant maturation must be a result of the very
recent burial and heating.
Since the inversion of vitrinite reflectance data
failed to resolve any early thermal history and the
Fig. 6. Contour map of the present day heat flow. present day heat flow is close to 1 HFU (heat-flow
units, 1 HFU = 10~6 cal/cm/s), a constant heat flow
of 1 HFU was chosen as thermal history input in the
the present day heat flow (Fig. 6) shows an almost two-dimensional model.
uniform heat flow throughout the basin with a lowest
heat flow in a N-S trending zone in the middle of Maturity
the study area. The assessed low heat flow is due
to overpressure in the shales causing a decreased The maturation history of a given source rock can
thermal conductivity. be expressed in various ways such as a thermal

L-ulu Age (My) Mona Age (My)

175 150 !25 100 75 50 25 0 150 125 100 75 50 25 0
1 1 1 1
90 - ' 9.0 2 9.0 _ I O
9 0 <^
5 8.0
- 8.0 ~ S 8.0 - 8.0V
7.0
- 7.0 * 7.0 - 7.0 5
-6.0
- 60 -6.0 - 6.0^

40 -- -
o5.0 5.0 0 5.0 5.0 0
S 4.0 2 Z 4.0 - 4.0-c
f 3.0 - 3.0 c | 3.0 - 3 0 |
2.0 - 2.0 2.0
- 2.0
1.0
~ 1.0 1.0 - 1.0 -
c) 25 50 75 100 125 150 175 c) 25 50 75 100 125 150
Time (My) Time (My)
Sten Age (My) Q Age (My)
150 125 100 75 50 25 r
) 75 150 125 1OO 75 50 25 C
1 1 1 1 I
9.0 ' . 1
9.0 m 9.0 9.0 ~
5 8.0 -8.0 2 D 8.0 8.0O
7.0
- 7.0 ; 7.0 7.0 ;
-6.0 6.0 5 1 6.0 - 6.0
o5.0
- 5-0 0 0 5.0 5.0-
Z4.0
*3.0
-4.0 n.0
I 3.0
- 4.0.2
3.0 c
3.0 c
2.0 _2.0 $ 2.0 - 2.0
1.0 1.0 1.0 - i.o
C ) 25 50 75 100 125 150 c 25 50 75 100 125 150 175
Time (My) Time (My)
I Age (My) w Age (My)
li50 125
1
100
1
75
1
50
1
25
1
C
) 30 0 250 200
1
150
I
100
1
50 C )
9.0 9.0 X 9.0 - 1
' 9.0
38. 8.0^ -8.0 - 8.0
7.0 7.0 ; ^ 7.0 7.0 "
-6.0 6.0 3 6.0 _ - 6.0 |
0 5.0 5.0 B 1 5.0 - - 5.0 r
40 T- 4 . 0 -
^ 4.0 -3 04.0 .
$3.0 - 3.0 c * 3.0 - K
2.0 - 2.0 % 2.0 - 2.0 |

s
1.0 1.0 1.0 1.0
1
c) 25 50 75 100 125 150 C ) 50 100 150 200 250 30 0
Time (My) Time (My)

Fig. 7. Heat flow variation with time assessed from inversion of measured vitrinite reflectance.
 448 R.O. Thomsen, I. Lerche and JA. Korstgrd

day. Present day oil generation rates show the high-

est generation rates in the western half of the area.
At 10 Ma the highest oil generation rates are seen
in the eastern part of the area, implying that oil
generation had peaked in the eastern part some-
time between 10 Ma and the present day, whereas
the western parts of the area are approaching peak
generation at present. This pattern is consistent with
the maturation trend seen in the time contour map
of the Late Jurassic reaching a vitrinite reflectance
of 0.6 (Fig. 8). Compared to the implications ob-
tained from vitrinite reflectance, with 0.6 chosen as
the lower limit for onset of oil generation, a differ-
ence in timing is observed, with the oil generation
model giving more recent maturation. Both results
Fig. 8. Contours of the time at which the top of the Upper are model dependent, the difference giving a range
Jurassic reached a vitrinite reflectance of 0.6. of uncertainty in assessing the maturation history of
the Upper Jurassic source rock.
In both cases, however, the modelled maturation
indicator value correlated to oil generation. Since stresses the significant impact of the Tertiary deep
vitrinite reflectance is the only thermal indicator burial and resulting heating on the maturation of the
available as a measure of basinal maturity, vitri- Upper Jurassic source rock.
nite reflectance variations with time and depth have
been modelled for wells and pseudo-wells, using the Fluid-flow pattern and possible
one-dimensional model, with a focus on the Upper migration paths
Jurassic shales. A vitrinite reflectance of 0.6 has been
chosen as the onset of oil generation, and 1.35 as the Once hydrocarbons are being generated in the
end of oil generation and onset of wet gas generation source rock they tend to migrate out of the source
(Waples, 1985). rock towards possible reservoirs. It was long being
A time contour map showing the time at which believed that the buoyancy effect induced by the
the Late Jurassic shales reached full maturity (RQ difference in specific gravity between oil (or gas) and
= 0.6) has been constructed (Fig. 8). Full maturity water (known as the "gravitational" theory (Hubbert,
(R0 = 0.6) was reached first in the N and E 30 Ma, 1953)) was the sole driving force for hydrocarbon
while the W, SE and SW reached full maturity migration. This belief is, however, only valid in the
only 5 Ma ago, indicating that maturation moved special case where the pore water is in hydrostatic
through the basin from north and east towards west, equilibrium (Hubbert, 1953). The deposition, and
southwest and south. In the two-dimensional model later compaction, of sediments in a basin provides a
the maturation of the area was expressed in terms of major water flow throughout the basin's geological
temperature evolution with the "oil window" chosen evolution. Thus, as hydrocarbons are generated in a
between 100C and 150C. This approach gives a basin they are subject to both the intrinsic buoyancy
consistent result with the Late Jurassic entering the drive as well as to a significant hydrodynamic drive.
"oil window" between 10 Ma and 1.7 Ma in the Tail Consideration and evaluation of possible migra-
End Graben, and between 1.7 Ma and present day in tion paths from the generating parts of the basin
the western part of the modelled section (Fig. 9). towards possible reservoirs should therefore include
an evaluation of the fluid-flow pattern in the basin
Oil generation throughout the time of migration. The change in
porosity, permeability, fluid pressure and excess fluid
The hydrocarbon potential is assessed by mod- pressure through time was simulated with the pur-
elling the oil generation through time using the pose of monitoring temporal changes in the above
simple scheme of Tissot and Weite (1978). The oil parameters in the Upper Jurassic shales. Because the
generation history is expressed as oil generation rate timing of hydrocarbon generation is between 30 Ma
in mg/g/m.y. (Fig. 10) Contours of this measure for and the present day, these changes were only mod-
the Upper Jurassic sediments at different times show elled through the Tertiary. Simulation of the porosity
that no significant oil generation took place in the variation gave 47% at the top of the Jurassic shale
Jurassic sequence until between 10 Ma and present and 27% at the bottom at 60 Ma, with a decreasing
The Danish North Sea: a basin prognosis 449
TEMPERATURE DISTRIBUTION

52.0 Ma. BP 50.0

100.0
150.0

Udll-OCC.-
C

Top part of
Upper Jurassic

Fig. 9. Two-dimensional simulation of line RTD-81-23 showing the temperature distribution at 52 Ma, 1.7 Ma and present day. The
Upper Jurassic reached a temperature of 100C some 2 Ma ago.
450 R. O. Thomsen, I. Lerche and JA. Korstgrd

of the fluid-flow pattern in terms of a lateral com-

ponent acting as a hydrodynamic drive on migrating
hydrocarbons (Fig. 11) and reveal potential migra-
tion paths for hydrocarbons. The low excess pressure
areas are in the S0gne Basin around the Lulu-l well,
across the southern Tail End Graben from the Adda
wells to the wells H-l and Bo-1, and in the Arne-
Elin Graben around well 1-1. High excess pressures
are found in the deep parts of the central Tkil End
Graben and in the Gertrud and Feda Grabens, in-
dicating a flow of fluids out of the central Tail End
Graben, Gertrud Graben and Feda Graben towards
the areas of low excess pressure.
The two-dimensional simulation illustrates this be-
havior slightly differently in that the highest excess
MA
P 5
fluid pressure gradient is seen out of the deepest part
of the Tkil End Graben towards the Ringk0bing-Fyn
High, indicating a significant lateral component of
flow in this area.

Evaluation

By examining the excess fluid pressure map and

fluid-flow pattern in conjunction with the oil gener-
ation maps, it is seen that oil generation occurred
in the areas of high excess pressure. Thus, the gen-
erated hydrocarbons flow towards the areas of low
excess pressure following migration paths outlined
by the fluid-flow pattern. The low excess pressure
areas with converging flow act as accumulation sites
and are shown in Fig. 12. Hydrocarbons generated
in the Gertrud Graben migrate towards the Otto salt
Fig. 10. Contour maps of oil generation rates (in mg/g/m.y.) structure where the excess pressure favors possible
for the Upper Jurassic shales at 10 Ma and the present day.
The maps show that oil generation peaked in the eastern part accumulation only in the Upper stratigraphic levels
between 10 Ma and the present day, while not reaching peak (area 2 on Fig. 12). Hydrocarbons generated in the
generation in the western part until the present day. central Tkil End Graben migrate towards the Lulu-l
well (area 1 on Fig. 12), the inverted structure of the
Arne-Elin Graben (area 2 on Fig. 12), the eastern
porosity with time until the present day yielding val- boundary fault along the Ringk0bing-Fyn High (area
ues of 32% at the top and 17% at the bottom of the 4 on Fig. 12), and an area between the Adda wells,
shale. The modelled present day porosities lie within well H-l and well Bo-1 (area 3 on Fig. 12).
the range of porosities assessed from wireline logs in
different wells (3%-30%) (Michelsen, 1982). A potential target
The modelled permeability gave a range from 2.10
mD (top) to 0.20 mD (bottom) at 60 Ma, decreasing The most prospective areas, as indicated by the
from 0.35 mD (top) to 0.10 mD (bottom) at the model, are shaded in Fig. 12. Three of these areas
present day. A permeability of 3 mD is reported have already been drilled: (1) Lulu (Lulu-l); (2)
from the J-2 Middle Jurassic sandy shale in the well Arne (1-1, Q-l, T-l, Otto-1); (3) Roar and Adda
U-l southeast of the study area (Michelsen, 1982). (H-l, Bo-1 and Adda-1, Adda-2) with hydrocarbons
The excess fluid pressure is the driving force for found. Hydrocarbon migration towards the eastern
fluidflow,with fluids flowing from high excess pres- boundary fault of the Tail End Graben along the
sure towards low excess pressure. A significant build- Ringk0bing-Fyn High indicates possible accumula-
up of excess fluid pressure in the Late Jurassic shales tion sites in the area marked 4 in Fig. 12. Any
took place between 60 Ma and the present day. Maps hydrocarbon accumulations in this area would likely
of excess fluid pressure allow for an interpretation be sealed either by the tight chalk or, alternatively,
 b
.

>:
I ^3

CJ1

Fig. 11. (a) Contour maps of the excess fluid pressure for the depths 2500 m, 3000 m, 4000 m and 5000 m.
 >3

Sis
Fig. 11 (continued), (b) Interpreted pressure drive for lateral fluid flow assessed from the contoured excess fluid pressure.
The Danish North Sea: a basin prognosis 453

50.0
100-0
150.0
36.0 Ma BP 200.0
250.0

Fig.ll (continued), (c) Excess pressure from the two-dimensional simulation of line RTD-81-23 showing commencement of a build-up
of excess fluid pressure around 36 Ma.

the slightly flexured chalk could leak hydrocarbons Conclusions

through to Early Tertiary sand tongues, sealed by
overlying shales. The consistency between the ob- The subsidence pattern though time from the
servations and the model predictions for areas i, 2 mid-Jurassic to the present day indicates that the
and 3 in Fig. 12, lends confidence to the suggestion northern Danish Central Graben developed through
that the fourth area along the TM1 End Graben- three stages of subsidence: a Late Jurassic differen-
Ringk0bing-Fyn High boundary fault is a high tial stage, a late Cretaceous/Early Tertiary uniform
potential, low risk exploration target. stage and a Late Miocene/Quaternary uniform stage.
454 R.O. Thomsen, I. Lerche and JA. Korstgrd

v^>\ l |
^ >

/% Hp\
56 \ \ .
Si \
^W^
>JX
JURASSIC
4

\ ( ^ L T *

T\
Structural elements

Legend

[Zj/ Migration path

I 0I 0
/>. Hydrocarbon
\^J accumulation 20 km \ \ s
ffi Salt structure
1 \ ^ ^ ^ ^
w Normal Fault \
-imr Reverse fault
==^ Strike slip fault
'/

Fig. 12. Map showing the most prospective areas, as indicated by the model (shaded areas). 1 Lulu; 2 = Arne; 3 = Adda (east) and
Roar (west), 4 = Ringk0bing-Fyn High-Tail End Graben boundary.

The porosities, permeabilities and excess pres- meridge clay), while type III kerogen predominates
sures within the Upper Jurassic shales indicate that in the Late Oxfordian.
fluid flow apparently did not change considerably A Tissot-Weite oil generation kinetic model
since Early Tertiary. Due to a significant pressure showed peaking in the eastern half of the basin
driven lateral component, fluid flow is directed out between 10 Ma ago and the present day, the west-
of the Gertrud Graben and central Tail End Graben ern half not reaching significant generation until the
towards areas of low excess pressure located in present day.
the S0gne Basin, the Arne-Elin Graben, the Heno The integrated evaluation showed that hydrocar-
Plateau and the southern Tail End Graben. The ther- bon migration was from the Gertrud Graben and
mal history was determined by inversion of vitrinite central Tfail End Graben towards the Lulu, Arne,
reflectance data from several wells. Resolution of Roar and Adda structures, and towards the eastern
thermal history degrades beyond 30-50 Ma due to boundary fault of the TM1 End Graben.
the recent deep burial and heating. Maturation of
pre-Cretaceous rocks took place within the Tertiary.
Acknowledgements
Modelled maturation in terms of vitrinite re-
flectance indicates that the Upper Jurassic shales We wish to thank The Statoil Group, Denmark:
entered the R0 = 0.6 "oil window" between 50 Ma Total CFP, Denerco K/S, LD Energi A/S, EAC En-
and 5 Ma. ergi A/S, BHP Petroleum (Denmark) Inc., Dansk
Pyrolysis data from nine wells showed that the Olie-OG Gasproduktion A/S, Statoil Efterforskning
Upper Jurassic shales contain kerogens of both types OG Produktion A/S, The Danish Research Academy
II and III. Type II kerogen predominates in some and Aarhus Universitet for financial support; the
stratigraphic levels in the Farsund Formation (Kim- Basin Modelling Group, University of South Car-
The Danish North Sea: a basin prognosis 455

olina, Columbia, SC, U.S.A., for help and advice and Michelsen, O. (Editor), 1982. Geology of the Central Graben.
Microlith, Denmark for providing the ISOPAC 3.0 Danm. Geol. Unders., Ser. B, No. 8, C.A. Reitzels Forlag,
Copenhagen.
software for constructing the contour maps.
Michelsen, O., Frandsen, N., Holm, L., Jensen, T.F., M0ller,
J.J. and Vejbaek, O.V., 1987. Jurassic-lower Cretaceous of
References the Danish Central Trough depositional environments,
tectonism, and reservoirs. Danm. Geol. Unders., Ser. A, 16,
Andersen, C , Olsen, J.C., Michelsen, O. and Nygaard, E., 45 pp.
1982. Structural outline and development. In: O. Michelsen M0ller, J.J., 1986. Seismic structural mapping of the Middle and
(Editor), Geology of the Danish Central Graben. Danm. Upper Jurassic in the Danish Central Trough. Danm. Geol.
Geol. Unders., Ser. B, 8: 9-26. Unders., Ser. A, 13: 1-42.
Cao, S., 1985. A quantitative dynamic model for basin analysis. Nakayama, K. and Lerche, I., 1987. Basin analysis by model
Master's Thesis, University of South Carolina, Columbia, S.C. simulation: effects of geological parameters on 1-D and 2-D
(unpublished). fluid flow systems with application to an oil field. Trans.
Cao, S., Glezen, W.H. and Lerche, I., 1986. Fluid flow, hy- GCAGS, XXXDVIII: 175-184.
drocarbon generation, and migration: a quantitative model Stouge, S., 1988. Biostratigraphy of selected wells from the
of dynamical evolution in sedimentary basins. 18th Annu. northwestern part of the Danish Central Trough. Rep.,
Offshore Technology Conf., Houston, Texas, OTC Paper 5182, CENBAS/Danm. Geol. Unders. (unpublished).
pp. 2267-2276. Stouge, S. and Toxwenius, B., 1988. Chronostratigraphic position
Cao, S. and Lerche, I., 1987. Geohistory, thermal history and of Cenozoic seismic markers of DGU and some log markers.
hydrocarbon generation history of the northern North Sea Rep., CENBAS/Danm. Geol. Unders. (unpublished).
Basin. Energy Explor. Exploit., 5: 315-355. Thomsen, E., 1988. Source Rock Analysis of Jurassic-Lower
Conford, C , 1984. Source rocks and hydrocarbons of the North Cretaceous Deposits in the Wells B-l, Bo-1, E-l, 1-1, Q-l,
Sea. In: K.W. Glennie (Editor), Introduction to the Petroleum W-l, Lulu-1, Mona-1, Sten-1. Results of Leco-Rock-Eval and
Geology of the North Sea. Blackwell, Oxford, pp. 171-200. Optical Maturity Determinations Tabulated Data. Rep.,
Fisher, M.J. 1984. Triassic. In: K.W. Glennie (Editor) Introduc- CENBAS/Danm. Geol. Unders. (unpublished).
tion to the Petroleum Geology of the North Sea. Blackwell, Thomsen, E., 1989, Source rock analysis of Jurassic-Lower
Oxford, pp. 85-103. Cretaceous deposits in the Danish Central Trough. Results
GECO, 1981. Geophysical Company of Norway A/S, Line RTD- of optical and organic geochemical analysis of the wells B-
81-20. 1, Bo-1, E-l, 1-1, Q-l, W-l, Lulu-1, Mona-1, Sten-1. Rep.,
Glennie, K.W. (Editor), 1984. Introduction to the Petroleum CENBAS/Danm. Geol. Unders. (unpublished).
Geology of the North Sea. Blackwell, Oxford, 236 pp. Thomsen, E., Lindgreen, H. and Wrang, P., 1983. Investigation
Goff, J.C., 1983. Hydrocarbon generation and migration from on the source rock potential of Denmark. Geol. Mijnbouw,
Jurassic source rocks in E Shetland Basin and Viking Graben 62: 221-239.
of the northern North Sea. J. Geol. Soc. London, 140: 445- Thomsen, R.O., 1989. The Danish Central Trough: burial history
474. and 1-D modelling Vol. I: An Integrated Quantitative
Gowers, M.B. and Saeb0e, A., 1985. On the structural evolution Analysis. Rep., CENBAS, 66 pp. (unpublished).
of the Central Trough in the Norwegian and Danish sectors Tissot, B. and Weite, D.H., 1978. Petroleum Formation and
of the North Sea. Mar. Pet. Geol., 2: 298-318. Occurrence. Springer-Verlag, New York, N.Y.
Hubbert, M.K., 1953. Entrapment of petroleum under hydro- Ungerer, P., Bessis, F., Chenet, P.Y., Durand, B., Nogaret, E.,
dynamic conditions. Am. Assoc. Pet. Geol., Bull., 37: 1954- Chiarelli, A., Oudin, J.L. and Perrin, J.F, 1984. Geological
2026. and geochemical models in oil exploration, principles and
Klint Jensen, P., 1983. Formation temperatures in the Danish practical examples. In: G. Demaison and R.J. Murris (Ed-
Central Graben. Danm. Geol. Unders., rbog 1982, pp. 9 1 - itors), Petroleum Geochemistry and Basin Evaluation. Am.
106. Assoc. Pet. Geol., Mem., 35: 53-77.
Lerche, I., 1988a. Inversion of multiple thermal indicators: quan- Vejbaek, O.V. and Andersen, C , 1987. CretaceousEarly Ter-
titative methods of determining paleoheat flux and geological tiary inversion tectonism in the Danish Central Trough.
parameters, I. Theoretical development for paleoheat flux. Tectonophysics, 137: 221-238.
Math. Geol., 20: 1-36. Waples, D.W., 1985. Geochemistry in petroleum exploration.
Lerche, I., 1988b. Inversion of multiple thermal indicators: IHRDC, Boston, Mass., 232 pp.
quantitative methods of determining paleoheat flux and geo- Wei, Z. and Lerche, I., 1988. Quantitative dynamic geology
logical parameters, II. Theoretical development for chemical, of the Pinedale anticline Wyoming, U.S.A.: an application
physical, and geological parameters. Math. Geol., 20: 73-96. of a two-dimensional simulation model. Appl. Geochem., 3:
Lerche, I., Yarzab, R.F. and Kendall, C.G.St.C, 1984. Determi- 423-440.
nation of paleoheat flux from vitrinite reflectance data. Am. Weite, D.H. and Yukler, M.A., 1981. Petroleum origin and
Assoc. Pet. Geol., Bull., 68: 1704-1717. accumulation in basin evolution a quantitative model. Am.
Lindgreen, H., Thomsen, E. and Wrang, P., 1982. Source rocks. Assoc. Pet. Geol., Bull., 65: 1387-1396.
In: O. Michelsen (Editor), Geology of the Danish Central Ziegler, P.A., 1982. Geological Atlas of Western and Central
Graben. Danm. Geol. Unders., Ser. B, No. 8. Europe. Elsevier, Amsterdam, 130 pp.

R.O. THOMSEN Geologisk Institut, Aarhus Universitet, DKSOOOrhus C, Denmark; and Department of Geological Sciences, University
of South Carolina, Columbia, SC 29208, U.S.A.
Present address: Saga Petroleum A/S, P.O. Box 490, N-1301 Sandvika, Norway
I. LERCHE Department of Geological Sciences University of South Carolina, Columbia, SC 29208, U.S.A.
J.A. KORSTGRD Geologisk Institut, Aarhus Universitet, DKSOOOrhus C, Denmark
 457

RIFT, a model of sedimentary basin evolution by finite rate,

non-uniform, pure shear extension of the lithosphere
Noelle E. Odling

A model of sedimentary basin formation by finite rate, non-uniform pure shear extension of the lithosphere is described.
This model can be applied to problems of superimposed rifting events and generally non-uniform extension. Existing models
of instantaneous pure shear extension (McKenzie, 1978) and subsequent modifications (Jarvis and Mackenzie, 1980, Royden
and Keen, 1980) have been applied, with success, to many extensional basins. All these models assume a linear, equilibrium
temperature profile throughout the lithosphere prior to rifting. However, where several rifting phases have occurred throughout
the history of a basin, the assumption of an initially linear, equilibrium temperature profile is not necessarily valid. The model
described here is developed from the models of McKenzie (1978) and Jarvis and McKenzie (1980) to calculate the effects of any
extension on any initial temperature profile.
In the present model, the lithosphere is divided into any number of layers (N) of equal thickness, each of which is assigned
an extension factor. In this way, any non-uniform stretching event can be modelled. Initial and thermal subsidences and heat
flow variation are calculated using the equations of McKenzie (1978) and McKenzie and Jarvis (1980) adapted for the multilayer
model. Extension of the lithosphere with a non-equilibrium temperature profile is modelled by adjusting the extension factors to
those necessary to obtain the final temperature profile from the equilibrium state. This extension will in general vary with depth,
a situation that can be simulated using the multilayer model.
Program RIFT (in standard Fortran) has been written to compute the model. The lithosphere is divided into 100 layers, each
of which can be extended by a different amount. Finite rate rifting is modelled as a series of small rifting events, each of which
is followed by a short cooling period. Subsidence values are computed for the water filled basin (tectonic subsidence). However,
prior to the second and subsequent rifting events, the basin is filled to a specified water depth with sediments composed of 50%
shale and 50% sandstone, taking compaction into account. This gives a more realistic estimate for crustal thickness immediately
prior to the next rifting event. The program outputs a sequence of tectonic subsidence and heat flow values at time intervals
specified by the user. The model is used to estimate the probable effects of Triassic rifting on subsidence and heat flow during
and after Jurassic extension in the northern North Sea. Effects are greatest on thermal subsidence and remnant Triassic thermal
subsidence can contribute as much as 30% to Jurassic thermal subsidence when Triassic extension factors are in the region of
1.5 and Jurassic factors around 1.25. The advantages and limitations of the model are discussed.

Introduction Interpretations of deep seismic data for the North

Sea basin have led to different structural models
The presence of large, normally tilted fault block for the North Sea basin that involve either pure
structures underlying thick sequences of largely un- shear extension (e.g., Barton and Wood, 1984; Gilt-
deformed sediments, observed on seismic sections ner, 1987) or low angle shear zones and asymmetric
from some sedimentary basins, indicated the impor- rifting (Beach et al., 1987). However, evidence from
tance of extensional tectonics in the formation of the geometry of the sedimentary basin fill has led
these basins. Subsequently, a quantitative, physical several authors to favour extension involving domi-
model of basin formation based on instantaneous nantly pure shear (Badley et al., 1988; Klemperer,
pure shear extension of the lithosphere was devel- 1988; White, 1988). Thus it seems probable that the
oped by McKenzie (1978). Applications of this model pure shear extension model of McKenzie is a valid
and subsequent modifications (Jarvis and McKenzie, approach to the problem.
1980, Royden and Keen, 1980) have been successful As the resolution and depth of penetration of
in modelling subsidence for various basins, e.g., the seismic data has improved, it has become apparent
North Sea (e.g., Barton and Wood; 1984, Giltner, that basins such as the North Sea are the result
1987), the East Labrador coast (Royden and Keen, of several periods of extension each of which is
1980) and the Pannonian Basin (Sclater et al., 1980). characterized by major normal faulting. In the case

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 457-467. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
458 NoeUe E. Odling

of the North Sea, it is probable that extension started TYiassic rifting on subsidence during and post-dating
very soon after the end of the Caledonian Orogeny the mid-Jurassic extensional event of the northern
in the Devonian and went through several stages of North Sea.
extensional faulting ending with rifting events in the
Tfriassic and mid-Jurassic (McClay et al., 1986). Existing models of extensional basin
In the pure shear extension models of McKenzie formation
(1978) and others, the lithosphere is assumed to be in
an initially equilibrium state (linear temperature pro- The basic model of sedimentary basin formation
file). Extension of the lithosphere produces a passive resulting from thinning of the lithosphere was de-
upwelling of hot asthenosphere and a temperature veloped by McKenzie (1978). Prior to extension, the
anomaly. Isostatic adjustments in response to instan- lithosphere is assumed to be in equilibrium, with a
taneous extension lead to an initial subsidence when linear decrease in temperature from that of the as-
the crust is of normal thickness or thicker and up- thenosphere at the base of the lithosphere to 0C at
lift in the case of thin crust. Cooling, accompanied the surface of the crust. The lithosphere is instanta-
by contraction and density increase, results in sub- neously extended by uniform, pure shear by a factor
sidence and basin formation. Since many basins are (= initial/final lithospheric thickness) resulting in
the result of more than one period of rifting (e.g., a temperature anomaly (Fig. 1A and B). This defor-
Tankard and Welsink, 1987, Badley et al., 1988) the mation is associated with a change in the elevation of
assumption of an equilibrium temperature profile the surface of the crust which is a balance between
prior to later events may not be valid. subsidence caused by replacement of light crust by
In the present paper, a general model of basin dense asthenosphere and uplift due to thermal ex-
formation resulting from non-uniform, pure shear pansion. This change in elevation (termed "initial
extension of the lithosphere, based on the models of subsidence") is, therefore, sensitive to the initial
McKenzie (1978), Jarvis and McKenzie (1980) and crustal thickness and the model predicts uplift if the
Royden and Keen (1980), is developed so that any crust is initially thinner than approximately 20 km.
variation of extension with depth may be modelled. Following extension, the temperature anomaly cools
As in the case of these previous models, the present and subsidence due to thermal contraction forms the
model is one dimensional in the sense that there basin. The cooling history is quantified by solving the
is no lateral variation in extension. As a test case, one-dimensional heat flow equation:
the model is used to estimate the possible effects of

z=0 z=0

;/ / /n=y / / .
//// n=M
/ /
0-

n=l
2=0 z=0

Fig. 1. A, B: Instantaneous uniform, pure shear extension of the lithosphere by a factor (McKenzie 1978). The resulting temperature
profile (B) is linear to the base of the thinned lithosphere below which temperature is constant at , the temperature of the
asthenosphere. C, D: instantaneous, non-uniform, pure shear extension of the lithosphere by factors which, in general, vary with
depth. The resulting temperature profile within the thinned lithosphere is non-linear (D).
RIFT, a model of sedimentary basin evolution byfiniterate, non-uniform, pure shear extension of the lithosphere 459

TABLE 1 resulting in a change in the width of the deformation

Symbol definitions and parameter values (mostly after McKenzie zone. A model in which varies continuously with
1978)
depth has been used by Rowley and Sahagian (1986)
Symbol Definition Parameter value to provide a mechanism for the uplift of rift flanks
crustal thickness 32 km
during extension. For this, they used an analytical
U
a lithospheric thickness 125 km expression for the vertical variation in /?, assuming
Pm mantle density at 0C 3.33 g c m - 3 a linear increase in the width of the extension zone
Pc crustal density at 0C 2.8 g c m " 3 with depth, and used modifications of the equations
Pw density of water at 0C 1.0 g c m - 3 of McKenzie (1978) to calculate uplift and subsi-
a thermal expansion coeffi- 3.28 x 1 0 - 5 o C dence.
cient for crust and mantle The observation of low angle shear zones on land
n thermal diffusivity coeffi- 0.00804 cm 2 s " 1
(Wernicke, 1984) and possible examples on seismic
cient for crust and mantle
K thermal conductivity of 0.0075 c a l " 1 c m " 1 s sections (de Charpal et al., 1978; Beach et al., 1987;
crust and mantle Tknkard and Welsink, 1987) also suggests that decou-
T temperature (C) pling zones which separate domains with differing
t time (Ma) magnitudes of extension, can exist within the litho-
z depth (km) sphere. In the model of Royden and Keen (1980),
extension factor the lithosphere is divided into two layers, separated
by a "decoupling" zone placed, in their case, at the
base of the crust. They then developed expressions
for the Fourier coefficients of the initial temperature
distribution for the two layer case and then ap-
where T is temperature, t is time, z is depth and plied the expressions of McKenzie (1980) to derive
K is the thermal diffusivity of the lithosphere (T&ble heat flow and subsidence with time. By simulating
1). McKenzie (1978) solved this equation by Fourier different magnitudes of instantaneous extension in
expansion and developed expressions for the vari- these two layers, preferential crustal and subcrustal
ation in heat flow and subsidence with time after stretching can be modelled. Preferential subcrustal
instantaneous extension for the case of a water filled stretching was found to reduce initial subsidence and
basin (tectonic subsidence). has been used by Royden and Keen (1980) to ex-
This model assumes instantaneous rifting whereas, plain subsidence patterns from the Labrador Shelf.
in reality, extension takes place over a finite time in- This model has also been employed by White and
terval. When rifting times are short, this assumption McKenzie (1988) to explain the "steer's head" geom-
is reasonable but if extension rates are sufficiently etry of basin fill. They used two different Gaussian
slow, significant heat is lost during the rifting event distributions to describe the lateral variation of in
resulting in significantly different heat flow and sub- the crust and upper mantle, while keeping the total
sidence patterns. Jarvis and McKenzie (1980) have extension of the two layers equal. By stretching the
modified McKenzie's initial model to take finite rift- upper mantle by lesser amounts over a wider area,
ing time into account by assuming a constant rate of uplift of the rift flanks is generated and subsequent
extension and solving the resulting one-dimensional subsidence results in a "steer's head" geometry of
heat flow equation. The result of heat loss during the infilling sediments.
rifting is to increase the subsidence during rifting at
the expense of that during cooling (the total subsi-
The present model
dence remaining the same). The results of this model
show that subsidence and heat flow patterns differ A multilayer lithospheric model for instanta-
significantly from those derived from the instanta- neous extension
neous extension model when the rifting times are The theories of McKenzie (1978) and Royden and
greater than approximately 10 Ma. Keen (1980) have been applied to a lithosphere di-
Further modifications of the instantaneous pure vided into any number of equally thick layers. In this
shear extension model has been developed by Roy- way, any variation in extension with depth, which
den and Keen (1980) and Rowley and Sahagian need not necessarily be described as an analytical
(1986) to deal with non-uniform extension through- expression (as in Rowley and Sahagian, 1986), can
out the lithosphere. As the rheological properties of be modelled. As in previous models, the initial tem-
the lithosphere vary with pressure, temperature and perature distribution of the lithosphere is assumed
chemical composition, it is possible that the degree to be that of the equilibrium state with the temper-
of thinning due to extension may vary with depth, ature increasing linearly with depth from 0C at the
460 NoeUe E. Odling

surface of the crust to , the temperature of the where bk represents the Fourier coefficients of the
asthenosphere, at the base of the lithosphere. On temperature distribution immediately after instanta-
rifting, each layer within the lithosphere is extended neous extension. The &th Fourier coefficient is given
by a factor , resulting in a general non-linear by (Jarvis and McKenzie, 1980: their eqn. 19):

\ +
temperature profile (Fig. 1C and D). 2
This extension is associated with a change in el-
evation which is a balance between uplift due to
; k-
sin- dz (5)

thermal expansion and subsidence due to the re- where T^ is the temperature distribution at time,
placement of light crust by denser mantle material. t = 0. For the multilayer model the Fourier coeffi-
It is assumed that isostatic equilibrium is maintained cients are determined by adding the results of the
and the change in elevation is then calculated by above integral applied to each layer, the derivation
equating the mass of a unit cross-sectional area col- of which is given in Appendix A:
umn to the base of the lithosphere before and after N N
extension. In the case of uniform pure shear, the
'
~ n-l
kw
sin** ( l -1 - 1- (6)
change in elevation, Siy is given by (McKenzie, 1978: n=l
his eqn. 1): Subsidence and heat flow variation with time for
the multilayer model can then be derived using the
(Pm ~ Pc) tc aTipn
equations of Jarvis and McKenzie (1980): their eqns.
Si =
a(l- aTitc/2a) - ? (2) 22, 23, 24 and 27. The elevation above the final level
pm(l - aTi) - p w to which the surface of the crust sinks is (Jarvis and
McKenzie, 1980: their eqns. 23 and 24):
A list of definitions of the variables in the above
and following expressions is given in Table 1. In the -
multilayer model, the contribution of each layer must *(*) =
7r[pm(l - aTi) - Pw + Btc/a)]
be calculated separately and the results summed. The
k2K2Kt
change in elevation is then given by:
bk
exp 1 - COSJ7T 1

*=>(-|+ + )(-
k =l

=
X (pm + Pc) + Pm~ (-l)Vc (7)
M-\ where:
+ I>(i-^ + 1 + ^ ) ( i - l
aT\ Pc_
B = p* 1 + 1 + (Pc ~ Pm)
(1-) Pm
1
pm(l - ^ N - As e(t) approaches zero as t approaches infin-
n=\ ity, the maximum post-extension subsidence (SG) is
a obtained by setting t = 0 in eqn. 7 (Jarvis and
(3)
Nipmil-aT^-pv]
McKenzie, 1980: their eqn. 27):
Equation 3 gives the elevation change when sub-
sidence occurs and the resulting basin is water filled
SG =
(tectonic subsidence). In the case of uplift, when the 7r[p m (l - otTx) - /?w + Btc/a\
basin is not water loaded, the denominator in the bk
above expression becomes N[pm(l - )]. |
>fc = l
1 - coskn 1 -
The instantaneous extension event causes an up-
welling of hot asthenosphere producing a tempera- + Pc COS&7T 1 - (-1)' (8)
ture anomaly, the decay of which results in thermal
contraction and decreasing heat flow, thereby creat- The subsidence at time, f, is then given by:
ing a basin. The solution of the one-dimensional heat
flow equation (eqn. 1) gives the following expression St = e(t) - SG (9)
for the temperature distribution with depth after where subsidence is positive and uplift negative.
time, t (adapted from Jarvis and McKenzie, 1980: Heat flow after time, t, is given by (Jarvis and
their eqn. 18): McKenzie, 1980: their eqn. 22):
(,) = 7(-^) KTi -k27T2tK
CO
F(t) = 1 + ^2 kbk(-l)k+1 cxp
2 2 8
+ **[-* ' '()] ^ (4)
a
k=i ^
k =l
(10)
RIFT, a model of sedimentary basin evolution by finite rate, non-uniform, pure shear extension of the lithosphere 461

T
() T
" T, () r, <) T
" T
\. n=N \ ,(N) V.
t
z
l " "^^
^^ i(D
Cooling
>
\^v
\ >v
- \ - - First extension \ \

a
n=l \
a a A
<7
1
() V
T, () 1
^^^n=N ^ \ z<N)

t
Z
X.
>
2

Second extension 2(l)

\
a
n=l \
a
() T, ( ^ !i ) < T,
\^ n=N nw(N)
XV
4
" "^\ \\^
\ . Cooling \ \\
I \ X.
\ \ \\
\ \ nw(D ^ \ \
a " \ a a x\
Fig. 2. The superposition of two rifting events. Before the temperature anomaly created by non-uniform rifting phase \ decays,
the temperature profile of the lithosphere is non-linear. This profile is then extended by a second rifting phase, 2, to give a new
temperature profile (2). To determine subsidence and heat flow variation during the decay of this anomaly, a second set of extension
factors /?nw are calculated. These are the extension factors required to produce temperature profile (2) from the equilibrium linear
profile. The second set of beta factors do not represent physical thinning of the lithosphere, but are the values required by the model
to calculate changes in the temperature profile with time.

Superimposing two rifting events and finite values required by the model to calculate cooling and
rate rifting resulting thermal subsidence. Using these modified
The temperature profile after an instantaneous extension factors, eqns. 3-10 described above may
rifting event is the result of the initial profile and the then be used to determine subsidence and heat flow
magnitude of the extension. The above expressions after rifting.
for heat flow and subsidence after instantaneous rift- To determine the extension values required, the
ing assume an initial equilibrium temperature profile lithosphere, with any temperature profile, is divided
for the lithosphere. However, if extension follows an into N layers, each of which is assigned an extension
earlier rifting event before the temperature anomaly factor () appropriate for the rifting event. There is
caused by that event has completely decayed, the ini- no theoretical limit to the difference in values of
tial temperature distribution may differ significantly adjacent layers. However, the aim of the model is to
from the equilibrium state. This will result in a differ- simulate smooth changes in with depth and thus
ent temperature perturbation after rifting and thus large jumps in between adjacent layers should nor-
different subsidence and heat flow patterns. Every mally be avoided. Knowing the initial temperature
temperature profile can be described as an extension at the base of each layer and the extension factors,
superimposed on the equilibrium temperature dis- the depth of each temperature value after instanta-
tribution. The magnitude of this required extension neous extension can be calculated. Temperature is
will, in general, vary with depth, a situation that can assumed to vary linearly with depth between these
be modelled using the multilayer model described values which results in a smooth temperature profile
in the previous section. In the model, the effects of if enough layers are used. In the equilibrium, linear
rifting on a non-equilibrium temperature profile are temperature profile, the temperatures at the base of
simulated by calculating a second set of extension each of TV layers of equal thickness are calculated
values which are those required to produce the final and the location of each of these temperatures in the
temperature profile from the equilibrium tempera- final temperature profile is determined. The vertical
ture state (Fig. 2). These recalculated values do not distance between these points with the original thick-
represent physical thinning of the lithosphere but the ness of each layer can then be used to calculate the
462 Noee E. Odling

modified extension values required to produce the The results of RIFT for a single rifting phase were
final temperature profile from the equilibrium state compared to those from the finite-rate, uniform pure
(Fig. 2). shear rifting model of Jarvis and McKenzie (1980).
Using this method, the above equations can be Increasingly better fits were obtained as the number
used to determine subsidence and heat flow variation of increments used to model finite rifting increased.
with time for superimposed rifting events where the RIFT was found to overestimate subsidence by a
second event follows the first before the temperature maximum of 4% for an extension factor of 2.0 when
perturbation has completely decayed. Finite rate rift- 10 or more increments were used. Errors in the heat
ing can be modelled as a series of small rifting events. flow were 1.5% or less.
The method could also be used to model the effects
of rifting on other non-equilibrium temperature pro- The model applied to superimposed
files caused, e.g., by heating events originating in the Triassic and Jurassic rifting in the northern
mantle. North Sea

The existence of a THassic phase of rifting in the

Program RIFT northern North Sea region is well known although
the timing of this event is, as yet, poorly constrained
The program RIFT (standard fortran) has been (Badley et al., 1988). Late Permian to Early THassic
constructed to compute subsidence and heat flow faulting is described by Badley et al. (1984) in the
during one to five rifting events comprising finite Oseberg region, while Late THassic faulting has been
rate, pure shear extension of the lithosphere and recorded from the western part of the Viking Graben
subsequent cooling of the temperature anomaly. The (Davies and Watts, 1977; De'Ath and Schuyleman,
lithosphere is assumed to be in an equilibrium state 1981; Beach et al., 1987; Karlsson, 1986; Gabrielsen
prior to the first rifting event, i.e., that temperature et al., 1989). Estimates for the thickness of THassic
decreases linearly from temperature T\ at the base sediments in the North Sea region range up to 3000
of the lithosphere, to 0C at the surface of the crust. m (Ziegler, 1982) which, if attributed to subsidence
Finite rate rifting is modelled as a series of small, following rifting, indicates that a major extension
instantaneous rifting events each of which is followed event took place in THassic times. THassic rifting
by a short cooling period. The lithosphere is divided was followed by a second and better documented
into 100 layers for the purposes of computation and rifting phase in Late Jurassic times described by
each layer is assigned an independent extension fac- many authors and dated as Bathonian to Ryazanian
tor, (user determined), for each rifting phase. At by Badley et al. (1988).
the end of each finite rate rifting increment and The thinning of the crust in the northern North
rifting phase, the temperature profile and values of Sea since Early THassic times has been estimated
crustal and lithospheric thickness are used as input from the depths of the Base THassic and the Moho,
for the following rifting event. During each rifting determined from reflection seismic and gravity data
phase the basin is assumed to be water filled for the by Klemperer (1988) and Holliger and Klemperer
case of subsidence, giving the tectonic subsidence of (1989). The thickness of this basement ranges from
the basin. However, at the end of a rifting phase, 32 km on the Shetland platform and western Norway
any basin that has formed is likely to be filled with (assumed to be the original crustal thickness) to
sediment, resulting in a modified crustal thickness. In 15 km in the deepest parts of the Viking Graben.
order to model the most realistic situation, the effect This gives a maximum extension factor, , since the
of filling the basin formed by a rifting phase with Early THassic as 2.1. However, discrepancies between
sediment composed of 50% sand, 50% shale to a the Moho derived from reflection seismic data and
user specified water depth are included in the model gravity data have led Holliger and Klemperer (1989)
(method after Sclater and Christie, 1980). The load- to speculate that the lower crust may have been
ing effects of this quantity of sediment are calculated thickened by basaltic intrusions in the deepest parts
using density, initial porosity and lithological com- of the Viking Graben, in which case the basement
paction factor quoted by Sclater and Christie (1980) thickness may be as low as 10 km, implying a factor
and compaction of the pre-rift sequence is also taken of up to 3.2.
into account. The crustal thickness is then adjusted Giltner (1987) modelled subsidence due to THassic
to include this sedimentation prior to the next rifting and Jurassic rifting events in the northern North
event. Other physical input parameters used in the Sea and obtained extension factors of up to 1.8 for
model follow values given in McKenzie (1978) and the THassic event and 1.1 to 1.2 for the Jurassic
are listed in Tkble 1. event (total of 2.16). Beach et al. (1987) obtained
RIFT, a model of sedimentary basin evolution byfiniterate, non-uniform, pure shear extension of the lithosphere 463

a value of 1.5 for Tttassic rifting and 1.8 to 2.2 Early Jurassic (225-200 Ma), were modelled and a
for Jurassic rifting from a subsidence curve from crustal thickness of 32 km (Holliger and Klemperer,
the north Viking Graben (total of up to 3.3). 1989) prior to THassic rifting was assumed. Jurassic
Vially (1988), in a regional study of subsidence in extensions of 1.25 and 1.5 were superimposed on
the northern North Sea obtained maximum THassic two THassic extension events of 1.25 and 1.5 and the
values of 1.5 (Horda platform) and a maximum results are plotted in Fig. 3. These results are com-
Jurassic value of 1.5 (Viking Graben) (maximum pared to those for a single Jurassic event (extension
total of about 2.0). Badley et al. (1988) estimated 15-131 Ma) using an initial crustal thickness of 32
maximum extension factors for Jurassic rifting at km (Fig. 3).
1.15 from initial subsidence and 1.49 from thermal Figure 3 shows that THassic rifting in general
subsidence, as average values for the Viking Graben. causes a decrease in initial subsidence and an in-
All of the above estimates assumed that the effects of crease in thermal subsidence in agreement with ob-
residual THassic thermal subsidence were negligible servations by Barton and Wood (1984) and Hellinger
prior to Jurassic extension. Lippard and Liu (this et al. (1989) on data from the Central Graben of
volume) have modelled THassic and Jurassic rifting the North Sea. The effect on initial subsidence is
in the northern North Sea, using model RIFT and a combination of a decrease caused by a thinner
arrived at THassic values of up to 1.5 and Jurassic crust and an increase induced by remnant THassic
values up to 1.8 (total of 2.7). thermal subsidence. The net result is a decrease of
An attempt has been made by Hellinger et al. up to 6% for the values modelled here, which
(1989) to determine the maximum effect of Car- is greatest when THassic extension factors are high
boniferous to Tfriassic rifting on subsidence during and the influence of a thinned crust is most domi-
and post-dating Jurassic extension in the Central nant. Jurassic thermal subsidence is increased by the
Graben of the North Sea, by modelling combinations addition of remnant THassic thermal subsidence in
of the various rifting events as single phases. For a all cases. The greatest effects occur when Tftassic
Carboniferous to THassic value of 1.69 and a Juras- factors are high (around 1.5) and Jurassic factors
sic value of 1.26, they estimated that up to 40% of are low (around 1.25) when Tftassic thermal subsi-
Jurassic subsidence may be attributable to remnant dence contributes up to 23% (Early Tftassic rifting
thermal subsidence of earlier events. To obtain an phase) and 30% (Late THassic rifting phase) to sub-
idea of the possible effects that THassic rifting may sidence post-dating Jurassic extension. In the case of
have had on subsidence during and after Jurassic total subsidence, the effects on initial and thermal
rifting in the northern North Sea, program RIFT subsidence tend to counteract each other but leave
was used to calculate theoretical tectonic subsidence an overall positive influence with THassic rifting con-
and heat flow curves for the superimposed THassic tributing up to 10% (Early TCassic rifting) to 16%
and Jurassic rifting phases. For simplicity, uniform (Late Tftassic rifting).
pure shear is assumed for both rifting phases. The The effects on initial subsidence of up to 6%
curves obtained for subsidence and heat flow af- probably lie within the errors on subsidence curves
ter the start of Jurassic extension are compared to derived from stratigraphic data and are, therefore,
curves of equivalent value for a single Jurassic rift- negligible in the determination of values from ini-
ing event, as a measure of the effects of the TYiassic tial subsidence. However, initial subsidence patterns
event. This treatment is not intended as a defini- are often complicated by footwall uplift and hanging
tive study of the basin development in the northern wall collapse and may be unsuitable for determina-
North Sea, but rather as an indicator of the probable tion of factors. Because of this, more reliance is
extent of the effects of Tftassic rifting on subsidence frequently placed on the shape of the thermal sub-
and heat flow after the start of Jurassic extension. A sidence curve (e.g., Barton and Wood, 1984). Here
fuller treatment of subsidence in the northern North the increased subsidence caused by remnant THassic
Sea from observed and theoretical subsidence curves subsidence will result in an overestimation of if
derived from program RIFT, including depth depen- observed subsidence curves are compared to those
dent stretching, is given by Lippard and Liu (this for a theoretical single event. Where TViassic exten-
volume). sion is low compared with Jurassic extension, the
A Jurassic extension of 25 Ma duration, from 156 discrepancies are of the order of 10% or less and
to 131 Ma (Bathonian to end Jurassic), (Badley et al., probably largely within the data uncertainties. How-
1988) was used. As the duration of THassic rifting is ever, where THassic extension are high compared
unknown, a value of 25 Ma, as for Jurassic rifting was with Jurassic extensions, TViassic thermal subsidence
assumed. Two "end member" timings of THassic rift- may compose up to 23% (Early THassic rifting) to
ing, Early THassic (250-225 Ma) and Late TYiassic- 30% (Late TViassic rifting), resulting in an overesti-
464 Noelle E. Odling

Time (Ma) Time (Ma)

250 210 170 130 90 50 145 105

v Triassic rifting 250 - 225 Ma.

1 I

-2J

2 3H

0 1.20 0 1.20

0.8<K 0.80
250 210 170 130 90 50 10 225 185 145 105
Time (Ma) Time (Ma)

Fig. 3. Subsidence and heat flow histories for superimposed Triassic-Jurassic rifting (solid curves). Two timings of Triassic rifting are
shown: Early Triassic, 250-225 Ma (left); and Late Triassic to Early Jurassic, 225-220 Ma (right). Jurassic rifting with extension factors,
/?, of 1.25 and 1.5 are superimposed on Triassic rifting with extensions of 1.25 and 1.5. The two-phase model curves are compared to
those for a single Jurassic event of equivalent values calculated with an initial crustal thickness of 32 km (dashed curves).

mation of Jurassic factors of the order of 0.1. The specifically designed to model superimposed rifting
effects of remnant THassic thermal subsidence on events where the temperature anomaly of an event
heat flow after the start of Jurassic extension is also has not completely decayed before the onset of the
shown in Fig. 3, together with the heat flow curves next.
for a single rifting event. The greatest contribution As a test case, the model has been used to assess
of THassic rifting occurs at the start of Jurassic exten- the possible influence of THassic rifting on subsi-
sion where it composes 11% ( of 1.25) to 18% (/? of dence and heat flow during and after Jurassic rifting,
1.5) of the total. In all cases, this decreases to 2% or in the northern North Sea. Two timings of THassic
less 100 Ma after Jurassic extension. rifting were modelled, Early THassic (250-225 Ma)
The above examples predict a present day, pre- and Late Triassic (225-200 Ma). As would be ex-
THassic basement thickness of 14.2 km in the central pected, the effects of Late THassic rifting were the
parts of the Viking Graben and 20.5 km on its larger, but only small differences were seen in the
flanks. This is in good agreement with the values results, up to 4% for initial subsidence and up to 9%
determined from seismic reflection data (Klemperer, for thermal subsidence. The effects of THassic rifting
1988) of less than 16 km in the central parts of the on initial Jurassic subsidence, when compared to a
graben and around 21 km on the flanks. Present single rifting event with an initial crust of 32 km, were
day total crustal thicknesses predicted by the model very small (6% or less). This is due to the effects of
range from 26.5 km to 28.2 km. This agrees quite the crust being thinner than initially (reducing sub-
well with the thicknesses presented by Klemperer sidence) which counteracts the effects of remnant
(1988) on the flanks of the Viking Graben (around thermal subsidence (enhancing subsidence). Rem-
26 km), but is larger than Klemperer's estimate (less nant THassic thermal subsidence generally increases
than 21 km) in the graben centre. A possible reason thermal subsidence post-dating Jurassic extension
for this discrepancy could be that the crust is more and makes the largest relative contribution when TH-
extended than the underlying mantle in this region, assic extension factors are high compared to Jurassic
resulting in less thermal subsidence. values, as has been suggested for the Viking Graben
(Giltner, 1987) and the Horda Platform (Giltner,
Discussion and conclusions
1987; Vially, 1988). For a THassic value of 1.5 and
a Jurassic value of 1.25, THassic remnant thermal
The model presented can be used to predict sub- subsidence contributes 23-30% of Jurassic thermal
sidence and heat flow patterns arising from any subsidence. This could cause an overestimation in
non-uniform, finite rate pure shear extension of the factors estimated using a single phase model of
lithosphere. The program that computes the model is about 0.1.
RIFT, a model of sedimentary basin evolution byfiniterate, non-uniform, pure shear extension of the lithosphere 465

Thus, the influence of TYiassic rifting results in an widening of the deformation zone in the mantle with
increase in the total subsidence and a change in the depth. Modelling presented here tentatively suggests
relative proportions of initial and final subsidence. that the crust may be preferentially stretched in the
This may go some way to explaining the discrepan- deeper parts of the Viking Graben which would also
cies between factors determined from initial and be consistent with this model. Thus, the models of
thermal subsidences, where a single rifting model is Rowley and Sahagian (1986) and White and McKen-
employed (Barton and Wood, 1984; Badley et al. zie (1988) may go some way to explaining some of
1988). A method that has been used frequently to the observations and modelling results of the above
improve the fit of observed and single phase theoreti- authors. However, it is clear that more data and
cal subsidence curves is to reduce the initial thickness detailed subsidence modelling will be required be-
of the crust, either using the subsidence curves as a fore the complex question of uniform versus depth
guide (Barton and Wood, 1984) or using estimates of dependent stretching in the North Sea region can be
TYiassic extension (Giltner, 1987). This will tend to resolved.
lead to an overestimation of factors as it ignores The model RIFT assumes Airy isostacy, i.e., that
the effects of remnant Tfriassic thermal subsidence. isostatic equilibrium is maintained throughout the
A uniform pure shear model was used in the test history of the basin and that any flexural loading ef-
cases for the northern North Sea. Most authors mod- fects of sediments are negligible. The question of the
elling subsidence in the North Sea region have used importance of flexural loading has been addressed by
a uniform pure shear model (e.g., Barton and Wood, Barton and Wood (1984) and Hellinger et al. (1989)
1984; Giltner, 1987; Hellinger et al., 1989). Badley et for the central North Sea and Fowler and McKen-
al. (1988) concluded from the mismatch between zie (1989) for the Rockall Plateau. They estimated
factors predicted from initial subsidence and thermal the elastic thickness of the crust in both areas to
subsidence using a single phase, uniform pure shear be around 5 km, a very small value, and concluded
model (0.1 to 0.3) in the northern North Sea, that a that there is therefore little error in assuming Airy
depth dependent stretching model involving prefer- isostacy. In the northern North Sea, Holliger and
ential mantle stretching, is more appropriate. How- Klemperer (1989) found the sediments of the Viking
ever, such a model of preferential mantle stretching Graben to be isostatically compensated implying that
everywhere in the basin results in space problems the strength of the lithosphere in this region is dras-
or, if mass is conserved, in a narrowing of the de- tically reduced. Thus it seems probable that, as in
formation zone with depth, an improbable geometry. the central North Sea, there is little error in assum-
For the Horda Platform to the east of the Viking ing Airy isostacy in the northern North Sea region.
Graben, Lippard and Liu (this volume) have used a Model RIFT is not applicable if the effects of flexural
two layer model with preferential mantle stretching, loading are considered important or in areas that are
in agreement with Badley et al. (1988), to explain known to be isostatically uncompensated, such as the
subsidence curves. Inner Moray Firth basin (Barr, 1985).
Models which predict a range of preferential The model RIFT is one dimensional in that it can
crustal and mantle stretching across the basin have be used to model the subsidence of only one point
been produced by Rowley and Sahagian (1986) and at a time within the basin, as is the case with the
White and McKenzie (1988). In these models, the model of McKenzie (1978) and the other modifica-
crust is preferentially stretched in the graben cen- tions described above. Two dimensional sections and
tre and the mantle preferentially stretched on the maps must be constructed from one-dimensional re-
graben flanks. In the model of White and McKenzie sults as in Barton and Wood (1984), Giltner (1987),
(1988), the zones between these two regions have ap- Hellinger et al. (1989) and others. Questions of area/
proximately uniform pure shear. Space problems are mass balance and the implied shape of the extending
avoided as total extension is the same at all depths. zone must, therefore, be addressed separately.
The vertical and horizontal distributions of imply a As with the models of McKenzie (1978), Jarvis and
widening of the deformation zone with depth. Deep McKenzie (1980) and Royden and Keen (1980), only
reflectors in the upper mantle were identified on vertical heat conduction is considered. Thus, effects
deep seismic reflection profiles by Klemperer (1988) from lateral heat conduction are not included. Such
and interpreted by him as probably representing effects could be expected to increase heat loss (and
shear zones. Their location at the margins of the therefore the subsidence rates) at a rift centre and
major Mezozoic extensional faulting in the north- to retard heat loss (and therefore decrease subsi-
ern North Sea and sense of dip outwards from the dence rates) from the rift flanks. Extension factors
faulted zone, although implying dominantly uniform derived from RIFT would then be overestimated in
pure shear in the crust, are also consistent with a the graben centre and underestimated on the flanks.
466 Noelle E. Odling

However, in the case of the northern North Sea,

the generally good agreement between the range of
estimated factors for Tftassic and Jurassic exten-
sion events (Giltner, 1987; Beach et al., 1987; Vially,
1988; Lippard and Liu this volume) and estimates
of the pre-Tftassic basement thickness (Klemperer,
1988) would suggest that this is not an important
factor. Model RIFT is not applicable if large magma Fig. A-1. Temperature profile of the lithosphere before and after
bodies in the lower crust are associated with rifting as non-uniform, pure shear extension. The equilibrium lithosphere
(linear temperature profile) is divided into N layers, each of
this would produce different heat flow and therefore which is independently assigned an extension factor n. Since the
subsidence patterns (Royden and Keen, 1980). extension factor can vary with depth, this results in a generally
non-linear temperature profile through the thinned lithosphere.

Acknowledgements
The general expression for the kth Fourier coef-
The author would like to thank I.K.U., owners of ficient, bk, of this temperature distribution is (Jarvis
program RIFT, for permission to publish the model. and McKenzie, 1980, their eqn. 19):
I would also like to thank Alastair Jenkins for his
mathematical help during the model construction 2 fa lm m fz 1 k^z , /K ,s
bk =
and 0ivind Sylte and Guojiang Liu for checking the
model algorithms. I thank S. Lippard for discussions In the multilayer model, the contribution that the
on North Sea geology. nth layer makes to this integral is:

s[*MJU)
Appendix A: Derivation of the Fourier coefficient
for the initial temperature distribution in the
multilayer model case
z n .
sin
J7TZ ,
dz (A-5)
The temperature of the lithosphere in the equilib- a
rium state is assumed to increase linearly from t =
where X and Y are, respectively, the depths to the
0C at the surface of the crust (z = a) to t = T\ at
base and top of the nth layer:
the base of the lithosphere (z = 0). The lithosphere
is divided into N number of layers of equal thickness 1 N
1
a/n with layer 1 at the base of the lithosphere (Fig.
A-1). In the equilibrium state, the base of the nth
layer lies at depth: and
N
(A-1) 1 1
- ^ - D
N L, m
m=n + l
The temperature at depth, z, in the nth layer is:
Evaluating the integral in eqn. 5 gives an expres-
r
-*(-;)**('-) sion for the nth contribution to the &th Fourier
coefficient:
After instantaneous extension of the lithosphere,
(nY ~ ) [COS^TrX - COS^TTY]
each layer is extended by factor /?, where is
the ratio of initial to final layer thickness. The new +
distance to the base of the nth layer is: ^ - ) -sinknY sinkwX
k-K k-

4- X cos kwX - Y cos kwY (A-6)

(A-2)
The final (n + l)th contribution comes from the
and the temperature at depth, z, in the nth layer is: portion of the asthenosphere lying between the orig-
inal base of the lithosphere (z = 0) and the base of

1 layer 1 (z = zi):
[
W
-"HJU-:-)- *-*) (A-3) 2
a Jo
21
. k-
z sin
a
J
dz
RIFT, a model of sedimentary basin evolution byfiniterate, non-uniform, pure shear extension of the lithosphere 467

where: Gabrielsen, R.H., Faerseth, R.B., Idil, S. and Steel, R., 1989.
Architectural styles of basin fill in the North Viking Graben.

z1=a[l i f J-N Terra Abstracts, 1: 33.

Giltner, J.P., 1987. Application of extensional models to the
Northern Viking Graben. Nor. Geol. Tidsskrift, 67: 339-352.
Hellinger, S.J., Sclater, J.G. and Giltner, J.P., 1989. Mid-Jurassic
Evaluating this integral gives: through mid-Cretaceous extension in the Central Graben of
the North Sea, Part 1. Estimates from subsidence. Basin Res.,

2_
,
1
sinkn 1- w , i f J_' 1: 191-200.
Holliger, K. and Klemperer, S.L., 1989. A comparison of the
N 2-J ft-. Moho interpreted from gravity data and from deep seismic
reflection data in the northern North Sea. Geophys. J., 97:

X COS
**(i-wti
N <-. A.
247-250.
Jarvis, G.T and McKenzie, D., 1980. Sedimentary basin forma-
tion with finite extension rates. Earth Planet. Sei. Lett., 48:
m=l
42-52.
Adding all contributions from the stretched litho- Karlsson, W, 1986. The Snorre, Statfjord and Gullfaks oilfields
sphere (eqn. 5) and that of the remaining thickness and the habitat of hydrocarbons on the Tampen Spur, offshore
to the base of the original lithosphere (eqn. 7) and Norway. In: A.M. Spencer (Editor), Habitat of Hydrocarbons
simplifying gives the final expression for the &th on the Norwegian Continental Shelf. Graham and Trotman,
London, pp. 181-197.
Fourier coefficient, &*:
Klemperer, S.L., 1988. Crustal thinning and nature of extension
in the northern North Sea from deep seismic reflection
N N
1 / 1 1
k-
profiling. Tectonics, 7: 803-821.
bk
-(-_0sin/:^1--X:- Lippard, S. and Liu, G., 1992. Tectonic modelling of the
northern North Sea using program RIFT. In: R.M. Larsen, H.
(A-8) Brekke, B.T Larsen and E. Talleraas (Editors), Structural and
Tectonic Modelling and its Application to Petroleum Geology.
References Norwegian Petroleum Society (NPF), Special Publication 1.
Elsevier, Amsterdam, pp. 43-54 (this volume).
Badley, M.E., Egeberg, T. and Nipen, O., 1984. Development McClay, K.R., Norton, M.G., Coney, P. and Davis, G.H., 1986.
of rift basins illustrated by the structural evolution of the Collapse of the Caledonian orogen and the Old Red Sand-
Oseberg feature, Block 30/6, offshore Norway. J. Geol. Soc. stone. Nature, 323: 147-149.
London, 141: 639-649. McKenzie, D., 1978. Some remarks on the developments of
Badley, M.E., Price, J.D., Rambech Dahl, C. and Agdestein, T., sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32.
1988. The structural evolution of the northern Viking Graben Royden, L. and Keen, C.E., 1980. Rifting process and thermal
and its bearing upon extensional modes of basin formation. J. evolution of the continental margin of Eastern Canada deter-
Geol. Soc. London, 145: 455-472. mined from subsidence curves. Earth Planet. Sei. Lett., 51:
Barr, D., 1985. Three dimensional restoration of normal faults 343-361.
in the Inner Moray Firth: implications for extensional basin Rowley, D.B. and Sahagian, D., 1986. Depth-dependent stretch-
development. Earth Planet. Sei. Lett., 75: 191-203. ing: a different approach. Geology, 14: 32-35.
Barton, P. and Wood, R., 1984. Tectonic evolution of the North Sclater, J.G. and Christie, P.A.F, 1980. Continental stretching:
Sea basin: crustal stretching and subsidence. Geophys. J. R. an exploration of the post-mid-Cretaceous subsidence of the
Astron. Soc, 79: 987-1022. North Sea basin. J. Geophys. Res., 85: 3711-3739.
Beach, A., Bird, T. and Gibbs, A.D., 1987. Extensional tectonics Sclater, J.G., Royden, L., Horvath, F., Burchfiel, B.C., Semken, S.
and crustal structure: deep seismic reflection data from the and Stegena, L., 1980. The formation of the intra-Carpathian
northern North Sea Viking Graben. In: M.P. Coward, J.F. basins as determined from subsidence data. Earth Planet. Sei.
Dewey and PL. Hancock (Editors), Continental Extensional Lett., 51: 139-162.
Tectonics. Geol. Soc. London, Spec. Publ., 28: 467-476. Tankard, A.J. and Welsink, H J . , 1987. Extensional tectonics and
Davies, E.J. and Watts, T.R., 1977. The Murchison oil field. stratigraphy of Hibernia oil field, Grand Banks, Newfound-
Norwegian Petroleum Society (NPF), Mesozoic Northern land. Am. Assoc. Pet. Geol., Bull., 71: 1210-1232.
North Sea Symposium, MNNSS/15: 1-24. Vially, R., 1988. La subsidence dans le Viking Graben (mer du
De'Ath, N.G. and Schuyleman, S.F., 1981. The Geology of the nord septentrionale). Rev. Inst. Fr. Petrole, 43: 629-646.
Magnus oil field. In: L.V Illing and G.D. Hobson (Editors), Wernicke, B., 1984. Uniform-sense normal simple shear of the
Petroleum Geology of the Continental Shelf of Northwest continental lithosphere. Can. J. Earth Sei., 22: 108-125.
Europe, Institute of Petroleum, London, pp. 342-351. White, N., 1988. Nature of lithospheric extension in the North
de Charpal, O., Montadert, L., Guennoc, P. and Roberts, D.G., Sea. Geology, 17: 111-114.
1978. Rifting, crustal attenuation and subsidence in the bay White, N. and McKenzie, D., 1988. Formation of "steer's head"
of Biscay. Nature, 275: 706-711. geometry of sedimentary basins by differential stretching of
Fowler, S. and McKenzie, D., 1989. Gravity studies of the the crust and mantle. Geology, 16: 250-253.
Rockall and Exmouth Plateaux using SEASAT altimetry. Ziegler, P.A., 1982. Geological Atlas of Western and Central
Basin Res., 2: 27-34. Europe. Shell/Elsevier, Amsterdam.

N.E. ODLING Continental Shelf and Petroleum Technology Research Institute (IKU), NHkonMagnusson'sgt. IB, P.O. Box 1883, Jarlesletta,
7001 Trondheim
Present address: Bergen Environmental Centre, IBM, Thorm0hlensgt. 55, Bergen, N-5008, Norway
 469

Structural styles in thrust belts developed through rift

basins: a view from the western Alps
A.I. Welbon and R.W.H. Butler

In the external French Alps Mesozoic passive margin development on the flanks of the Tethyan Ocean resulted in repeated
phases of extension with the formation of faults and associated stratigraphic variations. Extensional faults are found with great
lateral extent throughout the French subalpine chains, often roughly parallel but with antithetic orientations and a dip either
towards the foreland or hinterland. Oligocene and Miocene shortening in the area resulted in fold and thrust geometries which
show a strong orientational relationship to the original normal faults. These early structures and the related stratigraphy also
control the propensity for folding as opposed to simple thrust breaks.
We review published structural sections from the western Alps, together with unpublished data from the French-Swiss border
southwards to illustrate the structural style of these stacked rift basins. This recognition of the interaction of basin geometry and
thrusts is important in the modelling of the thrust belt structures and their geometric evolution.
The recognition of the structural styles in this tectonic setting presents an opportunity to improve our understanding of
structural geometry of hydrocarbon traps for thrust belts developed across rift basins. For example, over-ridden fault blocks,
normal faults reactivated as thrusts and folding against normal faults can provide important new exploration plays, but the
complexities of structural styles may make these elusive targets.

Introduction Recently, the effect of variations in basin geometry

on the development of thrust belts has been a focus
Research into the foreland fold and thrust belt of of interest, not just in foreland areas, but also in
the Rocky Mountains has led to the acceptance of the mechanisms of crustal stacking at depth during
certain geometrical styles in thrust belts (Dahlstrom, collision (Jackson, 1980). The relationship between
1970; Boyer, 1986; Mitra, 1986). These have been basin geometry and thrust belt style is of importance
based upon the recognition of a series of duplexes if we are to be able to model thrust belt structures
and thrust structures which deformed a layer-cake and understand their geometrical evolution.
or wedge-shaped stratigraphic template (for more In the European Alps, the stratigraphic template
examples see Boyer and Elliott, 1982; Suppe, 1983). is known to be complicated by normal faults and
Similar geometries are also present in other thrust stratigraphic variations related to the Mesozoic pas-
belts such as the Appalachians and Scandinavian sive margin of the Tethyan Ocean, (Tfrcart, 1984;
Caledonides (see Woodward and Gray, 1985; Morley, Lemoine et al, 1986; Tricart and Lemoine, 1986;
1986; Tbwnsend et al., 1986). Lemoine and TYmpy, 1987; Lemoine and Gracian-
However, in many parts of some thrust belts, a sky, 1988). The influence of this Alpine passive
layer-cake or wedge-shaped stratigraphic template margin configuration on compressional deformation
was not the pre-existing basin geometry through has been known for a long time. The presence of
which thrusting occurred. Often thrust belts have basement highs and lows had a control on nappe
developed within areas of recognised passive mar- geometry, with bounding thrusts often defined at
gin stratigraphy, consisting of pre-rift basement, tilt- the margins of the pre-existing blocks and basins
blocks and a clastic syn-rift cover, and an overlying (Argand, 1916). On a smaller scale, within these
post-rift sequence. Alternatively, continental margin major basins, pre-thrusting normal faults periodically
development could have related to a complicated controlled cover sequence development. These faults
overprinting of rifting events. In both cases thrusting and related stratigraphic variations also had an in-
took place through complex stratigraphy and normal fluence on thrust geometry (Welbon, 1988a; Butler,
faults and resulted in the formation of complicated 1989a). In this paper, we review geometry of the
thrust geometries (McMechan and Price, 1982; Bally, rift basins of the Tethyan margin in the foreland
1984; Dewey et al., 1986; McClay et al., 1989). of the French Alps and ideas for the nature of in-

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 469-479. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
470 A.L Welbon andR.W.H. Butler

teractions between pre-existing basin geometry and

Alpine compressional deformation. Much data on
this topic is dispersed in Alpine literature or in the-
ses, so here we collate it together with new data to
provide an overview of the nature of the structural
style in this thrust belt.
Finally, we go on to look at the implications of
this style of tectonics for the geometry of structural
hydrocarbon traps, and discuss the possibilities of ex-
trapolation of these styles to hydrocarbon provinces
such as the Rockies (Schmidt and Perry, 1988).

The French Alps

Geological setting
In the external French Alps repeated phases of 1 - Balme 2 - Sales/Pointe de Plate
extension related to Mesozoic passive margin de- Fig. 1. A geological map of the External French Alps showing
velopment of the Tethyan Ocean were followed by the location of the subalpine chains and basement massifs
limited Eocene then Oligo-Miocene shortening lead- containing the normal and thrust faults described in the text.
ing to the formation of the thrust belt (TYmpy, 1973;
Lemoine et al., 1986; Lemoine and TYmpy, 1987).
Excellent exposure makes the western Alps an ideal tivation taking place, whilst others were only active
locality to document structural styles in a thrust belt during one extensional episode. Normal faults may
and their control by inherited basin geometry. extend from Tertiary cover rocks (Welbon, 1988a, b)
Many review articles exist on the stratigraphy of as far down as the Hercynian crystalline basement in
the western Alps and its tectonic framework (e.g., the more central Alps (Lemoine et al., 1986; Gillcrist
Lemoine et al., 1986; TYicart and Lemoine, 1986). et al., 1987). During shortening the variation in basin
General agreement exists upon the main character- geometry resulted in the formation of many com-
istics in the region. Hercynian crystalline basement plicated thrust structures (Welbon, 1988a; Butler,
is overlain by TMassic pre-rift sandstones, shales and 1989a).
evaporites. The advent of Tethyan margin exten-
sion resulted in Jurassic argillite deposition super- Examples
seded by development of a major limestone unit, In Haute Savoie (Fig. 1), in addition to thrust
the Tithonian. More uniform regional subsidence in geometries related to thrust development in layer-
the Cretaceous began with deposition of interbed- cake cover sequences, several spectacular examples
ded sandy limestones and marls followed by car- of normal fault/thrust fault interactions are ex-
bonate platform development (the Urgonian lime- posed. Often associated with these outcrops is a
stone). Tertiary sedimentation began with syn-rift pronounced stratigraphic variation between hanging-
deposition of limestones and marls and became pro- wall and footwall rocks across thrusts. Examples of
gressively flysch-like in nature reflecting the onset these structures can be found in Welbon (1988a, b)
of compressional tectonics in the developing Alpine and in Butler (1989a, 1992).
orogen. In the Sales-Pointe de Plate area (Fig. 1) a thrust
The structural style in the external French Alpine fault exhibits a clear variation in stratigraphy from
Thrust Belt varies along strike from a series of fold hanging-wall to footwall, although the sequences are
dominated nappes at the French-Swiss border (the of the same age (Fig. 2) (Welbon, 1988a, b). Kine-
Helvetic Nappes, e.g., Ramsay et al., 1983) to a matic indicators (small scale folds, sheared cleavage,
more thrust dominated sequence to the southwest striae) show the latest movements on the fault to
(Doudoux et al., 1982; Butler, 1989a, b). A consistent have been thrust sense. The fault at Sales is well
feature along the whole External French Alps, and exposed from the level of the Early Cretaceous (Ur-
the Swiss Alps to the north, is the presence of lat- gonian) limestone facies up to Oligocene sandstones.
erally continuous intrabasinal faults (Lemoine et al., The Urgonian facies are similar on either side of
1986). These were active at various times from TYias- the fault, but the footwall Late Cretaceous units are
sic (Froitzheim, 1988) through Jurassic (Lemoine et 150 m thicker than rocks of the same age in the
al., 1986) to Oligocene (Butler, 1989a). Some normal hanging-wall. Similarly, the footwall Tertiary units
faults show long-lived tectonic histories, with reac- exhibit a greater thickness, 70 m as compared to 30
Structural styles in thrust belts developed through rift basins: a view from the western Alps 471

SE NW

NW SE
Fig. 2. A sketch of the Sales fault (after Welbon 1988a, b).
An increased stratigraphic thickness of footwall rocks relative to
those in the hanging-wall is evident. Footwall deformation in the
Late Cretaceous limestones consists of folds and thrusts which
lose displacement away from the fault.

m in the hanging-wall. Intense deformation in the

form of folds and thrusts characterises the footwall,
especially the Late Cretaceous units, but restoration
of these structures reveals that a considerable strati-
graphic thickness difference remains between the
rocks on either side of the fault. Shortening increases
towards the top of the Late Cretaceous units, which,
when restored, indicates that the fault was originally
near vertical or dipping in the opposite direction (i.e.,
NW) to the present SE dip. This is supported by the
presence of a NW-dipping Tbrtiary syn-sedimentary 300 m KEY
normal fault carried in the hanging-wall which was HI Tertiary
presumably synthetic to the original major normal [2 Late Cretaceous
fault. Slip surfaces between the Tertiary and Late IWF] Urgonian
Cretaceous rocks in the footwall (Welbon, 1988b)
Fig. 3. (A) A line drawing of a photograph of the Pointe de Plato
would also have accommodated increased shortening Fault which is highly oblique to its strike. This is an along-strike
in the upper stratigraphic units. From this data, it expression of the Sales Fault 3 km SW of the exposures in the
is concluded that the fault had an early history of Sales valley. A similar stratigraphic variation to that shown in
Fig. 2 occurs. The fault structure has a splay in its footwall
extension, prior to and during Eocene deposition, which cuts the Urgonian rocks. Key: densely dotted area =
and was then overturned and reactivated as a thrust Tertiary; blank area = Late Cretaceous; sparsely dotted area
during Oligo-Miocene shortening. = Urgonian limestone; dashed areas = Hauterivian-Berriasian.
Along strike at the Pointe de Plato, the fault (B) An evolutionary diagram of the Pointe de Plato Fault.
(a) Potential fault trajectory, (b) A fault cuts through the
has a similar variation in stratigraphic thickness as Urgonian rocks and overturns the normal fault carried in its
at Sales, but the fault has a splay in its footwall hanging-wall, (c) The overturned part of the normal fault is
that carries Urgonian limestone (Fig. 3A). The fault reactivated as a thrust.
has a thrust-sense offset and is interpreted as a
reactivated normal fault (Fig. 3B). Only the top part
of the normal fault has been used by a thrust after to that of the later thrust faults. Two major NW
it was overturned during folding. In this case the verging thrusts cut the graben (Fig. 4A, B). The
earlier folding is related to the thrust which carries upper one cuts through and overturns a normal fault
Urgonian facies lithologies in the footwall to the dipping away from it and ramps ahead of the SE-
reactivated fault. dipping normal fault. The lower NW verging thrust
At Balme, in the Arve valley (Fig. 1, Fig. 4A) the cuts through both sets of faults, carrying the graben
structures are more complicated than those at Sales. and is exposed 400 m NW of the graben. Again,
Here, a graben structure with similar thickening to like the Sales example where the early fault dipped
that found at Sales (i.e., within Late Cretaceous away from the advancing thrust, folding has been an
and Tertiary rocks) is exposed. The graben has two important part of the shortening process, resulting
oppositely dipping faults that have a strike similar here in partial overturning of the early fault.
 A.1. Welbon and R. W.U. Butler

In addition to the formation of major displace-

ment thrusts, minor thrusts (decollement) develop
within thrust sheets at contacts between less compe-
tent, well bedded, Late Cretaceous limestones and
the more competent, massive Urgonian limestones
(Fig. 4A, B). The origin of the internal deformation
of thrust sheets is the ramp areas of the overlying
major thrusts, and transference of displacement is
evidenced by footwall deformation linking the over-
lying thrust plane with the decollement (Welbon,
1988b). The minor thrusts pin at normal faults and
shortening above the decollement is accommodated
by folding against the normal fault (buttressing).
These structures are folded by hanging-wall anti-
clines developed above underlying major thrusts and
are, therefore, formed prior to the formation of the
major thrusts.
Further south in the Chartreuse (Fig. 1) strati-
graphic thickness variations suggest the presence of
a near vertical, SE dipping, normal fault cut by the
Corbel Thrust Zone (Fig. 5). The Thrust Zone (But-
ler, 1992) consists of a complicated zone of branching
thrusts developed up dip from the normal fault car-
ried in the hanging-wall. Folding above and below
the thrust is minor and, therefore, a simple offset of
the normal fault is observed. This is in contrast to the
geometries in the Balme and Sales localities where
folding within thrust sheets is sometimes sufficient to
overturn early faults.

Tertiary rocks are thickened, (b) A minor thrust develops at

the Urgonian-Late Cretaceous contact and pins at the NW
dipping normal fault, causing folding above the thrust, (c, d)
The upper major thrust offsets the NW dipping normal fault and
KEY the upper part of it is overturned in the hanging-wall anticline
Small scale structures developed above the thrust. Note the upper thrust ramps ahead
&7ZC Folding \^\ Late Cretaceous of the SE dipping normal fault and links into less competent
^rr Normal Faulting [JijJj^Tj Urgonian
units (thinly bedded Late Cretaceous limestones) above the
competent Urgonian facies massive limestones, (e) A minor
[|3 Tertiary ] Barremian - Berriasian thrust forms at the Urgonian-Late Cretaceous boundary, pins at
Fig. 4. (A) A diagram and (B) evolutionary model of the the SE dipping normal fault and continued shortening results
fault geometries in Balme (after Welbon 1988a, b). In B: A in buttressing (folding) against the normal fault, (f) The lower
graben with two oppositely dipping faults is cut by two NW major thrust transports both normal faults and the upper thrust
verging thrusts. The graben thickens Late Cretaceous lime- in its hanging-wall. It forms a ramp approximately 400 m NW of
stones and, in the case of the NW dipping normal fault, the graben.
 Structural styles in thrust belts developed through rift basins: a view from the western Alps 473

Chartreuse Hills
WNW ESE
Corbel Thrust Zone

Bas Dauphine St. Laurent Basin

Fig. 5. A section through the Chartreuse region showing a SE dipping normal fault offset by the Corbel Thrust Zone (after Butler, 1992).

ESE
Bas Dauphine Vercors Plateau

Fig. 6. A section through the Vercors (after Butler, 1989a). Note the backthrusting of cover over the Dome de la Mure massif and the
buried NW dipping normal fault at the front of the massif. Within the cover at Royans, near the thrust front, thrusts cut through a pair
of WNW dipping normal faults. At the thrust front an ESE dipping normal fault in the basement is over-ridden by a thrust.

In the Vercors region SW of the Chartreuse thrust front, an ESE dipping normal fault in the
(Fig. 1) there are more examples of the original basement is overridden by a thrust fault ramping
rift geometry in this part of the thrust belt (Fig. ahead of it (Fig. 6)
6). In the most internal part of the section at the At the front of the Jura mountains (Fig. 7) a
Dome de la Mure basement massif, a WNW dip- series of NW dipping normal faults bounding the
ping early normal fault is predicted from thickness Bresse graben are truncated by WNW verging thrust
variations in the pre-Aalenian cover rocks (Butler, structures (Butler, 1989b). The dip of these normal
1989a). This early normal fault is predicted to cut faults is antithetic to the thrust faults, and, like the
basement rocks and thicken Liassic sediments. The Sales example, illustrates that when normal faults dip
footwall to this fault, the Dome de la Mure basement away from an advancing thrust, they often present no
massif, was emplaced as a thrust sheet beneath the barrier to thrusting and are simply cut through.
cover rocks resulting in the formation of a backthrust In summary, from the geometry of the structures
between the basement and cover. This backthrust in the western Alps a progressive change in response
carries Jurassic and younger rocks across the thin to shortening of the early faults and their related
cover sequences on the crest of the Dome de la stratigraphy is suggested. More folding occurs in the
Mure massif. northern part of the French Alps at the Swiss border
More externally in the Vercors, at Royans, thrusts than to the southwest towards the Vercors. It is
cut through a pair of WNW dipping normal faults. also evident that a primary influence on thrusting is
Again, in a similar fashion to the structural styles the original orientation of early faults. Early faults
in the Chartreuse, there is little overturning of the dipping towards the thrust tend to cause ramping
faults, they are simply cut through. Finally, at the of the thrust ahead of the early fault, those dipping
474 A.I. Welbon and R. W.U. Butler

Bugey Hills Annecy Basin

^ 0 o*-L. Mid Miocene

Bresse Graben ^ Urgonian
*- Kim-Oxfordian
* Trias-Lias

Fig. 7. A section through the Jura Mountains after Butler (1989b). Note WNW verging thrusts cut from the Jura mountains through
the NW dipping normal faults bounding the Bresse graben.

away from the thrust are usually overturned and Intermediate geometries are possible between
partially reactivated, or cut through. The variation these end members dependent on the amount of
in geometry of the stratigraphic units adjacent to shortening and, like all rocks, other factors deter-
the fault and the differences in competence of these mining strain will also include lithology, pressure and
stratigraphic units also plays an important role in temperature conditions and strain rate (e.g., Knipe,
determining thrust geometry. 1985).
Although we have illustrated many examples of However, from comparing the Sales and Balme ex-
early normal fault-late thrust fault interactions, it amples where similar lithological units are involved
is important to note that many thrusts are demon- in thrusting, but with different original geometries
strably unrelated to normal faults or stratigraphic due to different extensional fault orientations and
variation (Doudoux et al., 1982; Ramsay et al., 1983; throws, it is evident that the degree of folding as
Butler, 1989b). Thrust spacing is smaller than the opposed to simple thrust breaks and any style of
spacing between significant basin heterogeneities. thrusting is greatly dependent on the original ori-
Thus, many thrusts within the thrust belt propagate entation of the early fault and the geometry of the
through parts of the stratigraphic sequence where stratigraphic units developed around the fault. Often
there was no apparent influence from pre-existing in the Alpine examples the fault itself is not a barrier
structures. The resulting geometries are those such to thrusting, but the juxtaposition by the fault of me-
as thrusts, duplexes, with fault-bend and fault prop- chanically strong lithologies against weak lithologies
agation folds (Boyer and Elliott, 1982; Boyer, 1986; is common (e.g., Balme) preventing the continuation
Mitra, 1986; Welbon, 1988b). of any decollement developed in the weak layer and
thereby resulting in folding or ramping of thrusts.
Discussion Normal faults dipping towards the thrust are likely
to cause ramping of the thrust ahead of the early
The examples of Alpine thrusts discussed in this fault. This is likely to be the case where intrabasinal
paper consistently show a series of geometries be- (normal) faults are involved, as in their hanging-
tween several end members. These are given below. walls they tend to thicken younger, less competent
(1) Thrusts cut through early faults (e.g., Balme, stratigraphic units. Thus, in the footwall to the early
Chartreuse, Vercors). fault a junction between competent and less com-
(2) Thrusts ramp ahead of normal faults dipping petent units will be at a higher level and will act
towards the advancing thrusts (e.g., Balme, Jura, as a stress riser, the thrust will be localised here
Vercors). and folding will be limited (Wiltschko and Eastman,
(3) Thrusts cut through and overturn normal 1983; Schedl and Wiltschko, 1987). Early faults dip-
faults dipping away from the advancing thrust (e.g., ping away from an advancing thrust with competent,
Balme). mechanically stronger units in the footwall produce a
(4) Thrusts overturn and reactivate the top part stress shadow across potential decollement surfaces
of normal faults originally dipping away from the in the hanging-wall. Hence these early faults are a
advancing thrust (e.g., Sales, Pointe de Plate). barrier to thrusting and are much more likely to be
(5) Thrusts pin at the early fault and shortening overturned before a thrust will propagate forwards,
continues above the thrust in the form of folding the thrust often using the overturned fault as an easy
(e.g., Balme). slip surface.
Structural styles in thrust belts developed through rift basins: a view from the western Alps 475

Fig. 8. A catalogue of possible thrust geometries found in areas where pre-existing faults and stratigraphic variations exist. These
structures may form hydrocarbon traps. In the following description "basement" (dotted area on the figure) refers to a competent
package of lithologies and "cover" (blank area on the figure) refers to a less competent package of lithologies. (a) Upper diagram.
Structures formed when a pre-existing fault dips towards an advancing thrust. (1) Simple reactivation of the early fault. (2) The
advancing thrust ramps ahead of the early fault and links into the cover above the competent basement. (3) The thrust forms along a
basement-cover contact and pins at the early fault. Continued shortening results in folding above the detachment. (4) The early fault is
cut through by the thrust, (b) Lower diagram. Structures formed when a pre-existing fault dips away from an advancing thrust. (1) The
thrust forms a detachment between basement and cover and continued shortening results in the formation of folds. (2) The early fault
is cut through by the thrust. (3) The early fault is cut through by the thrust and is overturned. (4) Folding by a results in overturning
into an orientation appropriate for reactivation of the top part of the early fault. N.B., where folding takes place against a fault (a,
example 3 and b, example 1), an assumption is made that the fault or the rock types across the fault exhibit relative competence
compared to the folded strata.

On the basis of the examples discussed, we have tribution of mechanically different stratigraphic units
made a catalogue of possible thrust geometries de- shown in the figure could be formed by oblique slip
veloped through areas of inherited structures and or dip slip faults.
stratigraphic variations (Fig. 8). The figure shows In a hydrocarbon province the resulting geometry
possible structures that can form from shortening after shortening in a thrust belt may provide the set-
across the two most common types of inherited ting for structural traps for hydrocarbons. Although
structures found in the Alps. The first set (Fig. 8a) Fig. 8 shows only two-dimensional representations
is where an early fault dips towards the advancing of the structures possible in a thrust belt through
thrust and the second set (Fig. 8b) is where a pre- interaction with inherited structures, the figure can
existing fault dips away from the advancing thrust. be used to discuss the potential geometries of hydro-
Note the original early fault geometry and the dis- carbon traps assuming structural closure in the third
476 A.L Welbon and R. W.U. Butler

dimension, at right angles to the page. Below is an structural traps in the over-ridden fault block. Alter-
outline of the structural styles of the structures and a native 3 illustrates a situation similar to that in 2,
brief outline of the possible hydrocarbon traps. but where folding is more dominant and as a result
Figure 8a shows the alternative structures formed the early fault carried in the hanging-wall to the
in the case of an early fault dipping towards oncom- thrust is overturned. Folding may also occur in the
ing thrusts. Alternative 1 shows simple reactivation footwall to the thrust. In addition to traps formed in
of a normal fault which may produce hanging-wall the over-ridden block, traps may form in the folded
and footwall traps. The hanging-wall traps consist of areas of the overturned normal fault. Alternative 4
anticlinal structures, whereas the footwall traps can shows overturning of the early fault being sufficient
be produced by combined footwall syncline forma- to allow reactivation of the top part of the normal
tion and sealing along the fault. Alternative 2 forms fault as a thrust. Similar traps to those in alternative
when shortening takes place above a more compe- 3 are possible.
tent unit which has been previously downfaulted. The In addition to the possible traps indicated in Fig.
thrust ramps ahead of the normal fault and links into 8, early faults may not interact with later thrusts and
less competent units in the footwall. Over-ridden can be passively tilted or folded within thrust sheets
fault blocks then become potential traps. Alterna- to form hydrocarbon traps. This is particularly likely
tive 3 is similar to the previous model, but illustrates if the early faults are transported passively in one
a situation where a thrust forms at a level between thrust sheet and are then folded by a culmination
two units of different competence and then pins at related to the emplacement of a lower thrust sheet.
the early fault or more competent footwall rocks. Reactivation of these faults may then occur if they
Continued shortening results in fold formation above are in an appropriate orientation, and this can con-
the detachment. Depending on the scale of these ceivably result in reactivation of an early fault with
structures, a potential hydrocarbon trap may form, its original sense of motion. The Mont Charvin area
but examples documented at Balme in the Alps are of Haute Savoie (for stratigraphic setting see Villars
small and show a very complex geometry unsuit- et al., 1988) may be an area where this type of re-
able for structural traps. Alternative 4 shows cutting activation has taken place (Fig. 9). Here there is an
through of the early fault and the carrying of the early Tertiary normal fault with a throw of more than
fault passively in the hanging-wall to the thrust. In 500 m which has been overturned by uplift of the
many external fold and thrust belts this is likely to underlying external Belledonne massif. Kinematic in-
form where the lithologies on either side of the fault dicators on part of the normal fault, including folds,
and the fault itself are mechanically weak. This can striae and an extensional crenulation cleavage in
create two possible traps in the over-ridden block on the hanging-wall, indicate that it was reactivated as a
either side of the early fault. backthrust with the same type of offset as the original
Figure 8b illustrates the alternatives in a scenario normal fault, although displacement was small.
where an early fault dips away from the advancing It is important to note, especially when consid-
thrust. Again the geometry of the early fault and ering the hydrocarbon potential of these structures,
the distribution of the mechanically different strati- that although the initial and final geometries of the
graphic units can be produced by a normal fault structures described here are known, it is not al-
or oblique slip fault. Alternative 1 shows a thrust ways clear what the sequence of formation of the
which forms a decollement between competent and intermediate structures was. Folds may have formed
less competent units and pins at the early fault. prior to thrusts or vice versa. In many cases it is
Continued shortening causes folding above the de- impossible to determine from geometry alone what
tachment. The early fault (or the hanging-wall rocks the sequence was, although overprinting of varying
in this case) are mechanically strong and act as a kinematic indicators on a meso- or microscale may
buttress to the footwall shortening. This occurred at provide the answer. A particular intermediate struc-
the SE end of the Balme example, in the original tural style may result in the loss of hydrocarbons
footwall to the early normal fault (now overturned before a trap is created.
by a major thrust) at the level of the Late Creta- Published data from other thrust belts has illus-
ceous limestones. Significant hydrocarbon traps are trated some thrust belt styles similar to those in the
unlikely to form if, as in Fig. 8a alternative 4, the western Alps. In the Rocky Mountains thrust belt
structure is complex and small scale. Alternative 2 thrust structures and hydrocarbon traps developed
shows the most common structural style encountered within duplexes and imbricate thrust stacks where
in the study area of those in Fig. 8b. This is where the shortening took place through a layer-cake or wedge-
early fault is cut through by the thrust and carried shaped stratigraphic template (Boyer and Elliott,
passively. Again there is the possibility of creating 1982; Mitra, 1986). In other regions of the Rockies,
Structural styles in thrust belts developed through rift basins: a view from the western Alps 11

ESE

-7N r? ^* -,.7^ 7/ 'T/ T - '

| ^ , , _ ^. , ,-'

Fig. 10. A schematic history of the Birmingham Anticlinorium

after Schedl and Wiltschko (1987) based on Thomas (1982).
(a, b) Initiation of the basement normal fault in the Eocambrian,
followed by drape folding after Mid-Cambrian, Mid-Silurian and
Early Mississipian movement, (c, d) Initiation of a thrust ramp
in the early Post-Pennsylvanian followed by transport forward
via later faults. The step in the basement is interpreted as acting
as a stress riser resulting in the ramping of the advancing thrust
ahead of the basement normal fault.

found to ramp up above basement topography (e.g.,

Thomas, 1982, 1983) (Fig. 10). Again the thrusts
link less competent units on either side of the nor-
mal faults/basement topography, the normal fault
KEY
footwall rocks having acted as a stress riser as pre-
g v j Tertiary EZ1 Basement
dicted from finite element modelling of Schedl and
Jurassic to
Late Cretaceous
Trajectory
of next fault
Wiltschko (1987) (Fig. 11).
Fig. 9. A diagram and model of the Mt. Charvin area. The In the Bass Straits/Gippsland basin of Australia
intrabasinal extension is well documented and results in thick- the reactivation of deeply buried extensional blocks
ening of the Late Cretaceous and Tertiary lithologies (Villars et
in a basin is common (Etheridge, 1986). Many styles
al., 1988). (a) Original pre-Thrusting geometry, with extensional
faults within the cover and at the basement cover contact, (b-d) of reverse and strike-slip reactivation are docu-
Tilting of this fault by the uplift of the Belledonne massif results mented, largely depending on the original orien-
in it being in an appropriate orientation for reactivation as a tation of the early fault. This type of reactivation
backthrust with the same type of offset as the original normal
fault.
may be the case at a passive margin during the early
stages of collision or where a basin has been weakly
affected by a compressional orogenic event (see also
normal faults and stratigraphic variations are more Ziegler, 1983, for examples of North Sea Alpine
common and have resulted in a control of structural inversion).
style in the thrust belt. Normal fault reactivation as
thrusts, ramping ahead of structures at basin margins Conclusions
or buttressing (folding) against normal fault footwalls
are known to have occurred in the foreland of the Accepted geometrical styles of thrusting from fore-
Rockies and much literature has been published on land areas with layer-cake or near so basin geometry
the nature of this style of tectonics (see Benvenuto include classic thrust duplexes and related folding.
and Price, 1979; Schmidt and Perry, 1988). The examples of thrust belt geometry given above
Appalachian thrust structures within cover rocks, show that inherited tectonic structures and strati-
where associated with early faults, are consistently graphic variations are a common feature of the
478 A.L Welbon andR.W.H. Butler

ened sedimentary basin with non-layer-cake stratig-

raphy is necessary for a successful assessment of
potential hydrocarbon traps during exploration. It
may be insufficient to rely on techniques such as
the projection of surface data to depth, (e.g., Suppe,
1983), with an assumption of lateral continuity of
rock type or to make structural predictions with-
out reference to stratigraphy and the influence rift
structures have had on its development.

Acknowledgements

A.W. acknowledges a Natural Environment Re-

/0 search Council (NERC) Ph.D. studentship No. GT4/
85/GS/44. A.W. also acknowledges a Royal Society
l-LLUM-14?. mSMii,
L /-/
European Science Exchange Programme Postdoc-
toral Fellowship and thanks Professor Arild An-
I///J Region of failure
dresen for providing research facilities at the De-
.=* Potential fault
partment of Geology, University of Oslo. Fieldwork
(R.W.H.B.) in the French Alps was funded by a Natu-
'''. Basement
ral Environment Research Council (U.K.) grant No.
Fig. 11. A diagram of failure regions and potential faults
for various geometries of basement faults and basement cover (GR3/6172). We thank Rob Knipe, Daniel Mller
contacts, after Schedl and Wiltschko (1987). The potential and Francois Villars for discussions on Alpine Geol-
geometry depends upon the basement topography and the
degree of basement-cover coupling.
ogy-
References

western Alps and other thrust belts and that their Argand, E., 1916. Sur l'arc des Alpes Occidentales. Eclogae
Geol. Helv., 14: 145-191.
presence has had an influence on thrust evolution. Bally, A.W., 1984. Tectonogenese et sismique reflexion. Bull.
The variation of early fault geometry and defor- Soc. Geol. Fr., 26: 279-285.
mation conditions within a thrust belt means that Benvenuto, G.L. and Price, R.A., 1979. Structural evolution of
later thrust fault geometries can show a wide vari- the Hosmer Thrust Sheet, southeastern British Colombia.
ety of forms. More complex structures develop with Bull. Can. Pet. Geol., 27: 360-394.
Boyer, S.E., 1986. Styles of folding within thrust sheets: examples
progressive shortening. Depending on early fault ori- from the Appalachian and Rocky Mountains of the U.S.A.
entation, reactivation, ramping of thrusts ahead of and Canada. J. Struct. Geol., 8: 325-339.
buried faults and folding against the early fault (but- Boyer, S.E. and Elliott, D., 1982. Thrust systems. Am. Assoc.
tressing) are common. Alternatively, early faults can Pet. Geol., Bull., 66: 1196-1230.
simply be cut through by later faults. The recogni- Butler, R.W.H., 1989a. The influence of pre-existing basin struc-
ture on thrust system evolution in the Western Alps. In: M.A.
tion of these thrust belt styles is important if we Cooper and G.D. Williams (Editors), Inversion Tectonics.
are to model them to understand their geometrical Geol. Soc. London, Spec. Publ., 44: 105-122.
evolution. Butler, R.W.H., 1989b. The geometry of crustal shortening in the
Structural complexity of this type in a hydrocarbon Western Alps. In: A.M.C. Segnor (Editor), Tectonic Evolution
of the Tethyan Region. Proceedings of the NATO Advanced
province can result in structural traps being formed.
Study Institute, C239, pp. 43-76.
The type of hydrocarbon trap, like structural style, Butler, R.W.H., 1992. Structural evolution of the western
will vary greatly depending on the orientation of Chartreuse fold and thrust system, NW French Subalpine
early faults and their related stratigraphic variations Chains. In: K.R. McClay (Editor), Thrust Tectonics. Chap-
relative to thrusts and the level of decollement. The man & Hall, London, pp. 287-298.
Dahlstrom, C.D.A., 1970. Structural geology in the eastern
position of buried normal faults in over-ridden blocks
margin of the Canadian Rocky Mountains. Bull. Can. Pet.
may increase or decrease exploration potential, de- Geol., 18: 332-402.
pending on the nature of the lithologies buried in the Dewey, J.F., Hempton, M.R., Kidd, W.S.F., Saroglu, F. and
early fault beneath the thrust. Analysis employing Segnor, A.M.C., 1986. Shortening of continental lithosphere:
a correlation between surface and seismic data may the neotectonics of Eastern Anatolia a young collision
zone. In: M.P. Coward and A.C. Ries (Editors), Collision
help constrain subsurface geometries.
Tectonics. Geol. Soc. London, Spec. Publ., pp. 3-36.
The consideration of the presence of these types of Doudoux, B., Mercier de Lepinay, B. and Tardy, M., 1982.
compressional structure in any thrust belt or short- Une interpretation nouvelle de la structure des massifs sub
Structural styles in thrust belts developed through rift basins: a view from the western Alps 479

alpins (Alpes occidentales): nappe de charriage oligocenes et Suppe, J., 1983. Geometry and kinematics of fault bend folding.
deformations superposees. C.R. Acad. Sei. Paris, Ser. II, 295: Am. J. Sei., 283: 684-721.
63-68. Thomas, W.A., 1982. Stratigraphy and structure of the Ap-
Etheridge, M.A., 1986. On the reactivation of extensional fault palachian fold and thrust belt in Alabama. In: W A. Thomas
systems. Philos. Trans. R. Soc. London, Ser. A, 317: 179-194. and T. L. Neathery (Editors), Appalachian Thrust Belt in
Froitzheim, N., 1988. Synsedimentary and synorogenic normal Alabama; Tectonics and Sedimentation. 95th Annu. Meet.
faults within a thrust sheet of the eastern Alps (Ortler zone, Geol. Soc. Am. Guidebook for Field Trip, Geology Society of
Graubunden, Switzerland). Eclogae Geol. Helv., 81: 593-610. America, pp. 55-78.
Gillcrist, R., Coward, M.P. and Mungnier, J-L., 1987. Structural Thomas, W.A, 1983. Basement-cover relations in the Ap-
inversion and its controls. Examples from the Alpine Foreland palachian fold and thrust belt. Geol. J., 18: 267-276.
and the French Alps. Geodyn. Acta (Paris), 1: 5-34. Townsend, C , Roberts, D., Rice, A.H.N. and Gayer, R.A,
Jackson, J.A., 1980. Reactivation of basement faults and crustal 1986. The Gaissa Nappe, Finnmark, North Norway: an ex-
shortening in orogenic belts and plate tectonics. Nature, 283: ample of a deeply eroded external imbricate zone within the
343-346. Scandinavian Caledonides. J. Struct. Geol., 8: 431-440.
Knipe, R.J., 1985. Footwall geometry and the rheology of thrust Tricart, P., 1984. From passive margin to continental collision:
sheets. J. Struct. Geol., 7: 1-10. a tectonic scenario for the Western Alps. Am. J. Sei., 284:
Lemoine, M., Bas, T., Arnaud, H., Gidon, M., De Graciansky, 97-120.
P.C., Rudkiewicz, J-L., Megard-Galli, J. and Tricart, P., 1986. Tricart, P. and Lemoine, M., 1986. From faulted blocks to
The continental margin of the Mesozoic Tethys in the Western megamullions and megaboudins: Tethyan heritage in the
Alps. Mar. Pet. Geol., 3: 179-199. structure of the western Tethys. Tectonics, 5: 267-276.
Lemoine, M. and Graciansky, P.C., 1988. Histoire d'une marge Irmpy, R., 1973. The timing of orogenic events in the Central
continental passive: les Alpes occidentales au Mesozoique. Alps. In: K.A. DeJong and R. Schlte (Editors), Gravity
Introduction. Bull. Soc. G6ol. Fr., 8: 597-600. Tectonics. Wiley, New York, N.Y.
Lemoine, M. and Irmpy, R., 1987. Pre-oceanic rifting in the Villars, F., Muller, D. and Lateltin, O., 1988. Analyse de la
Alps. Tectonophysics, 133: 305-320. structure du Mont Charvin (Haute Savoie) en termes de
McClay, K.R., Insley, M.W. and Anderton, R., 1989. Inversion of tectonique synsedimentaire paleogene. Consequences pour
the Kechika trough, Northeastern British Columbia, Canada. Linterpretation structurale des chaines subalpines septentri-
In: M.A. Cooper and G.D. Williams (Editors), Inversion onales. C.R. Acad. Sei. Paris, S6r. II, 307: 1087-1090.
Tectonics. Geol. Soc. London, Spec. Publ., 44: 235-257. Welbon, A.I., 1988a. The influence of intrabasinal faults on the
McMechan, M.E. and Price, R. A , 1982. Transverse folding and development of a linked thrust system. Geol. Rundsch., 77:
superposed deformation, Mt. Fisher area, Southern Canadian 11-24.
Rocky Mountain thrust and fold belt. Can. J. Earth Sei., 19: Welbon, AI., 1988b. Deformation styles and localisation of
1011-1024. thrust faults in the external French Alps. Ph.D. Thesis,
Mitra, S., 1986. Duplex structures and imbricate thrust systems: University of Leeds, Leeds (unpublished).
geometry, structural position and hydrocarbon potential. Bull. Wiltschko, D. and Eastman, D., 1983. Role of basement warps
Am. Assoc. Pet. Geol., Bull., 70: 1087-1112. and faults in localising thrust fault ramps. In: R.D. Hatcher,
Morley, C.K., 1986 The Caledonian thrust front and palinspastic H. Williams and I. Zeitz (Editors), Contributions to the
restorations in the southern Norwegain Caledonides. J. Struct. Tectonics and Geophysics of Mountain Chains. Geol. Soc.
Geol., 8: 753-765. Am., Mem., 158.
Ramsay, J.G., Casey, M. and Kligfield, R., 1983. Role of Woodward, N.B. and Gray, D.R., 1985. Southwest Virginia,
shear in the development of the Helvetic fold-thrust belt of Tennessee and Northern Georgia sections. In: N.B. Woodward
Switzerland. Geology, 11: 439-442. (Editor), Valley and Ridge Thrust Belt: Balanced Structural
Schedl, A.H. and Wiltschko, D.V, 1987. Possible effects of Sections, Pennsylvania to Alabama. Univ. Tennessee, Stud.
pre-existing basement topography on thrust fault ramping. J. Geol., 12: 40-53.
Struct. Geol., 9: 1029-1037. Ziegler, P.A., 1983. Inverted basins in the Alpine foreland. In:
Schmidt, C.J. and Perry, J., Jr., 1988. Interaction of the Rocky A. W Bally (Editors), Seismic Expression of Structural Styles
Mountain foreland and the Cordilleran Thrust Belt. Geol. A Picture and Work Atlas. Am. Assoc. Pet. Geol., Stud.
Soc. Am., Mem., 171. Geol. Ser., pp. 3.3.3-3.3.12.

A.I. WELBON Department of Geology, University of Oslo, P.O. Box 1047 Blindern, 0316 Oslo 3, Norway
R.W.H. BUTLER Department of Earth Sciences, The University, Leeds, LS2 9JT, U.K.
Present address: Department of Geological Sciences, Queens University, Union Street, Kingston, Ont. K7L 3N6, Canada
 481

Tertiary decollement thrusting and inversion

structures along Billefjorden and Lomfjorden Fault Zones,
East Central Spitsbergen
P. Haremo and A. Andresen

Structural investigations on East Central Spitsbergen along the Billefjorden (BFZ) and Lomfjorden Fault Zones (LFZ),
indicate that most of the stratigraphic thickness variations within the Mesozoic strata in this area are due to Tertiary
compressional tectonics related to the Eocene transpressive West Spitsbergen Orogeny. No convincing evidence of a distinct
Mesozoic deformational event, as suggested by previous workers, has been found. The Tertiary compressional tectonics
is characterized by a combined thin-skinned/thick-skinned (basement involved) structural style. Two regionally extensive
decollement zones are recognized in the Mesozoic strata; a lower one in the Lower Triassic Sassendalen Group, and an upper
one in the Upper Jurassic/ Lower Cretaceous Janusfjellet Subgroup. Both zones can be traced from the West Spitsbergen
Foldbelt eastwards to Storfjorden. Numerous reverse faults with variable eastward displacement are associated with both zones.
Reverse faults and east-verging folds are particularly frequent as the decollement zones cross BFZ and LFZ. Preliminary
calculations indicate an eastward displacement of minimum 3-4 km, possibly as much as 10 km, of Lower Cretaceous and
younger rocks across BFZ. Along-strike changes in fold and fault geometry and the appearance of thrust faults above the
Festningen Sandstone (Lower Cretaceous) towards the south, suggest the existence of transfer zones or lateral ramps across
which the upper decollement zone climbs up-section eastward as well as southward. The southward climbing of the thrusts is
inferred to be controlled by pre-deformational uplift and southward tilting of Spitsbergen during Upper Cretaceous time. Both
BFZ and LFZ represent reactivated steep, opposite dipping, basement involved reverse faults. This has resulted in an uplifting
(inversion) of the area between BFZ and LFZ. Interference structures along BFZ suggest that thick-skinned faulting preceded
thin-skinned thrusting, whereas the opposite age relationship is inferred for the thick- and thin-skinned deformation observed
along LFZ. Although the thick- and thin-skinned deformations appear to represent two separate deformational events, they are
both considered to be of Eocene age and associated with large scale transpression west of Spitsbergen.

Introduction The Billefjorden (BFZ) and the Lomfjorden Fault

Zones (LFZ) represent two of the most distinct and
Important hydrocarbon source and reservoir rocks extensive tectonic lineaments on Spitsbergen (Fig. 1).
occur in the Upper Palaeozoic and Mesozoic se- Both lineaments have a complex tectonic history,
quences on Spitsbergen. These rocks have been ex- probably going back into the Precambrian, and it
tensively studied and their depositional environment is well documented that BFZ played an important
is relatively well known. Considerably less is known role during the structural and sedimentological evo-
about their deformational history, particularly the lution of Spitsbergen in the Late Palaeozoic (Orvin,
extent and nature of the Tertiary West Spitsbergen 1940; Harland et al., 1974; Kellogg, 1975; Steel and
Orogeny. Another important problem concerning hy- Worsley, 1984). Both fault zones were reactivated
drocarbon accumulations on Spitsbergen, as for the during the Tertiary (De Geer, 1919; Hagermann,
entire western Barents Sea shelf (e.g., Nyland et al. 1925; Orvin, 1940; Livshits, 1965; Parker, 1966; Kel-
and Riis and Fjeldskaar, this volume), is the Ceno- logg, 1975). However, neither the geometry nor the
zoic uplift and erosion which on Spitsbergen have kinematics of these Tertiary structures nor the way
removed as much as 1.7 km of Late Tertiary and in which they relate to the narrow belt of intense
younger strata (Manum and TYondsen, 1978). In such deformation along western Spitsbergen have been
a setting, an understanding of the tectonic history of thoroughly described. In this paper the complex de-
major fault zones and how related structural traps formation structures found along BFZ and LFZ as a
are generated, is of great importance in detecting result of interacting thick- and thin-skinned Tertiary
possible hydrocarbon accumulations in this region of deformation are discussed.
the Barents Sea shelf.

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 481-494. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
482 P. Haremo and A. Andresen

) Festningen Sandstone

, Brentskardhaugen
jj j j n Conglomerate

Sassendalen
Group

Tempelfjorden
Group

A
'&' Hecla Hoek Triassic
Gipsdalen
Devonian U l i Jurassic/
Group
Cretaceous
Carboniferous/ Tertiary o o t> oo
Permian
Billef jorden
< Group

Fig. 1. Simplified geological map of Svalbard. BFZ = Billefjor-
den Fault Zone; LFZ = Lomfjorden Fault Zone; / = Isfjorden;
VM = Van Mijenfjorden. Old Red
z
O
Sandstone

Geological setting and stratigraphy

Spitsbergen displays an almost complete strati-

graphical succession from Late Precambrian to "- z

Palaeogene. The rocks are generally grouped into i'< Hecla- Hoek
five main units separated by major unconformities < z cr
LU d

(Figs. 1 and 2).

Fig. 2. Generalized stratigraphy of Spitsbergen.
Precambrian-Late Silurian
The oldest sequence on Spitsbergen is the Cale-
donian basement complex (Hecla Hoek) of Late deposited in north-south trending fault-bounded
Riphean to Silurian age. Hecla Hoek rocks are dom- grabens (Fig. 1) and are locally strongly deformed.
inated by sediments, with minor volumes of igneous This deformation is thought to be of Late Devonian
rocks, deformed and metamorphosed during the Ny age (Vogt, 1929; Birkenmajer, 1981). Both compres-
Friesland orogenic phase, but there is also evidence sional and transcurrent block movements have been
of an older (Greenvillian?) orogenic event (Orvin, proposed to have taken place along BFZ during
1940; Harland, 1965; Birkenmajer, 1975). Harland this "Svalbardian Orogeny" (McWhae, 1953; Har-
and Gayer (1972) argue that BFZ and LFZ were land, 1965; Harland et al., 1974; Harland et al., 1984;
already established and acted as major fault zones Lamar et al., 1986; Reed et al., 1987).
during the Precambrian.
Carboniferous-Permian
Late Silurian-Devonian Earliest Carboniferous represented a period of
Up to 8 km of late/post-orogenic molasse de- regional crustal extension on Spitsbergen, with for-
posits (Old Red Sandstone) of Late Silurian to mation of graben structures controlled by the old,
Middle Devonian age are unconformably overlying Caledonian, north-south trending lineaments, in-
the Hecla Hoek sequence (Holtedahl, 1914; Friend cluding BFZ and LFZ (Harland, 1969; Steel and
and Moody-Stuart, 1972). The Devonian strata were Worsley, 1984). The Lower Carboniferous deposits
Tertiary decollement thrusting and inversion structures along Billefjorden and Lomjjorden Fault Zones, Spitsbergen 483

(Billefjorden Group, Fig. 2) are characterized by in- between the Adventdalen Group and the Tertiary
terbedded fluviatile sandstones and shales showing deposits is marked by a slight angular unconformity
dramatic facies and thickness variations (Cutbill and formed as a result of the pre-Tertiary uplift of the
Challinor, 1965; Gjelberg and Steel, 1981). Resumed entire Spitsbergen area (Major and Nagy, 1972).
tectonic activity along boundary faults in Middle Car-
boniferous time as evidenced by erosion, renewed Tertiary
graben formation and local compression (Cutbill and Dextral strike-slip movement between Svalbard
Challinor, 1965; Birkenmajer, 1975) has been related and Greenland, related to opening of the North
to continued strike slip movements along the old Atlantic-Arctic Ocean Basins (Eidholm et al., 1987),
lineaments (Birkenmajer, 1975; Steel and Worsley, resulted on Spitsbergen in the Tertiary West Spits-
1984). Middle Carboniferous to early Permian strata bergen Orogeny and the formation of the Central
are characterized by continental red beds, carbonates Spitsbergen Basin (Lowell, 1972; Harland and Hors-
and evaporite sequences (Gipsdalen Group, Fig. 2) field, 1974; Steel et al, 1985).
reflecting more stable tectonic conditions. During The up to 2.3 km thick Van Mijenfjorden Group
Permian time a transition to a stable platform envi- (Fig. 2), making up the Central Spitsbergen Basin,
ronment with carbonate deposition developed (Tem- consists of alternating sandstones and shales of
pelfjorden Group, Fig. 2). mixed marine and continental origin, and is consid-
ered Palaeocene-Eocene in age (Ravn, 1922; Major
Mesozoic and Nagy, 1972; Manum and TYondsen, 1978, 1986;
The platform conditions established during Per- Steel et al., 1985). Steel et al. (1985) argue that
mian, persisted on Spitsbergen throughout the Meso- the Central Spitsbergen Basin (the younger part)
zoic, although resumed extensional movements along represents a foreland basin generated as a result
BFZ and LFZ during the Mesozoic have been pos- of Eocene transpression along the Hornsund Fault
tulated (Parker, 1966, 1967; Harland et al., 1974; Zone located west of Spitsbergen.
Mrk et al., 1982; Larsen, 1987). However, Haremo
and Andresen (1988) and Haremo et al. (1990) ar-
gue, based on structural, sedimentological and strati- Billefjorden fault zone
graphical evidences, that stable tectonic conditions
prevailed along the lineaments also during the Meso- Permian through Tertiary rocks are exposed along
zoic. the BFZ in the study area, extending from Gip-
The Mesozoic succession starts with the Sas- shuken in the north to Kjellstrmdalen in the south
sendalen Group (Early-Middle THassic, Fig. 2), a (Figs. 1 and 3). Due to the regionally southwesterly
300-400 m thick unit consisting mainly of black dip of strata, progressively younger rocks are exposed
shales and siltstones (Buchan et al., 1965; Mrk et towards the south, giving east-west cross-sections at
al., 1982). The overlying Kapp Toscana Group (Late different stratigraphical and structural levels along
Tfriassic-Early Jurassic, Fig. 2) consists of a 300-400 the fault zone (Figs. 3 and 4).
m thick sequence of dark grey and greenish shales The structural development along BFZ in
interbedded with light brown sandstones. Sandstones the area between Adventdalen/Sassendalen and
are the dominant lithology towards the top of the Kjellstrmdalen has been described in detail by
group. Haremo et al. (1990). The most characteristic struc-
The Adventdalen Group (Middle Jurassic-Lower tural features found along BFZ are two, region-
Cretaceous, Fig. 2) has a thickness of 800-1600 m. ally extensive, sub-parallel, east facing anticlines,
The base of the group is marked by the phosphatic the Skolten Anticline to the west and the TVonf-
Brentskardhaugen conglomerate which represents a jellet Anticline to the east (Figs. 3 and 4). These
large hiatus comprising parts of the Lower Jurassic anticlines, and the Drnbreen Syncline separating
and much of the Middle Jurassic (Bckstrm and them, have been mapped from Adventdalen in the
Nagy, 1985). The lower part of the Adventdalen north to Kjellstrmdalen in the south (Fig. 3). The
Group comprises the Janusfjellet Subgroup (Dypvik Skolten Anticline is particularly well developed be-
et al., 1991). This subgroup is composed of a 300- tween Adventdalen and Reindalen (Fig. 3). In this
400 m thick sequence dominated by black shales. The area it is developed as a rather tight anticline with
upper part of the Adventdalen Group consists of a a wavelength of 600-700 meters and an amplitude
300-800 m thick sequence of shales and sandstones of approximately 600 m (Fig. 4d and 5). The folded
(Helvetiafjellet and Carolinefjellet Formations). The Festningen Sandstone can be traced southwards to
base is marked by the 5-15 m thick fine to coarse Kjellstrmdalen where it plunges southward below
grained Festningen Sandstone (Fig. 2). The boundary gently folded Cretaceous and Tertiary strata. The
484 P Haremo and A. Andresen

| T | TERTIARY

ADVENTDALEN GROUP
Carolinefjellet Fm.
Helvetiafjellet Fm.

Janusfjellet Subgr.

KAPP TOSCANA GROUP

Wilhelmeya Fm.
De Geerdalen Fm.
Tschermakfjellet Fm.

SASSENDALEN GROUP

Barentsoya Fm.

TEMPELFJORDEN GROUP

Kapp Starostin Fm.

GIPSDALEN GROUP

Gipshuken Fm.

Nordenskioldbreen Fm.

^ Formation boundary
> Thrust fault
Anticline

Fig. 3. Geological map along the southern part of the Billefjorden Fault Zone, East Central Spitsbergen.
Tertiary decollement thrusting and inversion structures along Billejjorden and Lomfjorden Fault Zones, Spitsbergen 485

A'm

1000 X 1000

4o

1000

m B B'

FLOWERDALEN \ ..
1000 +

1000 1000

ARCTOWSKIFJELLET THRUST J U V D A L S K A M P E N MONOCLINE

yB I
1000f
/ tt l 41000

1000 + 1000

m D D' m
SKOLTEN ANTICLINE TRONFJELLET ANTICLINE

1000 T f 1000

1000 1000
\

10

Fig. 4. Profiles across the Billefjorden Fault Zone; (a) Gipshuken, (b) Flowerdalen, (c) Adventdalen, (d) Reindalen. LDZ = Lower
Decollement Zone; UDZ = Upper Decollement Zone; B = Brentskardhaugen conglomerate; F = Festningen Sandstone. For legend
see Fig. 3.

Skolten Anticline cannot be traced north of Ad- Skolten Anticline. To the south, in Kjellstrmdalen,
ventdalen (Fig. 3). Instead, a major west-dipping the TYonfjellet Anticline is developed as an asymmet-
reverse fault (Arctowskifjellet Thrust, Fig. 4c and 6) rical, kink-like, more open fold with an amplitude of
accomplishes the shortening seen south of the val- less than 200 meters (Fig. 3).
ley. Minimum shortening across the Arctowskifjellet The two fold structures, developed in Upper Juras-
Thrust is 1.5 km. sic and younger sequences, are not recognized below
The TYonfjellet Anticline occurs as an asymmet- the Lower Jurassic Janusfjellet Subgroup (Figs. 3
rical anticline at the Festningen Sandstone level, and 4c, d; Haremo et al., 1990). This structural dis-
located approximately 5 km east of the Skolten An- continuity and the existence of numerous imbricate
ticline (Figs. 3 and 4d). The anticline is particularly as well as minor duplex structures within the Janus-
well developed north and south of Reindalen. Its am- fjellet Subgroup itself, resulting in local thickening,
plitude is slightly less and it is not as tight as the indicate the existence of a decollement zone (Up-
486 P. Hare mo and A. Andresen

Fig. 5. The Skolten Anticline at Rudiaksla north of Reindalen, viewed from the south. The Festningen Sandstone is developed as a
tight, slightly east-verging anticline.

Fig. 6. The Arctowskifjellet Thrust viewed from the south (Adventdalen). This fault ramps up-section through the upper part of the
Adventdalen Group with a throw of at least 300 m. Shortening of the Festningen Sandstone across this fault is at least 1.5 km.
F = Festningen Sandstone.

per Decollement Zone) at this stratigraphical level ate, are repeated three times in a duplex structure
(Figs. 3 and 4; Haremo et al., 1990). (Adventelva Duplex, Figs. 3 and 4c; Haremo et
Compressional structures including folds and re- al., 1990). In addition, data from Flowerdalen and
verse faults of variable size and geometry, unre- Sassendalen (Fig. 3) demonstrate that the upper part
lated to the Skolten and TYonfjellet Anticlines, are of the Sassendalen Group is locally intensely imbri-
also found below the Upper Decollement Zone, in cated and thickened (Fig. 7), indicating the existence
the Sassendalen and Kapp Toscana Groups (Figs. 3 of another decollement zone (Lower Decollement
and 4c, d; Haremo et al., 1990). A particularly well Zone) located at this stratigraphical level (Figs. 3
developed example of the latter is displayed in Ad- and 4).
ventdalen, where the upper part of the Kapp Toscana A major west facing monocline (Juvdalskampen
Group and the lower part of the Janusfjellet Sub- Monocline) located slightly east of the Ttonfjellet
group, including the Brentskardhaugen Conglomer- Anticline (Figs. 3 and 4d), is another character-
Tertiary decollement thrusting and inversion structures along Billefjorden and Lomjjorden Fault Zones, Spitsbergen 487

Fig. 7. Local thickening of the upper, dark, cliff-forming part of the Sassendalen Group by intense reverse faulting and imbrication.
Viewed from the north (Sassendalen).

istic structural feature found along BFZ in the development of the Juvdalskampen Monocline.
Adventdalen-Reindalen area. This monocline, af- Reverse faults, located in the upper part of the
fecting the entire Mesozoic sequence, is not in any Adventdalen Group above the Festningen Sandstone
simple way related to any of the previously described level, are recognized south of Adventdalen (Figs. 3
fold and thrust structures (Haremo et al., 1990). and 4d). This indicates that lateral ramps may exist,
The Juvdalskampen Monocline is interpreted to be along which reverse faults climb up-section towards
controlled by an east-dipping reverse fault at depth south (Haremo et al., 1990). Lateral ramps may
(Fig. 4c). We believe that this fault is exposed fur- also be responsible for the observed local transverse
ther north in Flowerdalen and at Gipshuken (Figs. 3 folding of the Skolten and the TYonfjellet Anticlines
and 4a, b). The Juvdalskampen Monocline trends (Fig. 3).
north-northwest and intersects the TYonfjellet Anti-
cline with an acute angle (Fig. 3). The monocline Lomfjorden fault zone
apparently dies out southwards, although it may link
up with a similar east-west trending monocline in Triassic through Cretaceous rocks are exposed
the inner Kjellstrmdalen area. along the LFZ in the Agardhdalen area (Figs. 1
A dramatic thinning of the Mesozoic sequence in and 8). The structural development along LFZ in
the Adventdalen-Flowerdalen area, has previously this area is described in detail by Andresen et
been interpreted as the result of Mesozoic synde- al. (1992). The dominant structure is a regionally
positional extensional tectonics (Parker, 1966, 1967; extensive, east-facing, highly asymmetric anticline,
Harland et al., 1974). Haremo and Andresen (1988) the Eistraryggen Anticline (Figs. 8 and 9) affecting
and Haremo et al. (1990) argue against such a model the entire Mesozoic sequence including the Fest-
based on structural observations. Alternatively, it is ningen Sandstone. The difference in elevation at
suggested that an out-of-sequence thrust, associated Eistraryggen between the almost flatlying beds east
with the Upper Decollement Zone, cutting down- of the fold structure compared with similar beds west
section in the transport direction, is responsible for of the anticlinal hinge zone is approximately 500 m.
the thinning (Figs. 3 and 4b, c; Haremo et al., The bedding orientation along Eistraryggen is very
1990). Such an explanation is supported by numer- steep and locally inverted (Fig. 9a).
ous small scale out-of-sequence thrusts within the The anticline plunges gently to the south and
Janusfjellet Subgroup. These thrusts are observed to reappears at Klementievfjellet south of Agardhdalen
cut across previously imbricated, steep, west-dipping (Fig. 8). Here, several large parasitic fold structures
reverse faults. Our tectonic model for the reduced are developed in the core of the anticline. Generally,
thickness of the Janusfjellet Subgroup, considers the these folds are asymmetric, kink or chevron-like
out-of-sequence thrusting and thinning to post-date structures, all facing east (Fig. 9b). The eastern limb
488 P. Haremo and A. Andres en

Wilhelmoya Fm.
De Geerdalen Fm.
Tschermakfjellet Fm.
05 SASSENDALEN GROUP

1T
' Barentsoya Fm.

Formation boundary
Thrust fault
Anticline
Syncline

Glacier
Fig. 8. Geological map along the southern part of the Lomfjorden Fault Zone, East Central Spitsbergen.

of some of the parasitic folds is locally inverted. The medium scale reverse faults, locally forming duplex
fold geometry suggests that considerable shortening structures and imbricate fans (Fig. 10). These reverse
has taken place, and there can be no doubt about the faults are rotated into a sub-horizontal orientation
compressional origin of the Eistraryggen Anticline. when traced into the steep, eastern limb of the
We prefer to relate the development of the Eis- Eistraryggen Anticline, suggesting that decollement
traryggen Anticline to a west-dipping reverse fault at thrusting along LFZ predates the generation of the
depth (Fig. 9; Andresen et al., 1988, 1992). This is anticline (Fig. 9). Isolated reverse faults are also ob-
supported by observations along LFZ toward its type served in many places within the Sassendalen Group
area to the north, were pre-Mesozoic rocks including west of the Eistraryggen Anticline, indicating that
Hecla Hoek rocks are uplifted and exposed west of the Lower Decollement Zone found along BFZ con-
the fault (Flood et al., 1971; Hjelle and Lauritzen, tinues eastward to LFZ (Andresen et al., 1992).
1982). Small scale imbrications and duplexes are also ob-
A characteristic structural feature within the upper served in the Janusfjellet Subgroup, at Klementiewf-
part of the Sassendalen Group north of Agard- jellet suggesting the existence of a decollement zone
hdalen, is the development of numerous, small and (Upper Decollement Zone) also at this stratigraph-
Tertiary decollement thrusting and inversion structures along Billefjorden and Lomfjorden Fault Zones, Spitsbergen 489

m A Eistraryggen m1000
1000 _Kb/t
?
UDZ 9 " 1~^ -
. 7 7 . : _ . - . - . - - - - " ' * *\ \
500

+ 500

B B'm 1000
1000?
Klementievfjellet

500 ^.9 500

;^-?z^^^^^^^
4o

5001 *s -\ 500

1 km
Fig. 9. Profiles across the Lomfjorden Fault Zone, (a) Eistraryggen, (b) Klementievfjellet. LDZ = Lower Decollement Zone; UDZ
Upper Decollement Zone. For legend see Fig. 8.

ical level. The Janusfjellet Subgroup thins across deformational structures are particularly frequent
Klementievfjellet (Fig. 9b), a feature interpreted to along these two fault zones. Local thickening of
be the result of out-of-sequence thrusting, also con- both the Sassendalen Group and the Janusfjellet
firming the presence of an Upper Decollement Zone Subgroup due to small scale imbrication and duplex
in the Janusfjellet Subgroup this far east (Andresen development, is recognized in the area between the
et al., 1992). two fault zones, especially in Sassendalen (Lower
Decollement Zone, Figs. 3 and 7) and in the in-
Discussion ner parts of Reindalen and Kjellstrmdalen (Upper
Decollement Zone, Fig. 3). Similar structures are
Tertiary tectonic activity along BFZ and LFZ has also developed at the same stratigraphical levels west
been described by several authors (De Geer, 1919; of BFZ (Haremo and Andresen, 1988; Nttvedt et
Hagermann, 1925; Orvin, 1940; Parker, 1966; Har- al., 1988; Haremo et al., 1990), as well as along the
land et a l , 1974). The importance of the Tertiary western coast of Spitsbergen (Dallmann, 1988; Ma-
deformation has, however, been disputed (Andresen her, 1988; Ringseth, 1988). It therefore seems to be
et al., 1986; Haremo and Andresen, 1988; Dypvik well documented that the decollement zones recog-
et al., 1990; Haremo et al., 1990). The structural nized in the Mesozoic strata along BFZ and LFZ
development described herein, demonstrating that have a regional extent.
Tertiary deformation along BFZ and LFZ has been The data presented above, on the Tertiary defor-
underestimated, is also supported by other recent mation on Spitsbergen, suggest that the traditional
investigations (e.g., Bergh et al., 1988; Faleide et al., view of the West Spitsbergen Orogeny represent-
1988; Maher, 1988; Nttvedt et al., 1988). ing a transpressive orogen with structures typical of
The decollement zones recognized in the Mesozoic wrench tectonics (Lowell, 1972; Kellogg, 1975; Har-
strata are not restricted to BFZ and LFZ, although land and Horsfleld, 1974; Steel et al., 1985), has to be
490 P. Haremo and A. Andresen

Fig. 10. Duplex structures developed in the upper part of the Sassendalen Group at Eistraryggen. Viewed toward the south.

modified. The Tertiary deformation structures recog- along this detachment zone may have resulted in
nized in the foldbelt, as well as along BFZ and LFZ reactivation along LFZ.
indicate that Tertiary deformation is purely compres- During stage 3 (Fig. 11) deformation along BFZ
sional and similar to structures recognized in typical and LFZ is restricted mainly to decollement thrust-
fold and thrust belts (e.g., Dahlstrm, 1970; Boyer ing along the Upper and Lower Decollement Zones.
and Elliot, 1982; Mitra, 1986; Baker et al., 1988). Along BFZ, movement along the Lower Decolle-
Placed in this context an evolutionary model, il- ment Zone results in the development of the Ad-
lustrating the Tertiary deformational history along vent elva Duplex, the location of which is most likely
BFZ and LFZ, is presented in Fig. 11. Figure 11, controlled by an underlying fault. More or less simul-
stage 1, illustrates a pre-deformational (pre-Tertiary) taneous movement along the Upper Decollement
stage, in which structural discontinuity is restricted to Zone creates the TYonfjellet Anticline, above the
the existence of basement-rooted extensional faults Adventelva Duplex. Decollement thrusting extends
(BFZ and LFZ) of pre-Mesozoic age. In this model, eastwards, all the way to LFZ where imbricate fans
these faults are thought to act as "structure genera- and duplex structures are developed during stage 3.
tors" during later deformation. The next stage of Tertiary deformation along BFZ
In stage 2, Tertiary compression commences, re- (Fig. 11, stage 4) is characterized by out-of-sequence
sulting in the initiation of decollement zones in the thrusting, related to movements along the Upper
Sassendalen Group (Lower Decollement Zone) as Decollement Zone. The Tronfjellet Anticline is de-
well as in the Janusfjellet Subgroup (Upper Decolle- capitated and displaced eastwards. As it reaches the
ment Zone). Decollement thrusting is, however, Juvdalskampen Monocline, movement ceases and
thought to be limited during stage 2. Inversion along the Arctowskifjellet Thrust (or the Skolten Anticline
BFZ is the principal event during this stage. Reacti- to the south) is developed above the Adventdalen
vation of the steep, east-dipping, pre-Mesozoic fault Duplex. Along LFZ this late stage deformation is
lines results in the generation of the Juvdalskampen characterized by the development of the Eistraryggen
Monocline and possibly also in some reactivation Anticline as a response to reverse movement along
along the inferred western fault line (Fig. 11). In- the steep, west-dipping, basement-rooted fault. Out-
version along BFZ during this stage, is probably of-sequence thrusting resulting in tectonic thinning
not only a result of simple reactivation along old of the Janusfjellet Subgroup at Klementievfjellet, is
fault lines. Data from Gipshuken (Ringseth and An- related to movement on the Upper Decollement
dresen, 1988) and the inner part of Billefjorden Zone, and is thought to post-date inversion and to
suggest that a detachment level existed in the Car- represent the final deformational event along LFZ
boniferous evaporite deposits (Fig. 11). Movement (Andresen et al., 1992).
Tertiary decolkment thrusting and inversion structures along Billefjorden and Lomfjorden Fault Zones, Spitsbergen 491

BILLEFJORDEN FAULT ZONE LOMFJORDEN FAULT ZONE

1. ._*_
:">]
_ ^ c.

*Jkt Vkt

2.

3.

CD Tertiary HU SassendaJen Gr. (Tri.)

5 Helvetiafjellet-/Carolinefjellet Fm. (Cret.) GD Permian
DID Janusfjellet Subgr. (Jur./Cret.) (XU Carboniferous
5SD Kapp Toscana Gr. (Tri./Jur.) GEH Devonian
[HD Hecla Hoek

i5km
N
Fig. 11. Evolutionary model illustrating Tertiary deformational history along Billefjorden and Lomfjorden Fault Zone.

The Tertiary deformational history presented ment along the two decollement zones, is mainly
above suggest an eastward displacement of Lower based on the tectonic thinning of Mesozoic strata
Cretaceous and younger rocks across BFZ of mini- across the Juvdalskampen Monocline, indicating that
mally 3-4 km, and locally as much as 10 km. A fea- inversion pre-dates out-of-sequence thrusting (Fig. 4;
ture not included in the kinematic model presented Haremo et al., 1990). The opposite age relationship
above is the along-strike variation in geometries is inferred for the thick- and thin-skinned deforma-
observed along BFZ (Figs. 3 and 4). The most inter- tion observed along LFZ (Fig. 11). Here, inversion
esting observation in this respect, is the appearance post-dates decollement thrusting as reverse faults
of thrust faults at higher stratigraphical levels (above generated through decollement thrusting during the
the Festningen Sandstone) to the south. We prefer development of the Eistraryggen Anticline are ro-
to view these as caused by the existence of lateral tated into a sub-horizontal or even east-dipping ori-
or oblique ramps, probably controlled by the pre- entation (Fig. 9; Andresen et al., 1992).
deformational Late Cretaceous/Early Tertiary uplift The interaction between thick- and thin-skinned
and southward tilting of Spitsbergen. deformation described above, of which the latter
The model presented in Fig. 11 proposing that can be linked to the West Spitsbergen Orogeny,
inversion along BFZ preceded most of the move- suggests that they are formed more or less contem-
492 P. Haremo and A. Andresen

poraneously. As the thin-skinned folds affect strata kontra tertiaer tektonisk aktivitet lngs Billefjord-lineamentet
of Palaeocene age (Manum and TYondsen, 1986), syd for Isfjorden, Spitsbergen. 10. Nor. Geol. Vintermte
Trondheim 1987. Geolognytt, 21: 14 (abstract).
the deformation is considered post-Palaeocene in Andresen, A., Haremo, P. and Bergh, S., 1988. The southern
age. Since transpression along the Hornsund Fault termination of the Lomfjorden Fault Zone; evidences for
Zone west of Spitsbergen is thought to be of Late Tertiary compression on east Spitsbergen. Symp. Tertiary Tec-
Palaeocene-Middle Eocene age (Steel et al., 1985; tonics of Svalbard, Oslo, 26-27 April 1988, Norsk Polarinst.
Eldholm et al., 1987) an Eocene age is suggested for Rap., 46: 75-78 (extended abstract).
Andresen, A., Haremo, P., Swensson, E. and Bergh, S.G., 1992.
the deformational structures developed along BFZ Structural geology around the southern termination of the
and LFZ. Lomfjorden Fault Complex, Agardhdalen, East Spitsbergen.
Nor. Geol. Tidskr., 72 (in press).
Conclusion Baker, D.M., Lillie, R.J., Yeats, R.S., Johnson, G.D., Yousuf,
M. and Zamin, A.S.H., 1988. Development of the Himalayan
Most lateral thickness variations within the Juras- frontal thrust zone: Salt Range, Pakistan. Geology, 16: 3-7.
Bergh, S.G., Andresen, A., Bergvik, A. and Hansen, A.I., 1988.
sic and Thassic strata on Central Spitsbergen are
Tertiary thin-skinned compressional deformation on Oskar II
due to Tertiary compressional tectonics, linked to Land, Central Vest-Spitsbergen. Symp. Tertiary Tectonics of
the West Spitsbergen Orogeny. The Tertiary de- Svalbard, Oslo, 26-27 April 1988, Nor. Polarinst. Rap., 46:
formation is characterized by an interacting thin- 51-54 (extended abstract).
and thick- skinned structural style. Two regionally Birkenmajer, K., 1975. Caledonides of Svalbard and plate tec-
extensive decollement zones are recognized in the tonics. Geol. Soc. Denm. Bull., 24: 1-19.
Birkenmajer, K., 1981. The geology of Svalbard, the western
Mesozoic strata; a lower one in the Lower Thas- part of the Barents Sea, and the continental margin of
sic Sassendalen Group, and an upper one in the Scandinavia. In: A.E.M. Nairn, M. Churkin Jr. and EG.
Upper Jurassic/Lower Cretaceous Janusfjellet Sub- Stehli (Editors), The Ocean Basins and Margins, Vol. 5. The
group. Compressional structures (folds and reverse Arctic Ocean. Plenum Press, New York, N.Y., pp. 265-329.
faults) are particularly frequent along the basement- Boyer, S.E. and Elliot, D., 1982. Thrust systems. Am. Assoc. Pet.
Geol., Bull., 66: 1196-1230.
rooted Billefjorden and Lomfjorden Fault Zones. Buchan, S.H., Challinor, A., Harland, W.B. and Parker, J.R.,
Preliminary calculations indicate an eastward dis- 1965. The Triassic stratigraphy of Svalbard. Nor. Polarinst.
placement of minimally 3-4 km, possibly as much Skr., 135: 1-72.
as 10 km, of Lower Cretaceous and younger rocks Bckstrm, S.A. and Nagy, J., 1985. Depositional history
across BFZ. The complex Tertiary deformation rec- and fauna of a Jurassic phosphorite conglomerate (the
Brentskardhaugen bed) in Spitsbergen. Nor. Polarinst. Skr.,
ognized along BFZ and LFZ are caused by reac-
183: 1-61.
tivation of steep, opposite-dipping basement-rooted Cutbill, J.L. and Challinor, A., 1965. Revision of the strati-
reverse faults during the West Spitsbergen Orogeny. graphical scheme for the Carboniferous and Permian rocks of
This caused inversion of the Carboniferous-Permian Spitsbergen and Bjrnya. Geol. Mag., 102: 418-439.
Billefjorden Basin. Thick-skinned faulting preceded Dallmann, W.K., 1988. Thrust tectonics south of Van Keulenfjor-
den. Symp. Tertiary Tectonics of Svalbard, Oslo, 26-27 April
thin-skinned thrusting along BFZ, whereas the oppo-
1988, Nor. Polarinst. Rap., 46: 43-46 (extended abstract).
site relationship is inferred along LFZ. Dahlstrm, CD.A., 1970. Structural geology in the eastern
margin of the Canadian Rocky Mountains. Can. Pet. Geol.
Bull., 18: 332-406.
Acknowledgements
De Geer, G., 1919. Om Spetsbergens natur i Sveagruvans
omnejd. Ymer, H4: 240-277.
The present study is part of the BP R&D project
Dypvik, H., Andresen, A., Haremo, P., Nagy, J., Elverh0i, A ,
1068 "Study of the sedimentological and structural Eikeland, T.A., Johansen, H. and Strand, K., 1990. Sedimen-
evolution of the area between Kjellstrmdalen and tation and tectonics in the Jurassic/Cretaceous sequences of
Adventdalen/Sassendalen, Central Spitsbergen". We Central Spitsbergen. In: V.M. Kotlyakov and V.E. Sokolov
thank BP Petroleum Development (Norway), Ltd. (Editors), Arctic Research. Advances and Prospects. Proc.
for financial support to the project and for per- Conf. of Arctic and Nordic Countries on Coordination of
Research in the Arctic, Leningrad, December 1988, Part 1.
mission to publish the results. We also thank the Nauka, Moscow, pp. 348-354.
Norwegian Polar Research Institute, Norsk Hydro Dypvik, H., Nagy, J. Eikeland, T.A., Backer-Owe, K., Andresen,
and Statoil for valuable logistical support during A., Haremo, P., Bjasrke, T, Johansen, H. and Elverh0i, A ,
fieldwork. Einar Sverdrup, Eivind Swensson and two 1991. The Janusfjellet Subgroup (Bathonian to Hauterivian)
unknown referees are thanked for valuable com- on central Spitsbergen: a revised lithostratigraphy. Polar Res.,
9(1): 21-44.
ments on a preliminary draft of the manuscript. Eldholm, O., Faleide, J.I. and Myhre, A.M., 1987. Continental-
ocean transition at the western Barents Sea/Svalbard conti-
References nental margin. Geology, 15: 1118-1122.
Faleide, J.E., Gudlaugsson, S.T., Eiken, O. and Hanken, N.M.,
Andresen, A., Dypvik, H., Nagy, J., Elverh0i, A., Eikeland, .., 1988. Seismic structure of Spitsbergen: Implications for Ter-
Haremo, P., Nesteby, H. and Strand, K., 1986. Mesozoisk tiary deformation. Symp. Tertiary Tectonics of Svalbard, Oslo,
Tertiary decollement thrusting and inversion structures along Billefjorden and Lomfjorden Fault Zones, Spitsbergen 493

26-27 April 1988, Nor. Polarinst. Rap., 46: 47-50 (extended tional Lending Library for Science and Technology, U.K.,
abstract). pp. 59-75.
Flood, B., Nagy, J. and Winsnes, T.S., 1971. Geological map Lowell, D.L., 1972. Spitsbergen Tertiary Orogenic Belt and the
of Svalbard 1:500,000, Sheet IG, Spitsbergen southern part. Spitsbergen Fracture Zone. Geol. Soc. Am. Bull., 83: 3091-
Nor. Polarinst. Skr., 154A. 3102.
Friend, P.F. and Moody-Stuart, M., 1972. Sedimentation of the Mher, H.D., Jr., 1988. Minimum estimate of Tertiary shortening
Wood Bay Formation (Devonian) of Spitsbergen: regional suggested by surface structures exposed on Midterhuken,
analysis of a late orogenic basin. Nor. Polarinst. Skr., 157: Bellsund, Spitsbergen. Symp. Tertiary Tectonics of Svalbard,
1-77. Oslo, 26-27 April 1988, Nor. Polarinst. Rap., 46: 35-38
Gjelberg, J.G. and Steel, R.J., 1981. An outline of lower-middle (extended abstract).
Carboniferous sedimentation on Svalbard: effects of tectonic, Major, H. and Nagy, J., 1972. Geology of the Adventdalen map
climatic and sea level changes in rift basin sequences. Can. area. Nor. Polarinst. Skr., 138: 1-58.
Soc. Pet. Geol., Mem., 7: 543-561. Manum, S.B. and Trondsen, T, 1978. Rank of coal and dispersed
Hagermann, T.H., 1925. Stratigraphic and structural investiga- organic matter and its geological bearing in the Spitsbergen
tions within south-western Spitzbergen. Geogr. Ann., 7: 195- Tertiary. Nor. Polarinst., rbok, 1977: 159-177.
221. Manum, S.B. and Trondsen, T, 1986. Age of Tertiary formations
Haremo, P. and Andresen, A., 1988. Tertiary movements along on Spitsbergen. Polar Res., 4: 103-131.
the Billefjorden FaultZone and its relation to the Vest- McWhae, J.R.H., 1953. The Major Fault Zone of Central
Spitsbergen Orogenic Belt. Symp. Tertiary Tectonics of Sval- Spitsbergen. Q. J. Geol. Soc. London, 108: 209-232.
bard, Oslo, 26-27 April 1988, Nor. Polarinst. Rap., 46: 71-74 Mitra, S., 1986. Duplex structures and imbricate thrust systems:
(extended abstract). geometry, structural position, and hydrocarbon potential. Am.
Haremo, P., Andresen, A., and Dypvik, H., in prep. Mesozoic Assoc. Pet. Geol., Bull., 70: 1087-1112.
extension versus Tertiary compressionalong the Billefjorden Mrk, A., Knarud, R. and Worsley, D., 1982. Depositional and
Fault Zone south of Isfjorden, central Spitsbergen. diagenetic environments of the Triassic and lower Jurassic
Haremo, P., Andresen, A., Dypvik, H., Nagy, J., Elverh0i, A., succession of Svalbard. Can. Soc. Pet. Geol., Mem., 8: 371-
Eikeland, T.A. and Johansen, H., 1990. Structural develop- 398.
ment along the Billefjorden Fault Zone in the area between Nyland, B., Jensen, L.N., Skagen, J.I., Skarpnes, O. and Vorren,
Kjellstrmdalen and Adventdalen/Sassendalen, Central Spits- T, 1992. Tertiary uplift and erosion in the Barents Sea;
bergen. Polar Res., 8(2): 195-216. magnitude, timing and consequences. In: R.M. Larsen, H.
Harland, W.B., 1965. Discussion of the tectonic evolution of the Brekke, B.T Larsen and E. Talleraas (Editors), Structural and
Arctic-North Atlantic Region. Philos. Trans. R. Soc. London, Tectonic Modelling and its Application to Petroleum Geology.
Ser. A, 258: 59-75. Norwegian Petroleum Society (NPF), Special Publication 1.
Harland, W.B., 1969. Contribution of Spitsbergen to understand- Elsevier, Amsterdam, pp. 153-162 (this volume).
ing of the tectonic evolution of North Atlantic Region. Am. Nttvedt, A., Livbjerg, F and Midbe, P.S., 1988. Tertiary defor-
Assoc. Pet. Geol., Mem., 12: 817-851. mation of Svalbard various models and recent advances.
Harland, W.B. and Gayer, R.A., 1972. The Arctic Caledonides Nor. Polarinst. Rap., 46: 79-84.
and earlier Oceans. Geol. Mag., 109: 289-314. Orvin, A.K., 1940. Outline of the geological history of Spitsber-
Harland, W.B. and Horsfield, W.T., 1974. The West Spitsbergen gen. Skr. Svalbard Oglshavet, 78: 1-57.
Orogen. Geol. Soc. London, Spec. Publ., 4: 747-755. Parker, J.R., 1966. Folding, faulting and dolerite intrusions in
Harland, W.B., Cutbill, J.L., Friend, P.F, Gobett, D.S., Holliday, the Mesozoic rocks of the fault zone of central Spitsbergen.
D.W., Maton, P.I.,Parker, J.R. and Wallis, R.H., 1974. The Nor. Polarinst., Arbok, 1964: 47-55.
Billefjorden Fault Zone, Spitsbergen. Nor. Polarinst. Skr., Parker, J.R., 1967. The Jurassic and Cretaceous sequence in
161: 1-72. Spitsbergen. Geol. Mag., 104: 487-505.
Harland, W.B., Gaskell, B.A., Heafford, A.P., Lind, E.K. and Ravn, J.P.J., 1922. On the Mollusca of the Tertiary of Spitsber-
Perkins, P.J., 1984. Outline of Arctic post-Silurian continental gen. Result. Nor. Spitsbergeneksp. 1(2): 1-28.
displacements. In: A.M. Spencer (Editor), Petroleum Geology Reed, W.E., Douglass, D.N. and Lamar, D.L., 1987. Devonian
of the North European Margin. Proceedings of the North Old Red Sandstone sedimentation and tectonic history of
European Margin Symposium. Norwegian Petroleum Society, Billefjorden fault zone, Spitsbergen. Geol. Mijnbouw, 66:
Graham and Trotman, London, pp. 137-148. 191-199.
Hjelle, A and Lauritzen, ., 1982. Geological map of Sval- Riis, F and Fjeldskaar, W, 1992. On the magnitude of the Late
bard, 1:500,000, Sheet 3G, Spitsbergen, Northern part. Nor. Tertiary and Quaternary erosion and its significance for the
Polarinst. Skr., 154C: 1-15. uplift of Scandinavia and the Barents Sea. In: R.M. Larsen, H.
Holtedahl, O., 1914. On the Old Red Sandstone Series of Brekke, B.T. Larsen and E. Talleraas (Editors), Structural and
North Western Spitsbergen. Int. Geol. Congr. CRXIIth Sess., Tectonic Modelling and its Application to Petroleum Geology.
Canada 1913: 707-712. Norwegian Petroleum Society (NPF), Special Publication 1.
Kellogg, H.E., 1975. Tertiary Stratigraphy and Tectonism in Elsevier, Amsterdam, pp. 163-185 (this volume).
Svalbard and Continental Drift. Am. Assoc. Pet. Geol., Bull., Ringseth, N., 1988. The fold and thrust system of Midterhuk-
59: 465-485. fjellet, Bellsund. Symp. Tertiary Tectonics of Svalbard, Oslo,
Lamar, D.L., Reed, WE. and Douglass, D.N., 1986. Billefjorden 26-27 April 1988, Nor. Polarinst. Rap., 46: 39-42 (extended
fault zone, Spitsbergen: is it part of a major Late Devonian abstract).
transform? Geol. Soc. Am. Bull., 97: 1083-1088. Ringseth, N. and Andresen, A., 1988. The Gipshuken Fault
Larsen, V.B., 1987. A synthesis of tectonically related stratig- System evidence for Tertiary thrusting along the Billefjor-
raphy in the North Atlantic-Arctic region from Aalenian to den Fault Zone. Symp. Tertiary Tectonics of Svalbard, Oslo,
Cenomanian time. Nor. Geol. Tidsskr., 67: 281-293. 26-27 April 1988, Nor. Polarinst. Rap., 46: 67-70 (extended
Livshits, Y.Y., 1965. Tectonics of central Vestspitsbergen. In: abstract).
V.N. Sokolov (Editor), Geology of Spitsbergen, 1. Na- Steel, R.J. and Worsley, D., 1984. Svalbard's post-Caledonian
494 P. Haremo and A. Andresen

strata An atlas of sedimentational patterns and palaeo- Steel, R.J., Gjelberg, J., Helland-Hansen, W, Kleinspehn, K.,
geographic evolution. In: A. Spencer (Editor), Petroleum Nttvedt, A. and Rye-Larsen, M., 1985. The Tertiary strike-
Geology of the North European Margin. Proceedings slip basins and orogenic belt of Spitsbergen. Soc. Econ.
of the North European Margin Symposium. Norwegian Paleontol. Mineral., Spec. Publ., 37: 339-359.
Petroleum Society. Graham and Trotman, London, pp. 109- Vogt, T, 1929. Den norske fjellkjedes revolusjonshistorie. Nor.
135. Geol. Tidsskr., 10: 97-115.

P. HAREMO University of Oslo, Department of Geology, P.O. Box 1047, Blindern, 0316 Oslo 3, Norway
Present address: Norsk Hydro AI S, Drammensveiem 264, P.O. Box 200, N-1321 Stabekk, Norway
A. ANDRESEN University of Oslo, Department of Geology, PO. Box 1047, Blindern, 0316 Oslo 3, Norway
 495

Pre-Cretaceous structural development of the Danish

Central Trough and its implications for the distribution of
Jurassic sands
O. Michelsen, T.E. Mogensen and J.A. Korstgrd

The structural development of the Danish Central Trough has been studied to determine the age of graben development and
to reconstruct the paleogeographic distribution of Jurassic sands.
The Late Hercynian compression seems to have influenced the area. The presence of Rotliegende volcanics may be related
to transitional movements along a right lateral, NW-SE strike-slip fault system. The Zechstein, Iriassic and Early Jurassic times
are regarded as stable tectonic periods during which the Central Trough area may have acted as an inter-basinal barrier.
Graben development started along the Coffee Soil Fault in the Middle Jurassic; subdivision into the smaller half grabens took
place in the Late Jurassic.
Fluvio-deltaic, Middle Jurassic sands were deposited along the Coffee Soil Fault, indicating a transgression from the north.
Oxfordian to early Kimmeridgian, lower delta plain to offshore marine sands were deposited in the area of the later Heno
Plateau, Feda and Gertrud Grabens, indicating a progressive westward transgression from the eastern region of the trough.

Introduction Our study is based on released seismic and well

data (Fig. 1). North of profile 6 (Fig. 1) the seis-
A number of earlier regional papers have dealt mic database consists of surveys: CGT81, RTD81,
with the subject, presenting a simple picture of ELF8083, SP82, NH82. South of profile 6, pub-
graben development from the Permian to the Juras- lished lines from M0ller (1986), Vejbaek (1986) and
sic. Gowers and Saeb0e (1985) presented a more Cartwright (1987) have been used.
differentiated outline of the structural development,
which we regard as more realistic. The Jurassic
and Early Cretaceous structural outline described by Structural and depositional development
M0ller (1986) and Vejbaek (1986), respectively, will
be followed in this paper (see Fig. 1). However, Pre-Permian
the major bordering fault zone of the Ringk0bing- Precambrian rocks were drilled on the Mandal
Fyn High will here be referred to as the Coffee High (3/7-1 well) and, with a Caledonian overprint,
Soil Fault, following Gowers and Sasb0e (1985). The on the edge of the Ringk0bing-Fyn High (Per-l
sedimentary facies and depositional environments of well); Caledonian metamorphosed rocks were en-
the Jurassic sands are described by Frandsen (1986), countered by the P-l well on the Inge High (Fig. 1).
Jensen et al. (1986) and Johannessen (pers. com- These wells are situated on the footwall part of major
mun., 1989). The biostratigraphic correlation used in basement fault blocks (Fig. 1). These footwalls have
the paleogeographic reconstruction is based on data been relatively uplifted and suffered deep erosion.
published in the aforementioned papers or derived The uplift of the footwalls and the configuration of
from internal reports of the Geological Survey of the deep reflectors (see Fig. 2, A and B) are in
Denmark. Within the last few years the amount of accordance with recent models for the elastic and
released data has increased, creating the opportu- flexural behaviour of the upper crust during faulting.
nity to test earlier models. Our primary aims were These models, e.g., the "flexural cantilever model"
firstly to investigate how far back in geological time (Kusznir et al., 1988), predict both footwall uplift
graben development can be justified or proved by the and a concave upwards configuration of bed geome-
data and secondly to reconstruct the paleogeographic tries in the footwall. Application of these models
distribution of Jurassic sand. explains the locally high positions of Caledonian and

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 495-506. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
496 O. Michelsen, T.E. Mogensen and JA. Korstgrd

DANISH CENTRAL
TROUGH

JURASSIC
Structural elements

Legend:

op1 Well encountering pre Jurassic

but not Jurassic
Q-1 Well encountering Jurassic
/^) Salt structure
- Normal fault
-r" Reverse fault
^- Strike slip fault

3**** Seismic line

1
Seismic line (fig.3)

-H
10 20Km

Fig. 1. Jurassic structural elements in the Danish Central Trough. Location of wells available for the study and location of the seismic
lines presented in Fig. 2, modified from M0ller (1986).

Precambrian basements covered by Late Cretaceous Top Rotliegende may well delineate the Carbonifer-
sediments. ous sequence.
Carboniferous sediments are encountered in the At the Mid North Sea High, the uniformly thick
P-l well (at the Inge High) and the 2/10-1 well (in Carboniferous sequence forms W-E striking anti-
the adjacent Norwegian sector). The Carboniferous clines (Fig. 3) which may indicate a late Hercynian
sequence can be identified in the same area on the compressive phase, a compression that might have
seismic data and is probably present through most of formed the major W - E striking Mid North Sea
the Danish Central TVough. Seismic reflectors below High-Ringk0bing-Fyn High system (Pegrum, 1984).
 WSW ENE
2 sec. TWT

P r 1
O f i |e 2 sec. TWT

Profile 2
2 sec. TWL

A 4 sec TWT
\\^\^^^ "'S '
1 I H
\ \ 2 ~ ^
/
Profile 3 " AE ^ ^ ^ ^ ^ ^ ^
(4-5

2 sec. TWT

RFH 4 sec TWT

SSSSS^ -

LEGEND: Profile 4
(4-5?)

liiiliiiiiiiiiiil UPPER CRETACEOUS 2sec.TWT

1
'' LOWER CRETACEOUS ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^^^^^ ^^

-^ MIDDLE JURASSIC ^ ^ ^ 1 MH \ \ \ \^^^^^^^^^-':^^^^^^^,^

I.'" 1 LOWER JURASSIC Profile 5 AE \ \ ^ ^ ^ ^ ^ c ^ i i ^ ^ ^

f 1 (4-5?)
1 1 ZECHSTEIN
j 2 sec TWT
I I ROTLIEGENDE _^ , ^ ^ ^ ^ ^ ^ ^ ^ ^ 1 ,,;.; ynx^m-^^^^^ ,

A E
Profile 6 ^ ^ ^ ( 4 5?)

Fig. 2. Profiles based on seismic interpretation across the Central Trough (A and B, see text). 1-6 indicate connected faults. AE = Arne Elin Trend; IH = Inge High; MH = Mads High; MaH
Mandal High; RFH = Ringk0bing Fyn High. For locations see Fig. 1.
498 O. Michelsen, T.E. Mogensen and JA. Korstgrd

Fig. 3. Seismic section showing a Carboniferous fold on the Mid North Sea High. TC = Top Chalk; TPZ = Top Pre Zechstein; TCa
Top Carboniferous; ICa 1-2 = Intra Carboniferous 1-2. For locations see Fig. 1.

Rotliegende the later Danish Central Trough. Pegrum (1984)

According to Glennie (1984, 1986) and Ziegler and Cartwright (1987) also proposed transtensional
(1982, 1987), the Hercynian fold belt collapsed in movements in the Danish Central Though area, but
Stephanian-Autunian times, due to the development they related these to movements along a broad Torn-
of a NW-SE orientated right lateral transform fault quist Zone.
system. This system caused a complex pattern of con-
jugate shear faults and related pull-apart structures Zechstein
and volcanism over large areas of Northwest Europe Zechstein evaporites have been drilled in the
(Fig. 4). S0gne Basin to the north and in the Salt Dome
Rotliegende volcanics are encountered in a num- Province to the south (Michelsen and Andersen,
ber of wells drilled in the northern part of Danish 1983) and marginal carbonate facies were encoun-
Central TYough and adjacent areas, indicating tec- tered in the Diamant-1 well on the Heno Plateau.
tonic activity (Fig. 4). Ziegler (1982, 1987) refers Seismic mapping shows a distribution of the Zech-
the volcanism to evolution of the Mid North Sea stein sediments which is in accordance with that
High-Ringk0bing-Fyn High system, while Glennie described by Gowers and Saeb0e (1985) (see Fig. 5).
(1984, 1986) relates the volcanism to a postulated The evaporites are found in trends that correlate
opening of a graben structure in the Central TYough with the proposed NW-N trending right lateral fault
area. It is not possible from available data to estab- system. However, the present day distribution of the
lish whether the faults, e.g., the Coffee Soil Fault, Zechstein sediments is controlled by faults that were
were part of the fault systems proposed by Ziegler primarily active in the Jurassic and the distribution
and Glennie (op. cit). However, a few minor NNW may, therefore, be largely a result of Mesozoic ero-
trending Rotliegende faults are recognized in the sion.
S0gne Basin, which could be a part of a NW-N Deeply buried Zechstein evaporites have been in-
trending fault system. terpreted from seismic data to be present in the
Our suggestion is that the apparent NNW trend- TM1 End Graben (Fig. 2), thus indicating a connec-
ing distribution of the volcanics (Fig. 4) and the tion between the Northern and Southern Permian
observed N-NW fault orientations, may be related Basins through the S0gne Basin-TMl End Graben-
to transtensional movements, through the area of Salt Dome Province trend. However, the marginal
Pre-Cretaceous structural development of the Danish Trough and its implications for the distribution of Jurassic sands 499

LATE CARBONIFEREOUS AND EARLY PERMIAN

DISTRIBUTION OF
VOLCANICS AND FAULTS.
Legend
Highs in the North Sea P^sJ Major faults W:-:\ Volcanic \'-~^ Dykes

Fig. 4. Stephanian-Autunian distribution of volcanics and major faults. Modified from Glennie (1986). The inset map shows locations
for the wells penetrating Rotliegende volcanics in the investigated area.

carbonate facies on the Heno Plateau and the thin

evaporites and/or carbonates in the middle of the Tail
End Graben point towards an inter-basinal barrier in
the Central Though.
The Zechstein evaporites were possibly deposited
along trends subsiding due to thermal effects, caused
by the Rotliegende heat event.

Triassic
In the southern part of the Danish Central TYough
more than 1500 m of Tlriassic sediments are present,
representing the entire TYiassic time interval. The
facies succession shows that the sequence is a part
of the depositional pattern known from the south-
ern North Sea, which is characterized by repeated
marine transgression (Jacobsen, 1982; Michelsen and
Andersen, 1983). Farther to the north, in the Q-l,
W.Lulu-1 and Cleo-1 wells, 2-400 m of Lower to
lowermost Middle TViassic sediments have been pen- Fig. 5. Distribution of different seismic expressions of the
etrated (Fig. 6). These sediments are referred to the Zechstein sequence. From Gowers and Sasb0e (1986).
 GERT-1 KARL-1 W. LULU-1 CLEO-1

UPPER
CRETACEOUS

LOWER
CRETACEOUS
UPPER JRASSi5~ fcrzr

ROTLIEGENDE
UPPER
CARBONIFE- JURASSIC
ROUS

MIDDLE
JURASSIC

M.-L.
WELL LOCATION TRIASSIC

3*!'

Fig. 6. Lithostratigraphic subdivision of the Mesozoic and pre-Mesozoic in well sections from the northern Danish Central Trough.
Pre-Cretaceous structural development of the Danish Trough and its implications for the distribution of Jurassic sands 5

northern facies province, characterized by continen- Middle Jurassic

tal plain deposits. However, seismic and well data The Middle Jurassic sediments are known from the
indicate that TYiassic sediments were deposited in Edna-1, Jens-1, Lulu-1, M-8, Nora-1, O-l, U-l and
the entire trough area, which is in agreement with W.Lulu-1 wells in the S0gne Basin, Tail End Graben
Gowers and Saeb0e (1985). TYiassic sediments are en- and Salt Dome Province (Fig. 1). The sequence con-
countered in the V-l well on the later down-faulted sists of fluvial-deltaic sediments that show evidence
Poul Plateau, indicating that deposition in the TYias- of increasing marine influence upwards (Jensen et
sic also occurred on the Ringk0bing-Fyn High. al., 1986).
The seismic data show no clear asymmetric sedi- Both seismic and well data show a pronounced
mentation patterns in the Danish Central Trough, ex- increase in thickness towards the east, suggesting
cept for halokinetically induced sedimentation, pro- that sedimentation was controlled by fault activity
files 1-2 (Fig. 2). This is in contrast to the rapid along the Coffee Soil Fault. This feature has also
and asymmetric subsidence in the nearby, but differ- been observed in the northern S0gne Basin (Gowers
ently orientated, Horn Graben. The thinning of the and Saeb0e, 1985). The Middle Jurassic sequence is,
TYiassic sequence towards the north and northwest thus, the oldest clear syn-rift, wedge-shaped deposit
in the Danish TYough area is probably the result of that is recognized in the Danish Central TYough.
Middle-Late Jurassic erosion.
The halokinetically induced rim-synclines are Late Jurassic
clearly seen in the S0gne Basin and also identi- During the Late Jurassic, marine conditions pre-
fied in the Tail End Graben (Fig. 2 and Mogensen, vailed in the Central TYough region, caused by the
1988). Minor fault activity in the TYiassic probably eustatic sea-level rise and increased subsidence. Up
triggered the salt movements. to 4000 m of clay-dominated sediments were de-
The thickness and facies distribution of the TYiassic posited to the east, in the Tail End Graben. In the
sediments indicate that the Late Permian structural Feda and Gertrud Grabens more than 2000 m were
framework seems to have continued through the deposited and approximately 600 m seems to be
THassic. The trough area probably acted as an inter- present in the Grensen Nose and Outer Rough Basin
basinal barrier together with the Mid North Sea (Fig. 7; M0ller, 1986; Michelsen et al., 1987).
High-Ringk0bing-Fyn High trend, separating the In the Salt Dome Province-TM End Graben-
northern and southern TYiassic facies provinces. S0gne Basin trend there seems to have been con-
tinuous sedimentation from the Callovian to the
Lower Jurassic earliest Early Cretaceous. The age distribution of the
The Lower Jurassic deposits are only drilled in the formations indicates a transgression from the north
Salt Dome Province, in the Edna-1, Jens-1, M-8, O-l through this basinal trend during the late Middle
and U-l wells (Fig. 1), but it is generally accepted Jurassic (Michelsen et al., 1987; fig. 3). The present
that the Early Jurassic sea covered most of the North study, furthermore, indicates that during the Late
Sea area (see also Gowers and Saeb0e, 1985). Jurassic the transgression continued westwards onto
The drilled sequence is referred to the lower the Heno Plateau and Feda Graben areas.
part of the Fjerritslev Formation, as known from The age distribution of sediments subcropping
North Jylland (Michelsen, 1978). Lithostratigraphic the Middle Jurassic stratigraphic level supports this
and biostratigraphic correlations show that sedimen- model (Figs. 8 and 9).
tation rates may have been the same over broad (1) To the south, in the Salt Dome Province, the
areas. Middle Jurassic is separated from the Lower Jurassic
Seismic data indicate that the Lower Jurassic sed- sediments by a hiatus corresponding to the Pliens-
iments are present in the southern part of the Tail bachian and Toarcian stages. The Lower Jurassic
End Graben, while they are absent in the S0gne overlies Upper TYiassic sediments.
Basin. (2) In the S0gne Basin the Middle Jurassic rests on
The drilled Lower Jurassic sequence is truncated lower Middle to Lower TYiassic sediments. According
so that the Upper Sinemurian beds are the youngest to the seismic data this also seems to be the case in
remaining (Michelsen, 1978). The erosion was prob- parts of the Feda Graben and Gertrud Graben.
ably caused by the central North Sea uplift, as de- (3) At the eastern margin of the Heno Plateau,
scribed by Ziegler (1982). The overlying sediments the Oxfordian Lola Formation rests on lower TYiassic
are dated to the Bajocian to Callovian stages (Jensen sediments.
et al., 1986). The erosional event possibly also af- (4) Farther westwards on the Heno Plateau, the
fected the TYiassic sequence in the northern and late Oxfordian to early Kimmeridgian Heno Forma-
northwestern Danish Central TYough. tion overlies Zechstein or Rotliegende sediments.
502 O. Michels en, T.E. Mge ns en and JA. Korstgrd

Fig. 7. Thickness and distribution of the seismic sequence corresponding to the Middle and Upper Jurassic section. From M0ller (1986).

(5) In the Gert Ridge area, north of the Danish Central TYough area during the early Middle
Heno Plateau, the Oxfordian basal sand rests on Jurassic uplift and the initial rifting phase (Fig. 9).
Rotliegende sediments. Continued rifting, causing subdivision of the trough
(6) To the north, in the Norwegian sector, the Up- area into smaller halfgrabens, must have taken place
per Jurassic rests on Rotliegende at the Mandal High after this erosion phase. Subdivision of the trough
and the uppermost Jurassic Farsund Formation over- area into the Feda Graben and Gertrud Graben oc-
lies Zechstein carbonates at the Mid North Sea High. curred in earliest Kimmeridgian, after deposition of
This subcrop pattern may be explained by pro- Oxfordian basal sand, which is drilled in the Gert-1
gressively deeper erosion from east to west in the well (Fig. 6). Seismic profiles (Fig. 2) show that the
Pre-Cretaceous structural development of the Danish Trough and its implications for the distribution of Jurassic sands 5

SUBCROP MIDDLE JURASSIC LEVEL NW-N Rotliegende grain. This caused a separation
of the S0gne Basin and the Tail End Graben and
footwall uplift of the Mandal High, the southern
S0gne Basin and the Ringk0bing-Fyn High, inducing
complex structures at the triple point between the
S0gne Basin, the Ringk0bing-Fyn High and the Tail
End Graben (Fig. 10). The separation of the S0gne
Basin and the Tkil End Graben and the subsequent
starvation of the S0gne Basin, is clearly seen from
the distribution of Cretaceous deposits covering the
Legend
:
Late Cimmerian Unconformity surface (Fig. 11).
C = Chalk Grou p
Fa = Farsun d Fm.
H = Heno Fm. Middle Late Jurassic paleogeography
L = Lola f m.
J B= Basa l Sand and related sand deposits
Jm = M i d del J u r a s cs i
F = Fjerritsle v F m.
T = Triassi c
Tm = M i d del Triassi c
The structural preconditions for reconstructing the
Tl = Lowe r Triassi
Z - Z e c h s tne i
c Middle and Late Jurassic paleogeography are de-
R = Rotliegend
Pc = Precambria
e
n
scribed above. The early Middle Jurassic erosional
Fig. 8. Age and distribution of sediments subcropping the base
surface is documented by well and seismic data. In
Middle Jurassic stratigraphic level, indicating the well locations. the southeastern Danish Central TYough the ero-
sion cut down to the middle of the Lower Jurassic
sequence, whereas in the northwest the erosion trun-
Inge and Mads Highs were part of the Mid North cates Paleozoic rocks.
Sea High until Late Jurassic times. The broadening
of the Central Ttough through the time-span from Middle Jurassic
the late Middle Jurassic to the Early Cretaceous, Following the early Middle Jurassic uplift and the
indicates slow stretching according to the stretching subsequent erosion fault activity is recognized in the
model of McKenzie (1978). Slow stretching causes eastern Danish Central Ttough, especially along the
thermal recovery in the lithosphere and subsequent Coffee Soil Fault. Fluvio-deltaic sediments were de-
strain hardening, widening of the rift system and posited in the eastern part of this major halfgraben
finally abortion of the rift. system (Fig. 12). The environment became increas-
The WNW-NW orientation of Middle to Late ingly marine in the course of time. The age distri-
Jurassic faults cuts to a certain degree the older bution recognized in the Norwegian, Danish and

PA LI N S P A S T I C PROFILE

S0GNE BASIN
MID NORTH SEA
I DOGGER HIGH
-4 HENO PLATEAU h ~ TAIL END GRABEN
SALT DOME PROVINCE
Fig. 9. An approximately east-west palinspastic profile of the Danish Central Trough, including the Permian to lowermost Lower
Cretaceous sequences. Note the major hiatus at the base Middle Jurassic stratigraphic level.
504 O. Michehen, T.E. Mogensen and JA. Korstgrd

marine conditions prevailed with deposition of shelf

muds (Lola Formation), which are also found to
the west on the Heno Plateau (Fig. 9). Farther
to the northwest, in the Gert Ridge area, coarse-
grained beach and lower delta plain deposits are
found (the basal sand, Fig. 6). The lower part of
the basal sand seems to be time-equivalent with the
shelf muds (Lola Formation) on the Heno Plateau,
whereas the upper part is time-equivalent with the
Heno Formation mentioned below, which consists of
shoreface and offshore bar sediments (Fig. 12). This
distribution of environments indicates that the sea
transgressed westwards from the eastern part of the
-"" Major Mid to Late Jurassic fault
M M Uplifting of Mandal High and Ringkobing-Fyn High
trough area during the earliest Oxfordian, whereas
11111 11 Uplifting of Segne Basin
:
an apparent regression caused the coast line to mi-
1^; ';;^ Down faulted part of Mandal High
1::::::::::::::::::::1 Major Mid to Late Jurassic half graben
grate eastward again during the latest Oxfordian.
Threshold area
k ^ ^ j Intermediate area between Sogen Basin / Tail End
^ ^ ^ Graben and Ringkobing-Fyn High Kimmeridgian
The subdivision of the northwestern trough area
Fig. 10. The mid to Late Jurassic structural framework of the into the Feda and Gertrud Grabens occurred in the
northeastern Danish Central Trough. From Mogensen (1988). earliest Kimmeridgian. Shelf muds (Lola Formation)
were deposited in this area as well as in the eastern
and southern parts of the Central Ttough region. On
the Heno Plateau, shoreface and offshore coarse-
grained sediments (Heno Formation) were deposited
in the early Kimmeridgian (Fig. 12). Over a short
period, these latter environments prograded into the
newly established Feda and Gertrud Grabens (see
the Gert-1 and Karl-1 wells, Fig. 6) and farther into
the Norwegian sector (Bergan et al., 1989). During
the latest Kimmeridgian, rapid subsidence took place
in the Tail End, Gertrud and Feda Grabens and shelf
muds (Farsund Formation) were deposited in the
entire trough region.

Volgian
The rate of subsidence in the grabens increased
in the Volgian and a thick sequence of shelf mud
(Farsund Formation) was deposited in the entire
trough region (Fig. 12). Locally, coarse-grained sed-
iments deposited by density flows (Poul Formation)
Fig. 11. Distribution of Lower and Upper Cretaceous deposits
interfinger with the Farsund Formation. At the end
overlying the late Cimmerian unconformity. From Mogensen
(1988). of the period the Inge and Mads Highs were prob-
ably separated from the Mid North Sea High and
the Outer Rough Basin and the Grensen Nose were
Dutch sectors indicates that the transgression came established. Elevation of the Mandal High and star-
from the north during the Callovian (Michelsen et vation of the S0gne Basin seems to have occurred at
al., 1987). the same time.

Oxfordian Summary
The fault activity along the Coffee Soil Fault sys-
tem continued into the Oxfordian. Farther west- This study has shown that the present Danish
wards, the faults delimiting the younger Feda Graben TYough has a more differentiated structural devel-
and Gertrud Graben were possibly initiated at the opment, than is generally accepted in the literature.
end of this period. In the eastern Central TYough The main conclusion is that graben development, as
Pre-Cretaceous structural development of the Danish Trough and its implications for the distribution of Jurassic sands 505

Middle Kimmeridgian L. Kimmeridgian - Volgian

palaeogeography palaeogeography

Legend
:

CO Land
Sea
ED Sandy facie
s

Fig. 12. Middle to Late Jurassic paleogeographic maps of the Danish Central Trough.

reflected by fault controlled, asymmetric sedimenta- Bryne Formation) were deposited in the eastern part
tion, started in the Middle Jurassic and continued of this halfgraben. Caused by a transgression coming
throughout the Late Jurassic. The dynamic basin from the north, the environment became increasingly
development is summarized below. marine in the course of time.
(1) The trough region and adjacent areas, were (3) During the Oxfordian, continued transgression
uplifted in early Middle Jurassic times. A major half- led to deposition of the clay-dominated Lola For-
graben, the Tkil End Graben and the adjacent areas mation, which also covers the Heno Plateau to the
to the west, developed along the Coffee Soil Fault, west. Farther northwestwards, updip the halfgraben
bordering the Ringk0bing-Fyn High. This represents structure, coarse-grained beach and lower delta plain
the oldest documented phase of graben develop- sediments (the basal sand) were deposited.
ment. The area to the west was deeply eroded during (4) During the latest Oxfordian, a short-lived re-
the Middle Jurassic. gression caused an eastward migration of the near
(2) Simultaneously with the halfgraben develop- shore sedimentation, and the Heno Formation cov-
ment fluvio-deltaic, sand-dominated sediments (the ered the entire Heno Plateau.
506 O. Michelsen, T.E. Mogensen and JA. Korstgrd

(5) High subsidence rates prevailed in the TM1 Glennie, K.W., 1986. Development of N.W. Europe's Southern
End Graben throughout the Late Jurassic. The Feda Permian Gas Basin. In: J. Brooks, J.C. Goof and B. van
Hoorn (Editors), Habitat of Paleozoic Gas in N.W. Europe.
Graben and Gertrud Graben started to develop in
Geol. Soc. London, Spec. Publ, 23: 3-22.
earliest Kimmeridgian and shelf muds (Lola For- Gowers, M.B. and Sasb0e, A., 1985. On the structural evolution
mation) were deposited in these areas. Deposition of the Central Trough in the Norwegian and Danish sectors
of more coarse-grained sediments continued on the of the North Sea. Mar. Pet. Geol., 2: 298-318.
Heno Plateau. Over a short period these latter shal- Jacobsen, E, 1982. Triassic. In: O. Michelsen (Editor), Geology
of the Danish Central Graben. Danm. Geol. Unders., Ser. B,
low marine facies prograded into the Feda Graben
8: 32-36.
and the Gertrud Graben and farther into the Norwe- Jensen, T.E, Holm, L., Frandsen, N. and Michelsen, O., 1986.
gian sector. Jurassic-Lower Cretaceous lithostratigraphic nomenclature
(6) During the latest Kimmeridgian and the Vol- for the Danish Central Trough. Danm. Geol. Unders., Ser. A,
gian, rapid subsidence took place in the entire Cen- 12, 65 pp.
Kusznir, N.J., Marsden, G. and Egan, S., 1988. Fault block
tral Though region and shelf muds (the Farsund
rotation during continental lithosphere extension: a flexural
Formation) were deposited. Locally, coarse-grained cantilever model. Geophys. J., 92: 546 (abstract).
sediments deposited by density flows intermix the McKenzie, D., 1978. Some remarks on the development of
Farsund Formation. The Grensen Nose and Outer sedimentary basins. Earth Planet. Sei. Lett., 40: 25-32.
Rough Basin developed late in the Late Jurassic. The Michelsen, O., 1978. The Lower Jurassic of the Danish Nords0
O-l boring, Central Trough, Danm. Geol. Unders., Arbog
broadening of the Central Though was probably the
1977: 77-87.
result of slow rifting. Starvation of the S0gne Basin Michelsen, O. and Andersen, G, 1983. Mesozoic structural
and footwall uplift of the Mandal High took place. and sedimentary development of the Danish Central Graben.
The Inge High and Mads High seem to have been Geol. Mijnbouw, 62: 93-102.
parts of the Mid North Sea High until this time. Michelsen, O., Frandsen, N., Holm, L., Jensen, T.F., M0ller,
J.J. and Vejbask, O.V., 1987. Jurassic-Lower Cretaceous of
the Danish Central Trough depositional environments,
References tectonism, and reservoirs. Danm. Geol. Unders., Ser. A, 16,
45 pp.
Bergan, M , T0rudbakken, B. and Wands, B., 1989. Strati- Mogensen, T.E., 1988. Permo-Jurassisk strukturel udvikling af
graphic correlation of Upper Jurassic sandstones within the Central Trget i Graenseomrdet mellem Danmark og
Norwegian Central Graben: sedimentological and tectonic Norge. Masters Thesis, Department of Geology, Aarhus
implications. In: J.D. Collinson (Editor), Correlation in Hy- University, Aarhus.
drocarbon Exploration. Graham and Trotman, London, pp. M0ller, J.J., 1986. Seismic structural mapping of the Middle and
243-251. Upper Jurassic in the Danish Central Trough. Danm. Geol.
Cartwright, J.A., 1987. Transverse structural zones in continental Unders., Ser. A, 13, 40 pp.
rifts an example from the Danish Sector of the North Sea. Pegrum, R.M., 1984. The extension of the Tornquist Zone in the
In: J. Brooks and K.W. Glennie (Editors), Petroleum Geology Norwegian North Sea. Nor. Geol. Tidsskr., 64: 39-68.
of North West Europe, Vol. 1. Graham and Trotman, London, Vejbaek, O.V., 1986. Seismic stratigraphy and tectonic evolution
pp. 441-452. of the Lower Cretaceous in the Danish Central Trough.
Frandsen, N., 1986. Middle Jurassic deltaic and coastal deposits Danm. Geol. Unders., Ser. A, 11, 46 pp.
in the Lulu-1 well of the Danish Central Trough. Danm. Geol. Ziegler, P.A., 1982. Geological Atlas of Western and Central
Unders., Ser. A, 9, 23 pp. Europe. Elsevier, Amsterdam, 107 pp.
Glennie, K.W., 1984. Introduction to the Petroleum Geology Ziegler, P.A., 1987. Late Cretaceous and cenozoic Intraplate
of the North Sea. Blackwell Scientific Publications, Oxford, compressional deformations in the Alpine foreland a
236 pp. geodynamic model. Tectonophysics, 137: 389-420.

O. MICHELSEN Geological Survey of Denmark, Copenhagen, Denmark

Present address: Department of Geology, University ofAarhus, DK-8000 Arhus C, Denmark
T.E. MOGENSEN Aarhus University, Arhus, Denmark
J.A. KORSTGRD Geologisk Institut, Aarhus Universitet, DK-8000 Arhus C, Denmark
 507

Small faults in sandstones from Spitsbergen and

Haltenbanken. A study of diagenetic and deformational
structures and their relation to fluid flow
E. Sverdrup and K. Bj0rlykke

A study of medium to small scale faults has been carried out on outcrops of Cretaceous sandstones from eastern Spitsbergen
(Kvalvagen) and on Jurassic reservoir sandstones from Haltenbanken, offshore mid-Norway. Structural and textural evidences
from the faults indicate that they formed at shallow depth when the sand was uncemented. At Kvalvagen the faulting can
be demonstrated to be syn-sedimentary, and formed during the progradation of a deltaic sequence. Rapid fluid escape along
developing faults gave rise to deformation zones with fault parallel laminae. The high frequency of fault structures observed
in cores from the reservoir sandstones of the Njord Field, Haltenbanken, formed at shallow burial depth. This is inferred by
the fact that feldspar dissolution and precipitation of authigenic kaolinite, which probably relates to meteoric water flushing,
post-date movements on the faults.
Enhanced pressure solution and stylolitization have developed on the fault structures at both localities at greater burial
depths, due to enrichment of clay minerals and micas along the fault planes when they formed. Steeply dipping stylolites, such
as those observed in some Haltenbanken reservoir sandstones, can therefore be explained by selective pressure solution along
fault planes due to overburden stress, and need not imply horizontal tectonic stress.
Permeability measurements of the fault planes show that they are important barriers to horizontal fluid flow and may
significantly influence the reservoir properties.

Introduction seismic sections. Small-scale faults found in cores,

however, are more difficult to assess. By relating
Permeability reduction in reservoir sandstones movements along such faults to a sequence of diage-
by faults has been observed by several workers netic reactions, a relative timing of fault movements
(e.g., Pittman, 1981; Nelson, 1985; Gabrielsen and and later deformation along fault planes can be ob-
Koestler, 1987), and is generally closely related to tained. The aims of this paper are to exemplify how
deformational textures such as dense physical pack- fault structures formed at shallow burial depth might
ing of grains, clay enrichment, grain size reduction become diagenetically altered during late burial, to
by grain breakage, brecciation and/or fault related demonstrate how small-scale faults in sandstones af-
mineralization. In addition, the primary texture and fect permeability, and to explain the development of
mineralogy of the faults might be diagenetically al- steeply inclined (fault-parallel) stylolites.
tered during deep burial. To predict the effect of
faults in sandstones, whether they act as barriers or
Syn-sedimentary faulting
not, as well as their geometrical distribution within
in lower Cretaceous deltaic sandstones
a reservoir, it is important to distinguish between
at Kvalvagen, Eastern Spitsbergen
early (pre-lithification) faults and tectonic defor-
mation structures generated at deeper burial. The The section at Kvalvagen, Eastern Spitsbergen,
picture is, however, generally diffuse where syn- Arctic Norway (Fig. 1) displays syn-sedimentary
sedimentary and shallow generated tectonic faults faults of Lower Cretaceous age. Detailed descrip-
are to be distinguished, as such deformational depths tions of the geological setting and deformational
tend to produce similar textures. history are given in Nemec et al. (1988) and Sver-
In outcropping sedimentary sequences the relative drup and Prestholm (1990).
timing of fault displacement can be inferred from The cliff exposure shows a strike section through
field data. In buried sedimentary basins large scale an upper Jurassic deltaic system, which represents
faults can be interpreted, and partly dated, from the distal development of the Helvetiafjellet and

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 507-517. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
508 E. Sverdrup and K. Bj0rlykke

mouth bar, lagoon and barrier bar deposits (Fig. 2).

SVALBARD iir
Syn-sedimentary faulting in the distributary chan-
nel sandstones and the upper part of the Janusfjellet
Formation is interpreted to be due to the collapse
of the prograding delta front (Nemec et al., 1988).
The faults cutting through channel sandstones do not
display discrete slip-planes but appear as sandstone-
dominated fault zones (Sverdrup and Prestholm,
1990). Vertical throws on the faults fall in the range
of 3-35 m, their dips vary between 40 and 60 and
widths of the fault zones are between 5 and 40 cm
(Fig. 3). The fault zone rims are sub-parallel and
the boundary to the adjacent channel sandstones lo-
cally exhibits a lithological or structural contrast, or
a dark, clay/micaceous seam (Fig. 4). The fault zones
Fig. 1. Location map Kvalvgen, Spitsbergen, Norway. are characterized by fault-parallel laminae (Fig. 5).
Small-scale normal faults occur within 2 m of the
major fault zones (Fig. 3). One set of such small-
Janusfjellet Formations (Fig. 2). Open marine shale scale faults is located in the hanging wall block,
and thin sandstone beds (the Janusfjellet Formation) is synthetic to the major faults, shows considerable
are overlain by afiuviodeltaicsuccession (the Helve- continuity and throws in the order of 1-2 cm. An
tiafjellet Formation). The lower part of the Helveti- other set of small-scale faults, located in the footwall
afjellet Formation consists of a sandstone unit some blocks, is both syn- and antithetic to the major faults
20 m thick, which represents a distributary channel (conjugated) with induced millimetric to centimetric
fill sequence. The channel sandstone is medium- to offsets. The small-scale fault structures show no sys-
coarse-grained and finely laminated. Subordinates tematic increase of density towards the major fault
are thin siltstone layers and beds of coarser, locally zones.
pebbly material. The channel sandstone is overlain by The channel sandstone has a very low porosity

Fig. 2. Collapsed channel sandstones of the Helvetiafjellet Formation are draped by mouth bar, lagune and barrier bar sediments.
Small faults in sandstones from Spitsbergen and Haltenbanken 509

Variations in grain-size and matrix content define the

fault-parallel lamination within the fault zones.
The fault zones probably generated from chan-
nel sand which got fluidized and remobilized due
to water flushing through developing faults during
the delta collapse (Sverdrup and Prestholm, 1990).
Variable grain size and mineralogical composition of
the fluidized material caused separation of particles
which constituted the fault zone laminae. As no sec-
ond generation faults are observed to cut through
the described fault zones, the process is suggested
to have occurred as a rapid, single-phase injection.
It has not been determined whether the small-scale
faults represent water-escape structures or ordinary
faults.
Thin section studies of the fault zones show no
clear evidence of a relative denser physical packing
NB! NOT TO SCALE
or preferred alignments of grains here compared to
Fig. 3. Idealized sketch of the fault structures at Kvalvgen. the adjacent, undeformed sandstone, which usually
characterizes syn-sedimentary faulting (e.g., Allen,
1982). Furthermore, as expected, there is no cata-
measured from thin sections (0-5%). Quartz-grains clasis of either quartz or feldspars within the fault
and cement comprise 70-90% of the rock. Feldspars zones. The fault structures could, therefore, easily be
(1-5%) and lamina of clay minerals and micas make overlooked in core and/or thin section studies alone.
up the main compositional variations in the sand- The post-faulting textural development of the fault
stones. zones is explained on page 515.
By examining grain size and mineralogical com-
position of the fault zones it has become clear that
the material herein is derived from the channel sand- Early diagenetic faulting in reservoir
stones. No porosity difference exists between the sandstones at Haltenbanken
undeformed sandstones and the fault zones. In the
microscopic scale the fault zone rim locally appears The rocks from the Njord Field at Haltenbanken
as a distinct planar feature of a few grain-diameters are of Middle Jurassic age and have been inter-
enriched in clay and micaceous material (Fig. 6). preted as shallow marine deposits (Gjelberg et al.,

Fig. 4. Photograph of a typical fault zone at Kvalvgen. Notice the sharp rim and the discordant character of the fault zone.
510 E. Sverdrup and K. Bj0rlykke

Sandstones are from depths around 3000 m be-

low sea level, which correspond to their maximum
depth of burial. They are homogeneous or faintly
laminated, very-fine to fine-grained. Porosities of the
sandstones are in the 15-25% range. K-feldspar (5-
15%) and muscovite (0-15%) are abundant in the
rocks. Matrix, when present, is composed of fine-
grained clay minerals (kaolinite, illite) and organic
material.
Faults displayed in cores appear as numerous
small-scale structures concentrated within certain in-
tervals (Fig. 7). Dark color contrasts commonly make
the traces of individual faults visible. The fault struc-
tures are single or multiple and the individual fault
planes are linear or curved (Fig. 7). Multiple fault
planes are either conjugated or sub-parallel. Inter-
weaving fault planes make up anastomosing zones.
The displacements of individual faults are on the
order of 0.5-2.0 cm, and on the microscopic scale
Fig. 5. Intra fault zone laminations appear as dark seams. A few they appear as micro-fault zones with widths between
small-scale normal faults which offset the primary sedimentary 0.1 and 5.0 mm. Porosity is considerably lower within
laminae in the channel sandstone are also displayed (scale in
these micro-fault zones (2-7%) compared to the ad-
cm.).
jacent, undeformed sandstones (Fig. 8a). The faults
are characterized by dense physical packing and pre-
1987). Seismically defined faults, which probably re- ferred orientation of grains and enrichment of matrix
late to movement along the western boundary of the material. Locally, grain size reduction, due to me-
Halten Terrace (Kristiansund-Bod0 Fault Complex) chanically deformed micas and pressure solution of
during the Late Jurassic to Early Cretaceous time quartz, is present (Fig. 8b).
(Gabrielsen and Robinson, 1984) cut through the The absence of mechanically deformed quartz and
sedimentary sequence. feldspar grains suggests that the faults formed rel-

, RIM OF .
FAULT ZONE / FAULT /
FAULT-PARALLEL
HiCROSTYLOLiTES
/ Z0NF /

2mm
/ /
Fig. 6. Photograph (microscale) of the fault zone rim. The rim is defined by a clay seam. The sandstone within the fault zone at this
particular position has a lower clay content than the undeformed channel sandstone immediately outside. Notice the fault parallel
microstylolites along the rim.
Small faults in sandstones from Spitsbergen and Haltenbanken 511

~--2r-

Fig. 7. Small-scale fault structures as they appear in the cores from the Njord Field, Haltenbanken, Offshore Mid-Norway (see text for
further descriptions).

atively shallow, prior to any extensive cementation The two different generations of calcite cement
of the sandstone. This is in accordance with Malt- have been analysed with respect to oxygen and car-
man (1984) who pointed out that intragranular de- bon isotopes (Fig. 11). The early, "pre-faulting" cal-
formation dominates in a lithified material due to cite cement generally displays heavier <S180(PDB) val-
compaction and cementation, and which also has ues than the late "post-faulting" calcite cement which
been visualized through Knipe's (1986) "Deforma- is rather negative. Early cement with <$180(PDB) >
tion Mechanism Path Diagrams". The grain defor- 3%o and <512C(PDB) < 8%o probably reflects pre-
mation during faulting was characterized by inde- cipitation from marine water in the oxydized zone
pendent paniculate flow (Borradaile, 1981) where (<5180(SMOW) = -2%o) at low temperatures (T =
individual grains, except from micas, remained unde- 15-20C). Early cements with more negative values
formed. Relative "enrichment" of clay minerals and (-4%o > <5180(PDB) > -moo) and 613C values be-
micas in the micro-fault zone is explained by vol- tween -10 and -20 could have precipitated from
ume loss caused by the collapse of the primary grain meteoric water in the sulphate reducing zone (Ir-
framework and by clay smear of primary sandstone win et al., 1977). Assuming an average oxygene
laminae. The deformation also produced preferential isotopic composition for this cement at -6%o, the
alignment of flaky and elongated grains. <$180(SMOW) value for the water at the time of pre-
The movement on the faults can be dated relatively cipitation (Jurassic time) was in the range between
to two phases of carbonate cementation, feldspar dis- - 5 and -6%o (T = 12-15C). The oxygene iso-
solution and pressure solution of quartz. Two gener- tope values of the late calcite cement are generally
ations of carbonate cement (pre- and post-faulting) more negative (<S180(PDB) = -10.5-14.0%o). By as-
have been distinguished in the sandstones. Early, mi- suming an isotopic composition for the pore water at
crosparry, ferroan calcite occurs within nodules (con- (5180(SMOW) = 3%o, which is close to the present
cretions) and thin layers in fine-grained argillaceous day formation water composition, the corresponding
intervals in the cored sequence. The concretions are crystallization temperature becomes 50-77C. This
not cut by faults. The calcite cemented layers, how- is consistent with values obtained (and discussed
ever, exhibit brittle deformation where cut by faults further) by Bj0rlykke et al. (1986) and Saigal and
(Fig. 9). These relations strongly suggest that this Bj0rlykke (1987).
calcite cement precipitated prior to the faulting pro- Moulds of feldspar partly filled with kaolinite are
cess. In certain zones of the sandstones, poikilotopic common in the sandstones and are also found within
non-ferroan calcite cement is present. Petrographical the micro-fault zones (Fig. 12). Since this texture
relations suggest that this cementation took place has not been deformed, the displacement on the
after the faulting. The cement shows no brittle de- faults must pre-date the feldspar dissolution, which
formation (fracturing, etc.) adjacent to the faults and in the Haltenbanken area is interpreted to be due to
the cement has not precipitated within the densely meteoric water flushing at a relatively shallow burial
packed, low-porosity micro-fault zones (Fig. 10). depth (Bj0rlykke et al., 1986).
512 E. Sverdrup and K. Bj0rlykke

Undeformed sandstone

Fig. 8. Top: SEM picture showing the grain arrangements, the porosity loss and the enrichment of clay minerals and fine-grained micas
within the micro-fault zone (right). The left side of the picture shows the undeformed sandstone texture. Bottom: SEM picture of a
micro-fault zone where pressure solutions of quartz and microstylolites are well developed. Note also the oversized, secondary pores
outside the fault generated from dissolved feldspars.

Quartz overgrowth occurs at a great burial depth, genetic phenomenon in this area.
and sandstones at Haltenbanken buried less than The textural development of the fault structures
2.5-3 km deep generally display a low quartz cement in relation to the burial and diagenetic history of
content (Bj0rlykke et al., 1986). Furthermore, sty- the sandstones is schematically illustrated in Fig. 13.
lolites are reported to be present in Haltenbanken The two processes: faulting and diagenesis, are both
sandstones at burial depths greater than 2.5 km, demonstrated as important factors in the textural
which supports quartz dissolution being a late dia- development of the fault structures.
Small faults in sandstones from Spitsbergen and Haltenbanken 513

-5
t i i t i i * ) i i nt > i i
/ \
/
Late | < ol.
carbonate \ /
cement \ \ i t-10
\ y
\ i
\ II
Early\ |
diagenetic ' -20
meteoric \ _ |
Carbonate 1

t-30
Early diagenetic\
marine Carbonate

40
Vfi/Tf 2Z

13 C (PDB)
/.'oV

Fig. 11. Isotope data for different generations of carbonate

cements from the Njord Field. Early cement which predates
Cm faulting (black squares) is characterized by low 180 values.
The marine cement is formed at the sea floor in the sulphate
reducing zone. Late cement which postdate faulting (open
Fig. 9. Structural relationships between the early calcite ce- circles) is formed at higher temperatures and greater burial
mented nodules (a-c) and layers (d). depths as indicated by the more negative 8lsO values.

(a)

cm

IMAGE ANALYSIS OF
SECTION OUTLINED IN
\ FIG. 10a
white color =
Late calcite cem

Fig. 10. Textural relationship between the late calcite cement and a fault plane.
514 E. Sverdrup and K. Bj0rlykke

iiiitiiiWr^tI
iiliMIlllllllIfJT'I
f -JHif, 1

-*&*" * "*;

s
f.

Fig. 12. Secondary porosity within a micro-fault zone due to dissolved feldspar. The mould is partly filled with authigenic kaolinite
which indicates that fault movement pre-dates feldspar dissolution and kaolinite precipitation.

0 T DEPOSITION 1 TEXTURAL IMPLICATIONS

MARIN
E
| \ FAULT- ~]
0
Early calcite /-v / " ~ \ /-J

FAULTING

PreK
00 (pre-lithification)
00 o
LU
Z tx
LU
BURIAL
< * Feldspar
UNDEF0RM
DE K
\ \l
DEPTH
Q
,_, dissolution
> o

CC
early
L
u

1000m(?)t
Diagenetic
\/
<>
modifications
T
00
00 Late calcite
LU < late
z Quartz dissolution
LU
O o
< Q. Stylolitization
O
J
<
cr:
D a.
cD ro
3000 m

QUARTZ FELDSPAR MICA CLAY M. KAOLINITE Notice both horizontal and vertical
stylolites in fig. 13c.
Fig. 13. Composite diagram showing the textural development of a sandstone containing a fault structure during the burial (and
diagenetic) history.
SmaU faults in sandstones from Spitsbergen and Haltenbanken 515

Diagenesis superimposed on fault textures 2

In spite of a shallow deformational depths, the
fault structures at Kvalvgen and Haltenbanken
show great dissimilarity with respect to fault tex-
tures at present. This is partly due to the different
CTn >
deformational processes, but is mainly explained by
differences in the post-faulting diagenetic history in
the two areas. In the Njord Field, the fault zones
are relatively tight with porosities in the lower edge ^"n

of 5%, compared to an average of about 20% in

the surrounding sandstone. This is caused by dense
packing of grains and subsequent pressure solu-
tion. At Kvalvgen, the sediments have experienced
deeper burial (approximately 3500 m; Sverdrup and
Prestholm, 1990), and extensive quartz cementation Fig. 14. Stress relationships on a dipping plane. = vertical
has reduced the porosities to generally less than stress component; ^ = horizontal stress component; = stress
component perpendicular to plane; = dip of plane.
5% both in the fault zones as well as in the unde-
formed sandstone. Consequently, any differences in
the early, post-deformational packing arrangements component. The normal stress component on any
within the fault zones and the channel sands have plane is given by the relation:
disappeared and are not observed. A common fea-
ture at both localities is the presence of stylolites
along the fault structures.
where represents the dip of the plane (Fig. 14).
Pressure solution and stylolitization From this equation can be calculated for any
dipping surfaces. For = 60 the normal stress com-
The most prominent diagenetic \ -esses which ponent () equals 12.5 MPa. This value corresponds
have affected the fault structures at both localities, to a vertical normal stress at a depth close to 1000 m,
are pressure solution and stylolitization. Stylolites with normal pore pressure. As intergranular pressure
are developed along primary laminae as well as fault solution has been observed in sandstones at depths
surfaces (Figs. 6 and 8b). of as little as 900 m (Wilson and Sibley, 1978), and
Stylolites are generally considered to be formed reports from depth of 1000 to 2000 m are common
perpendicular to the maximum principal stress com- (e.g., Thompson, 1959; Nagtegaal, 1978), the stress
ponent (e.g., Bchner, 1981; Nelson, 1981). In a value of 12.5 MPa should be sufficiently high for
subsiding basin, chemical compaction is induced by the pressure solution to develop taking into account
the vertical lithostatic stress which may produce hor- the temperature at 2800 m. The situation described
izontal (layer-parallel) stylolites. Vertical or steeply above, as for the Kvalvgen example, might be visu-
dipping stylolites are commonly interpreted in terms alized in a "Deformational diagram" (Rutter, 1976)
of horizontal (tectonically induced) compression in a (Fig. 15). Here, pressure solution can be demon-
basin (e.g., Bj0rlykke et al., 1986; Mitra, 1988). strated to be an active process along the fault planes
Sverdrup and Prestholm (1990) have shown that in both localities. The stylolitization observed, taking
the stress component of the vertical, lithological place along the inclined, clay/mica-rich fault sur-
stress acting perpendicular to fault surfaces, may be faces, has probably evolved as such inhomogeneities
sufficient for stylolites to generate when tempera- in sandstones tend to act as precursors for dissolu-
tures are favourable. A calculation performed for the tion of quartz (e.g., Heald, 1959; Weyl, 1959; Boles,
Kvalvgen rocks concluded that the stress compo- 1984).
nent perpendicular to a fault dipping 70 equals 18
MPa. Similar calculations can be performed for the
Permeability of faults
rocks at Haltenbanken.
In the Njord Field at Haltenbanken the vertical To investigate the permeability effects of small-
effective stress at 2800 m is close to 25 MPa. In scale faults, such as those described from Hal-
sandstones, the ratio between horizontal and vertical tenbanken, the permeability perpendicular to small-
stress (<jzA7h) varies between 0.2 and 0.4 (Howard scale faults was measured in core plugs and com-
and Fast, 1970), but in general the horizontal com- pared with the permeability of the adjacent, un-
ponent may be taken to be one-third of the vertical deformed sandstones. As also was recognized by
516 E. Sverdrup and K. Bj0rlykke

Grain size = o,2mm plug containing both fault and undeformed sand-
^00 -
stone, respectively, x is the net thickness of the
\ Sir a m rat > E - M
"tight" zone (fault) as a fraction of the plug length.
1 1")0 - By this relationship very low permeability values
f.
~ (in the range between 0.06 mD and 5.28 mD) were
~ \ s t . r in r a t ( K - l r ) obtained for the fault planes. By plotting the perme-
I 100 - ability values against matrix content and grain size
r - in the parent, undeformed sandstones, a log-normal
~~ r
trend appears (Fig. 16). A few values did, however,
)0 -
not fit into this trend. These values relate to faults
o
where pressure solution and stylolitization have been
10 -
r
1 1 1 1 1 1 11 1 1 1 1 1 1 i most extensive, and the faults show lower perme-
)() 70 90 110 l.'K) lr)() 170 190 J10 ,) )
Temp. ( (') abilities than should be expected from their parent
Fig. 15. Deformational diagram constructed from Rutter (1976). sandstone matrix content and grain size.
The black square describes the stress and temperature situation
for a surface dipping 60 at approximately 2800 m depth in
the Njord Field. The open circle represents the situation for a Conclusions
surface dipping 70 at approximately 3500 m depth (Kvalvgen
faults). The plots are both within the pressure solution field (1) The faults at Kvalvgen and the Njord Field
of the diagram. The pressure solution is indicated as an active
process taking place at a slow strain rate (e < 10~ 15 s""1). formed at a shallow burial depths. Fault textures in
both localities have later been modified by diagen-
esis, resulting in loss of or modifications of earlier
Gabrielsen and Koestler (1987) during their study soft-sediment deformational textures.
on small-scale faults from the Ttoll Field, a relative (2) Diagenetic reactions like feldspar dissolution,
permeability reduction commonly characterizes the precipitation of kaolinite, carbonate and quartz ce-
samples containing faults. The method used for per- mentation, can, both texturally and by the use of
meability measurements, however, also admits the geochemical methods, help to date movements on
permeability of the fault planes themselves to be faults.
calculated by the following relationship: (3) Syn-sedimentary or early tectonic faults are
* ^(sst) * ^(sy)
frequently enriched in clay minerals and micas along
Kt(fault) their slip surfaces relative to the surrounding sand-
if(sst) - (1 ~x)K{$y)
stones. Stylolites can evolve along the fault planes at
where if(fauit)> ^(sst) and K^ are the permeability of greater burial depth, purely due to the overburden
the fault, the undeformed sandstone plug and the stress.

PERMEABILITY (mD)

Fig. 16. Log-normal plot of fault permeability versus matrix content (percentage value) above grain size (sorting) of the parent,
undeformed sandstone. There is a positive correlation between increased grain size and increased permeability. High matrix content in
the parent sandstone tend to lower the permeability of the fault. The plots which deviate from the trend belong to faults with the most
extensive pressure solution and stylolitization.
Small faults in sandstones from Spitsbergen and Haltenbanken 517

(4) Stylolitization along low-porosity, early fault K. Glen nie (Editors), Petroleum Geology of North-West
surfaces may cause further reduction in permeabil- Europe. Graham and Trotman, London, pp. 1105-1129.
Heald, M.T., 1959. Significance of stylolites in permeable sand-
ity across the low-permeable structures, because clay
stones. J. Sediment. Petrol., 29: 251-253.
minerals and micas are concentrated as quartz dis- Howard, G.C. and Fast, C.R., 1970. Hydraulic fracturing. Soc.
solves. This is important in moderate to deep buried Pet. Eng., AIME Monogr., 2: 210 pp.
reservoirs affected by faulting, where pressure solu- Irwin, H., Curtis, C. and Coleman, M., 1977. Isotopic evidence
tion and stylolitization are common. for source of diagenetic carbonates formed during burial of
organic rich sediments. Nature, 269: 209-213.
Knipe, R.J., 1986. Deformation mechanism path diagrams for
Acknowledgements sediments undergoing lithification. Geol. Soc. Am., Mem.,
166: 151-160.
Maltman, A.J., 1984. On the term "soft-sediment deformation".
Financial support for this study was provided by
J. Struct. Geol., 6: 589-593.
the Norwegian Research Council for Scientific and Mitra, S., 1988. Effects of deformation mechanisms on reservoir
Applied Research (N.T.N.F.). Norsk Hydro a.s. are potential in Central Appalachian Overthrust Belt. Bull. Geol.
acknowledged for their co-operation and for provid- Soc. Am., 72(5): 536-554.
ing rocks samples from the Njord Field. H. Johansen Nagtegaal, P.J.C., 1978. Sandstone-framework instability as a
function of burial diagenesis. J. Geol. Soc. London, 135:
of the Institutt for Energiteknikk (I.F.E., Kjeller) is
101-105.
thanked for discussions and assistance in connec- Nelson, R.A., 1981. Significance of fracture sets associated with
tion with the Laser sampling procedure for isotope stylolite zones. Am. Assoc. Pet. Geol., Bull., 65(11): 2417-
analysis. 2425.
Nelson, R.A., 1985. Geologic analysis of naturally fractured
reservoirs. Contrib. Pet. Geol. Eng., 1, 279 pp.
References Nemec, W, Steel, R.J., Gjelberg, J., Collinson, J.D., Prestholm,
E. and 0xnevad, I.E., 1988. Anatomy of a collapsed and
Allen, J.R.L., 1982. Sedimentary Structures: Their Character and re-established delta front in the lower Cretaceous of Eastern
Physical Basis. Elsevier, Amsterdam, 663 pp. Spitsbergen: gravitational sliding and sedimentation processes.
Bj0rlykke, K., Aagaard, P., Dypvik, H., Hastings, D.S. and Am. Assoc. Pet. Geol., Bull., 72(4): 454-476.
Harper, A.S., 1986. Diagenesis and reservoir properties of Pittman, E.D., 1981. Effect of fault-related granulation on
Jurassic sandstones from the Haltenbanken area, offshore porosity and permeability of quartz sandstones, Simpson
Mid-Norway. In: A.M. Spencer et al. (Editors), Habitat of Group (Ordovician), Oklahoma. Am. Assoc. Pet. Geol., Bull.,
Hydrocarbons on the Norwegian Continental Shelf. Graham 65: 2381-2387.
and Trotman, London, pp. 275-286. Simpson Group (Ordovician), Oklahoma. Am. Assoc. Pet. Geol.,
Boles, J.R., 1984. Influence of mica surface on pore water pH. Bull., 65: 2381-2387.
Chem. Geol., 43: 303-317. Rutter, E.H., 1976. The kinetics of rock deformation by pressure
Borradaile, G.J., 1981. Particulate flow of rock and the formation solution. Philos. Trans. R. Soc. London, Ser. A, 283: 203-219.
of cleavage. Tectonophysics, 72: 305-321. Saigal, G.C. and Bj0rlykke, K., 1987. Carbonate cements in
Buchner, F., 1981. Rhinegraben: horizontal stylolites indicating clastic reservoir rocks from offshore Norway relation-
stress regimes of earlier stages of rifting. Tectonophysics, 73: ships between isotopic composition, textural development and
113-118. burial depth. In: J.D. Marshall (Editor), Diagenesis of Sedi-
Gabrielsen, R.H. and Robinson, C , 1984. Tectonic inhomo- mentary Sequences. Geol. Soc. London, Spec. Publ. 36: 313-
geneities of the Kristiansund-Bod0 Fault Complex off-shore 324.
Norway. In: A.M. Spencer (Editor), Petroleum Geology of the Sverdrup, E. and Prestholm, E., 1990. Synsedimentary deforma-
North European Continental Margin. Norwegian Petroleum tion structures and their implications for stylolitization during
Society, Graham and Trotman, London, pp. 397-406. deeper burial. Sediment. Geol., 68: 201-210.
Gabrielsen, R.H. and Koestler, A.G., 1987. Description and Thompson, A.F., 1959. Pressure solution and porosity. Soc. Econ.
structural implications of fractures in late Jurassic sandstones Paleontol. Mineral., Spec. Publ., 7: 92-110.
of the Troll Field, northern North Sea. In: A.G. Koestler and Weyl, P.K., 1959. Pressure solution and the force of crystallization
S. Olaussen (Editors), Tectonics and Deposition in Sedimen- a phenomenological theory. J. Geophys. Res., 64: 2001-
tary Basins. N.G.T., 67, 4: 371-381. 2025.
Gjelberg, J., Dreyer, T, H0ie, A., Tjelland, T. and Lilleng, T, Wilson, TV. and Sibley, D.F., 1978. Pressure solution and
1987. Late Triassic to Mid-Jurassic sandbody development on porosity reduction in shallow buried quartz arenite. Am.
the Barents and Mid-Norwegian shelf. In: J. Brooks and Assoc. Pet. Geol., Bull, 62: 2329-2334.

E. SVERDRUP Saga Petroleum A/S, Kj0rbovn. 16, P.O. Box 490, N-1301 Sandvika, Norway
K. BJ0RLYKKE Department of Geology, University of Oslo, P.O. Box 1047 Blindem, N-0316 Oslo 3, Norway
 519

Subparallel faults and horizontal-stress

orientations: an evaluation of in-situ stresses inferred from
elliptical wellbore enlargements

R.L. Dart and H.S. Swolfs

Stress-induced wellbore-enlargement (breakout) data from the Marietta basin in south-central Oklahoma and the Bravo dome
area of the central Texas Panhandle show a possible variation in the orientation of horizontal stress trajectories across areas
bounded by subparallel faults. Regional studies of breakout and hydraulic fracturing-stress data indicate a northeast-southwest-
oriented maximum principal horizontal stress (SHmax) oriented oblique to westnorthwest-striking subparallel faults in these two
study areas. Subtle variation in long-axis orientations of elliptically shaped wellbore enlargements from wells located between and
near the ends of mapped subparallel faults in the Marietta basin and in the Bravo dome area indicates a possible local rotation
of SHmax. From the observed variation in the mean orientations of wellbore enlargements from selected wells, it is inferred that
as much as 52 and 22 of stress trajectory rotation occurs between subparallel faults in the Marietta basin and the Bravo dome
area, respectively. Assuming the areas bounded by these subparallel faults are functioning as stress field discontinuities, factors
affecting local stress trajectory reorientation may include the amount of fault overlap, fault separation and differences in Young's
modulus of the faulted rock units. Another factor contributing to or controlling suspected stress trajectory reorientation might
be the influence of anisotropic rock-fabric properties.

Introduction Breakout-data analysis has become an important

tool in evaluating present-day stress conditions in
Borehole field studies (Cox, 1970; Brown, 1978; areas where data from more traditional stress-
Bell and Gough, 1979, 1982; Gough and Bell, 1981, measurement techniques (earthquake focal mecha-
1982, Hickman et al., 1985; Plumb and Hickman, nisms, hydraulic-fracturing, etc.) may not be avail-
1985; Tfeufel, 1985; Zoback et al, 1985) and the- able. An understanding of the in-situ state of stress is
oretical and laboratory studies simulating borehole essential in assessing seismic hazards, recent crustal
stress conditions (Mastin, 1984; Haimson and Her- deformation and the potential for slip on existing
rick, 1985; Zoback et al., 1985) show that well- faults. In the petroleum industry, knowledge of lo-
bore enlargements known as "breakouts" are stress- cal stress-field conditions can assist in the successful
induced spall zones that typically elongate vertically engineering of borehole drilling and the design of
within the wellbore and are the result of compres- wellbore hydraulic-fracturing treatment (Dart and
sional shear failure associated with horizontal com- Zoback, 1989).
pressive stresses about the wellbore. Breakouts are This study is intended to illustrate that routine
observed to cause the wellbore to be elliptically en- analysis of stress-induced or stress-related borehole
larged in cross section and are typically aligned in features may yield useful information concerning
the direction of minimum horizontal principal stress subtle variations in the local stress field, especially
(SHm[n). The idea that breakouts reflect local in where structural features like the pattern of frac-
situ stress conditions is supported by the agreement turing locally influence the orientation of regional
between horizontal stress directions inferred from horizontal stresses.
breakouts and stress orientations obtained using
other types of stress data such as earthquake fo-
Structural setting
cal mechanisms, hydraulic-fracturing measurements
and geologic indicators of stress (Bell and Gough, The Texas Panhandle-southern Oklahoma study
1982; Gough and Bell, 1981, 1982; Hickman et a l , region (Figs. 1 and 2) is an area of deep sedimen-
1985; Plumb and Hickman, 1985; Dart and Zoback, tary basins filled with thick sequences of Cambrian
1989). and younger Paleozoic marine rocks separated by

Structural and Tectonic Modelling and its Application to Petroleum Geology edited by R.M. Larsen, H. Brekke, B.T. Larsen and E. Talleraas.
NPF Special Publication 1, pp. 519-529. Elsevier, Amsterdam. Norwegian Petroleum Society (NPF), 1992.
 R.L. Dart and H.S. Swolfs

Fig. 1. Marietta basin study area. Well locations are plotted as solid dots with the mean orientations of wellbore data distributions
(bars). Wellbore data distributions are shown as either orthogonally oriented trends (bars of equal length) or a unimodal trend (bars of
unequal length). Faults are solid lines. Regional state of stress (SHmax) is the orientation of the large open arrows. The distribution of
wellbore enlargement data is shown in the rose diagram (inset).

uplifted basement structures (Jordan, 1967; Johnson The Marietta basin in Oklahoma is located at the
et al., 1972, 1984; Johnson and Denison, 1973). The eastern end of the Amarillo-Wichita uplift (Fig. 1).
primary west-northwest structural trend of basins The Marietta basin, a northwest-trending synclinal
and basement highs, extending across southern Okla- structure, is bounded on the northeast by the Criner
homa into the Texas Panhandle, developed through- uplift and on the southwest by the Waurika-Muenster
out the Paleozoic Era during periods of basement uplift. During the early depositional history of this
faulting, igneous intrusion, basin subsidence and oro- area, predating the formation of the Marietta syn-
genic deformation (Ham et al., 1964; Hoffman et cline, a thick sequence of Cambrian to Ordovician
al., 1974; Wickham, 1978). The stratigraphic history carbonate sediment was deposited. This early depo-
of the region appears to be dominated by episodes sitional sequence is overlain by Upper Ordovician
of carbonate and shale deposition attributed to eu- through Lower Mississippian limestone, sandstone
static fluctuations and intervals of clastic deposition and shale strata (Henry, 1968).
associated with subareal erosion of uplifted areas. The Wichita orogeny in Late Mississippian to
Subparallel faults and horizontal-stress orientations: in-situ stresses inferred from elliptical wellbore enlargements 521

Fig. 2. Bravo dome study area. Well locations are plotted as solid dots with the mean orientations of wellbore data distributions (bars).
Wellbore data distributions are shown as either orthogonally oriented trends (bars of equal length) or a unimodal trend (bars of
unequal length). Faults are solid lines. Regional state of stress (SHmax) is the orientation of the large open arrows. The distribution of
wellbore data is shown in the rose diagram (inset).

Middle Pennsylvanian (early Atokan) time and the deposition within the syncline began in late Mid-
Arbuckle orogeny in Middle Pennsylvanian (early dle Pennsylvanian (Desmoinesian) time (Frederick-
Desmoinesian) to Late Pennsylvanian (early Vir- son and Redman, 1965). Basin subsidence associated
gilian) time were significant events in the struc- with sediment loading continued into post-Permian
tural and depositional history of the area (Tomlinson time (Frederickson and Redman, 1965). During the
and McBee, 1959). Formation of the Criner and structural and tectonic history of southern Okla-
Muenster-Waurika uplifts and the faulting and fold- homa, significant amounts of dip-slip and left-lateral
ing that occurred during the initial development strike-slip displacement have occurred on faults
of the Marietta syncline accompanied the Wichita within the Amarillo-Wichita uplift (Harlton, 1963;
orogeny (Tomlinson and McBee, 1959). Significant Butler, 1980; Donovan, et al. 1982; McLean and
522 R.L. Dart and H.S. Swolfs

Sterns, 1983). The Frontal fault system, a regional bore, and (2) fracture borehole enlargements in-
fault system associated with the Amarillo-Wichita ferred to be associated with either drilling-induced
uplift, strikes west-northwest/east-southeast across hydraulic fracturing, or with favorably oriented, pre-
southern Oklahoma into the Marietta basin. existing fractures or closely spaced point sets inter-
The Bravo dome (Fig. 2) is a Late Mississippian secting the wellbore (Dart and Zoback, 1989). It is
to Early Pennsylvanian age tectonic feature posi- assumed that the orientation of borehole enlarge-
tioned between the Dalhart and Palo Duro basins ments, due to preexisting fractures, reflects the gen-
(Budnik and Smith, 1982). Formation of the Bravo eral trend of these fractures, and that fractures inter-
dome coincides with development of the Amarillo- secting the wellbore, whether hydraulically induced
Wichita uplift in Early Pennsylvanian time. By mid- or preexisting, can contribute to a structural failure
Permian time, carbonate and clastic (arkosic) sed- of the wellbore when aided by the in-situ state of
iment deposited in response to Late Mississippian stress. Fracture wellbore enlargements are inferred,
to Early Pennsylvanian tectonism completely cov- therefore, to be stress related when they orient in
ered structural features like the Bravo dome and the direction of SHm2LXy orthogonal to the trend of
filled the smaller basins in the region (McGookey breakouts. Thus, an azimuthal data distribution con-
and Goldstein, 1982). A sequence of Permian age taining breakout and fracture wellbore enlargement
halite-bearing formations interbedded with marine orientations should have a bimodal appearance with
mudstone, siltstone, anhydrite and dolomite rock bimodal trends being orthogonal (Dart and Zoback,
units covers the Bravo dome structure (Presley, 1989; Morin et al., 1989). Wellbore enlargement data
1980). This overlying sedimentary rock accumula- distributions are shown in Figs. 1 and 2 as either
tion is cut by a series of northwest-striking high-angle unimodal or bimodal.
faults that define basement structures (Dutton et al., Well data sets used in this study were selected from
1982). The most prominent of these faults is the a larger set of available data. The criteria used for
west-northwest-striking Potter County fault separat- data selection were that each data set consist of four
ing the Bravo dome from the Dalhart basin (Fig. 2). or more in-situ stress-associated elliptical wellbore-
Basement faults associated with Paleozoic tecton- enlargement intervals and that each azimuthal dis-
ism of the Amarillo-Wichita uplift are probably re- tribution, either unimodal or bimodal, must have an
verse faults (Dutton et al., 1982). Deformation of angular deviation (standard deviation) of <20. In
Miocene Ogallala Formation strata associated with the case of bimodal distributions, each orthogonal
these faults is indicative of Cenozoic fault movement trend had to have an angular deviation of <20. The
in the Tbxas Panhandle (Budnik, 1987). data used are give in Tables 1A and 2A. Additional
data tables (Tables 1B-D, 2B-D) show a comparison
Wellbore enlargement data of the Marietta basin and Bravo dome area data
sets. The data presented in these tables illustrate
Wellbore enlargement data consist of the mea- similarities in stratigraphy and in the distribution of
sured long-axis orientations of certain elliptically wellbore-enlargement azimuthal data.
shaped features observed on standard petroleum
industry four-arm, high-resolution dipmeter and Stress trajectory rotation
fracture-identification well logs. These features are
inferred to be in-situ stress related. Stress-related In the Marietta basin and Bravo dome areas,
wellbore enlargement features are intervals within the direction of SHmaiX (Figs. 1 and 2) is oriented
the wellbore where the borehole diameter has sud- oblique (east-northeast; von Schonfeldt et al., 1973;
denly increased in one direction beyond the normal Hooker and Johnson, 1969; Borjeson and Lamb,
bit size. Within this interval, normal logging tool 1987) to west-northwest/east-southeast-striking sub-
rotation is interrupted and the elliptically shaped parallel faults. We assume that reorientation or ro-
enlargement interval is tracked vertically. In order tation of the stress trajectories will occur in the
for a wellbore enlargement feature of this kind to vicinity of these fault-bounded areas, and that due
be considered stress related, its long-axis azimuth to fault geometry and the mechanical properties of
must not coincide with the bearing of the vertical stress-field discontinuities, the sense of rotation and
deviation of the hole or show an obvious correlation the amount of rotation will vary from point to point
to mechanical "drilling-induced" enlargement of the across the fault-bounded area.
wellbore. It has been shown (Segall and Pollard, 1980;
The two types of wellbore enlargement features Rodgers, 1984; Hempton and Neher, 1986) that
important to the stress study are: (1) stress-induced en echelon faults having a left-stepping or right-
breakouts associated with shear failure at the well- stepping left-lateral strike-slip sense of displacement
Subparallel faults and horizontal-stress orientations: in-situ stresses inferred from elliptical wellbore enlargements 523

TABLE 1

A. Marietta basin drill hole and wellbore enlargement data

Map location 1 2 3 4 5 6 7 8 9 10 11
# of drill*
holes 3 1 1 1 1 1 2 1 1 2 1
Surface
elevation 893** 866 793 855 840 876 926** 854 845 785** 918
Total
depth 7905** 6871 9934 8028 7708 7760 7725** 8399 8799 9042** 10547
Logged 6602- 4471- 5407- 6028- 5590- 5708- 6174- 6333- 6739- 6951- 8911-
interval 7904** 6464 6527 9771 7799 7745 7710** 8383 8798 9061** 10516
Data 7133- 5987- 5726- 7214- 5818- 6508- 7151- 6833- 8337- 7794- 8939-
interval 7545** 6446 6446 7894 7619 7698 7619** 8189 8635 8729** 10211
Vertical
ft. of 90/257 216/- 132/20 54/52 -/421 48/125 149/14 92/59 -/162 26/260 28/88
enlargement
(NE/NW)
# of
enlargement 2/4 4/- 4/1 4/2 -/10 3/2 5/1 2/2 -/4 1/6 2/2
intervals
(NE/NW)
Distribution bi uni bi bi uni bi bi bi uni bi bi
Mode
orientation NE NE NE ENE
- NE NE NE
- NE ENE
Mean
Angular
43 54 67 77
- 56 46 61
- 58 84

deviation
Mode
5 20 11 7
- 12 9 12
- - 6

orientation NW
- NW NW NW NW NNW NW WNW WNW NNW
Mean
Angular
312
- 318 328 323 323 343 232 289 298 341

deviation 4
- - 10 18 5
- 5 11 11 12

*Two or more drill holes may be located in close proximity, at the same map location. ** Average value.
Depths are measured from ground surface. A slash (/) separates data from bimodal distributions, modes are oriented generally NE (0
to 90 degrees) and NW (270 to 360 degrees). Abbreviations: bimodal (bi), unimodal (uni).

B. Marrietta basin borehole stratigraphic data

Clastic
Sandstone Limestone Shale Rock Ss-Ls-Sh-Cong Ss-Sh Total
Upper
Pennsylvanian 832( .032) 832( .032)
Middle
P e nn s y l y a n i a n 239( .009) 2476( .094) 806(.031 6709(.25) 140K.053) 1840(.070) 1347K .51)
Lower
P e n_ns_y I v a n i a n 2397(.091) 2397( .091)
Upper
Devonian 492( .019) 492( .019)
Upper
Ordovician 6054(.23) 188(.007) 6242(.24)
Middle ~
Ordovician 346( .014) 2042(.077) 58K.22) 2969(.11)

Total 1071(.041) 6400( .24) 5198(.20) 806(.031 9687(.37) 140K.053) 1840(.070) 26403(1.0)

The data are the vertical borehole feet logged and its precentage of the total footage logged, in parenthesis. Abbreviations: sandstone
(Ss), limestone (Lm), shale (Sh) and conglomerate (Cong). A hyphen (-) indicates units with mixed lithologies.

C. Marrietta basin wellbore enlargement data

Sandstone Limestone Shale Ss-Ls-Sh Ss-Ls-Sh-Cong Ss-Sh Total

Middle 16 ( .007)7 43( .019)/ 180( .078)/ 408 ( .18)/ 18(.008)/ 665(.29)/
Pennsylvanian 1( .016) K.016) 4( .065) 7(.12) K.016) 14(.23)
Lower 46( .020)/ 21( .009)/ 134( .058)/ 70(.031)/ 27K.12)/
Pennsy1vanian 1(.016) 1( .016) 6(.10) K.016) 9(.15)
Upper 17 ( .007)/ 17(.007)/
Deyonjian 1( .016) K.016)
Upper 76(.033)/ 1096( .48)/ 1172(.51)/
Ordovician 1( .016) 25( .61) 26(.43)
Middle" 35( .015)/ 24(.011)/ 9K.039)/ 18(.008)/ 168(.073)/
Ordovician 3( .049) 2( .033) 5(.082) K.016) IK.18)
Total 92( .040)/ 1220( .53)/ 242(.11)/ 499(.22) 152(.07)/ 88(.038) 2293(1.0)/
2( .033) 30( .49) 8(.13) 12(.20) 7(.12) 2(.033) 61(1.0)

The data are the vertical feet of borehole enlargement (% of the total vertical feet of enlargement)/number of enlargement intervals
(% of total the number). Abbreviations: sandstone (Ss), limestone (Lm), shale (Sh) and conglomerate (Cong). A hyphen (-) indicates
units with mixed lithologies.

D. Marrietta basin wellbore enlargement data

Azimuthal orentation (degrees)

W-E WNW NW NNW N-S NNE NE ENE E-W
(270-281) (281-304) (304-326) (326-349) (349-360/0-11) (11-34) (34-56) (56-79) (79-90) Total
Middle 52(.023)/ 288( .13)/ 126( .055)/ 90 ( .039)/ 83(.036)/ 26 ( .011)/ 665(.29/
Pennsylvanian K .016) 6(.10) 4( .066) 1( .016) K.016) K.016) 14(.23)
Lower 39(.017)/ 8( .003)/ 84( .036)/ 70(.031)/ 59(.026)/ I K . 0 0 5 ) / 27K.11)/
Pennsylvanian 1(.016) 1(.016) 2( .033) 1( .016) 3(.049) K.016) 9(.15)
Upper 17(.007)/ 17 ( .007)/
Devonian K.016) 1(.016)
Upper 225( .10)/ 57(.025)/ 330(.14)/ 48(.021)/ 96( .042)/ 134(.058)/ 282(.12)/ 9(.004)/ 1172(.51)/
Ordovician 3( .049) 2(.033) 8(.13) 2( .033) K.016) 5(.082) 5(.082) K.016) 26( .18)
Middle 44( .019)/ 15(.007)/ 42(.018)/ 36(.016)/ 3 K . 0 1 4 ) / 168(.073)
Ordovician 1(.016) 2( .033) 3(.049) 3(.049) 2(.033) IK.18)

225(.10)/ 148 ( .065)/ 6 6 K . 2 9 ) / 273(.12)/ 160(.070)/ 179(.078)/ 235(.10)/ 372(.16)/ 40(.017)/ 2293(1.0)/
Total 3( .049) 4( .066) 15(.25) 10(.16) 2( .033) 2(.033) I K .18) IK.18) 3( .049) 61(1.0)

The data are the vertical feet of borehole enlargement (% of the total vertical feet of enlargement)/number of enlargement intervals
(% of total number).
524 R.L. Dart and H.S. Swolfs

TABLE 2

A. Bravo dome drill hole and wellbore enlargement data

Map location 1 2 3 4 5 6 7 8 9 10
# of d r i l l
holes 1 1 1 1 1 1 1 1 1 1
Surface
elevation 3500 3451 3615 4020 3715 3706 3542 3590 3539 3257
Total
depth 4757 6881 4475 4202 4076 3733 4064 3596 3750 2781
Logged 2740- 837- 2469- 2192- 2075- 2883- 1965- 400- 1650- 1438-
interval 4747 6839 4469 4198 4075 3583 4093 3618 3739 2779
Data 3110- 1227- 2788- 2322- 2340- 3031- 2428- 826- 2453- 2291-
interval 3747 5691 3655 3244 3157 3505 4019 2786 3100 2611
Vertical
f t . of 69/16 143/96 142/87 91/60 -/55 21/75 53/75 116/13 -/53 166/-
Enlargement
(NE/NW)
# of
enlargement 2/2 5/5 3/4 2/6 -/4 1/3 3/3 3/1 -/4 4/-
intervals
(NE/NW)
Distribution bi bi bi bi urn bi bi bi uni uni
Mode
orientation NNE NE NE NE
- NE NE NNE
- NNE
Mean
Angular
33 59 45 30
- 35 25 33
- 19

deviation
Mode
17 14 8 20
- - 11 9
- 19

orientation WNW NW NW NW NW NW WNW WNW NW

-
Mean
Angular
293 336 314 322 326 328 302 302 317
-
deviation 2 9 4 4 12 8 6
- 19
-
A slash (/) separates data bimodal distributions, with modes oriented generally NE (0 to 90 degrees) and NW (270 to 360 degrees.
Abbreviations: bimodal (bi), unimodal (uni).

B. Bravo dome area borehole stratigraphic data

Clastic Silts-Ls- Igneous

Dolomite Limestone Shale Rock Anhy-Cong rock Total
Upper
Permian 8379( .36) 1482(.063) 274(.012) 10135(.43)
Middle
P_e_nnsy_l y an i an 1297 ( .055) 7518(.32) 3556(.15) 1237K.53)
Lower
P e n n sy ly a n i a n 109( .005) 206(.009) 315(.013)

Pr e c a m b r i a n 670(.029) 670(.029)

Total 8379(.36) 2888( .12) 7518(.32) 3762M6) 274(.012) 670(.029) 23491(1.0)

The data are vertical borehole feet logged and its percentage of the total footage logged, in parenthesis. Abbreviations: sandstone (Ss),
limestone (Lm), shale (Sh) and conglomerate (Cong). A hyphen (-) indicates units with mixed lithologies.

C. Bravo dome area wellbore enlargement data

Clastic Igneous
Dolomite Limestone Shale Rock Rock Total
Lower 682(.51)/ 32( . 0 2 4 ) / 714(.54)/
Permian 33(.60) 1 ( .018) 34(.62)
Middle 515(.39)/ 82(.062)/ 597(.49)/
Peinnsy 1. v a n i a n 16(.29) 3(.055) 19(.35)
19(.014)/ 19(.0140/
Pre c'ambri an 2{.036) 2(.036)

682(.51)/ 32(.024)/ 515(.39)/ 82(.062)/ 19(.014)/ 1330(1.0)/

Total 33(.60) 1(.018) 16( .29) 3(.056) 2(.036) 55(1.0)

The data are the vertical feet of borehole enlargement (% of the total vertical feet of enlargement)/number of enlargement intervals
(% of the total numer).

D. Bravo dome area wellbore enlargement data

A z i m u t h a l o r e n t a t i o n (degrees)
W--E WNW NW NNW N-S NNE NE ENE E-W
(270--281) (281-304) (304-326) (326-349) (349-360/0-11) (11-34) (34-56) (56-79) (79-90) Total
Lower 45( . 0 3 4 ) / 152(.11)/ 86( . 0 6 5 ) / 39(.029)/ 152(.ll)/ 1 5 6 M 2 ) / 96(.072)/ 726(.55)/
Permian 5( .091) 12( .22) 6(.ll) 2(.036) 3(.055) 5(.091) 2(.036) 35(.64)
Middle 61( . 0 4 6 ) / 152(.ll)/ 12(.009)/ 14(.011)/ 338(.25)/ 8(.006)/ 585(.44)/
Pennsylvanian 2() 5() K.018) K.018) 8(.15) K.018) 18(.33)
10(.008)/ 9(.007)/ 19(.014)/
P r e c a m b r i an K.018) K.018) 2( . 0 3 6 )

106( . 0 8 0 ) / 3 0 4 ( . 2 3 ) / 86( . 0 6 5 ) / 6K.046)/ 175(.13)/ 494(.37)/ 104(.078)/ 1330(1.0)/

Total 7 ( .13 17(.31) 6( .11) 4(.073) 5(.091) 13(.24) 3(.055) 55(1.0)

The data are the vertical feet of borehole enlargement (% of the total vertical feet of enlargement)/number of enlargement intervals
(% of total number).
Subparallel faults and horizontal-stress orientations: in-situ stresses inferred from elliptical wellbore enlargements 5

Discussion of results

We make the assumption that reorientation of

horizontal stress trajectories across an area bounded
by subparallel faults can be indicated by a corre-
sponding change in the orientations of stress-related
wellbore-enlargement features in wells located across
the study area.
The two areas of interest, the southwestern edge of
the Marietta basin and the eastern flank of the Bravo
dome, are structurally similar in that they both have
a thick sequence of relatively flat-lying sedimen-
tary rock offset by west-northwest/east-southeast-
striking subparallel basement faults. The distribution
of wellbore enlargement orientations from wells lo-
cated adjacent to and between these subparallel
faults exhibits bimodal trends that are orthogonal in
character (Figs. 1 and 2). Because SHm2LX is regionally
oriented northeast (Hooker and Johnson, 1969; von
Schonfeldt et al., 1973; Borjeson and Lamb, 1987),
these bimodal-orthogonal distributions are assumed
composed of northwest-oriented wellbore breakouts
and northeast-oriented fracture wellbore enlarge-
ments (Dart and Zoback, 1989).
Fig. 3. Schematic diagram illustrating the influence of Young's Observed variation in the calculated azimuthal
modulus (E) of displacement discontinuities ( D ) on stress means of wellbore-enlargement data distributions
trajectory orientation (\ and ). Shown are displacement from wells located between subparallel faults in the
discontinuities of fault-bounded areas with a Young's modulus
less than (A), equal to (B) and greater (C) than that of the
two study areas has the appearance of a systematic
adjacent rock (modified after Hudson and Cooling, 1988). rotation. This apparent rotation is inferred to be
indicative of stress-trajectory reorientation between
subparallel faults and may best be seen among wells
can initiate stress trajectory rotation in the vicinity of 8 through 11 in the Marietta basin (Fig. 1) and wells
these en echelon faults. The distance separating en 7 through 9 in the Bravo dome study area (Fig. 2).
echelon faults and the amount of offset or overlap of The maximum amount of apparent stress-trajectory
the ends of these faults will influence the extent of rotation is 52 between wells 9 and 11 (Fig. 1) in the
stress trajectory rotation. Marietta basin and 22 between wells 7 and 9 (Fig. 2)
Rotation of stress trajectories will also occur when in the Bravo dome area. A bimodal-orthogonal dis-
the subparallel faults function as displacement dis- tribution of wellbore-enlargement orientations from
continuities. If differences do exist in mechanical the Marietta basin strongly contrasts with the uni-
properties of the rock units involved, the fault zone modal distribution of wellbore enlargement orien-
or fault-bounded area (Fig. 3) will function as a tations, inferred to be breakouts, from the eastern
discontinuity within the regional stress field. If a Anadarko basin adjacent to and north of the Mari-
fault-bounded area functions as a stress-field dis- etta basin (Dart, 1989; and Dart and Zoback, 1989).
continuity within the regional stress field having a In the eastern Anadarko basin, drill holes with stress
Young's modulus greater than that of the adjacent associated wellbore enlargements were not or were
rock (Fig. 3C), it is predicted (Hudson and Cooling, only rarely located between or in the vicinity of
1988) that the stress trajectory pattern will rotate mapped subparallel faults.
either counterclockwise or clockwise when crossing Drilling-induced hydraulic fracturing is more likely
boundary faults. to develop in a normal-faulting stress regime (Healy
Stress trajectory rotation may also be predicted if et al., 1984; Stock et al., 1985). Correspondingly,
an anisotropic rock fabric, such as strongly oriented formation of hydraulic fracture-related wellbore en-
jointing, is present. In this case the rock fabric largements in drill holes located in the Marietta
property can act as a stress guide to redirect the basin and Bravo dome study areas are common
stress trajectory orientation (Amadei et al., 1988). and may indicate local extensional (normal) fault-
ing stress conditions. Rodgers (1984) and Hempton
526 R.L. Dart and H.S. Swolfs

and Neher (1986) described structural settings where Values for the Young's modulus of the strati-
extensional and compressional stresses can develop graphic section of the two study areas are not known.
between and near the ends of en echelon strike-slip To our knowledge, the necessary field and labora-
faults in response to regional horizontal compres- tory studies have not been made to determine the
sion oblique to their strikes. Subparallel faults in relative similarity or difference in Young's modulus
the Marietta basin and Bravo dome area are not of the rock units between or beyond the subparallel
en echelon by definition; however, these two areas faults in the Marietta basin or the Bravo dome area.
appear to structurally resemble the areas described Because the rock units on opposite sides of these
by Rodgers (1984) and Hempton and Neher (1986). subparallel faults are offset vertically (Harlton, 1963;
Extensional stress conditions (normal-faulting stress Butler, 1980; Dutton et al., 1982; Donovan et al.,
regime) might exist within the Marietta basin and 1982; McLean and Sterns, 1983), a Young's modulus
the Bravo dome study area where 5i7max is ori- difference may indeed exist. However, any varia-
ented oblique (northeast) to the strike of subparallel tion would probably be small because the rock units
faults. on either side of the offset appear to be similar

9730'
1

0
1 I
io^>^l r ^H
I

^ /
//
^-**><^ TV.
/ / Marietta Basin
^ N ^,
^^^ /
/
/
7^ / /
__- v ^*^. / /
tX^ ^ '
\5K5S? ^
^ ^ /
/
/
/
34c _, ^ \
OKLAHOMA /7-'^* / -
N s ^ /
/* TEXAS\
""^ ^S^-' ^ ^ / /
Muenster-Waurika^Xe^^C /
^ >-^9 ^ 1 Uplift ^^^
/ /\ /
1
1
\.
J ^L "**

Fig. 4. Schematic diagram illustrating the inferred local rotational pattern of stress trajectories (dashed lines) across the Marietta basin
(A) and Bravo dome (B) study areas.
Subparallel faults and horizontal-stress onentations: in-situ stresses inferred from elliptical wellbore enlargements 527

lithologically. Therefore, the influence of Young's extending into the Marietta basin, in response to the
modulus on stress trajectory rotation in the two study east-northeast/west-southwest regional 5i7max, may
areas is thought to be, at best, of minor importance. have generated a left-lateral sense of displacement
Preexisting fractures in an array of orientations on these faults. In this scenario, a left-lateral sense of
may exist within blocks bounded by subparallel faults. displacement could generate a local extensional state
Well-established fracture trends or joint set patterns of stress in the vicinity of these faults where stress
could function as anisotropic rock fabric elements trajectories would be locally rotated and where the
to exert an influence on stress trajectory orientation. orientations of stress sensitive features like elliptical
Studies of fracture and lineament patterns exist for wellbore enlargements would reflect this local state
the Bravo dome area and the Slick Hills area of of stress anomaly.
southern Oklahoma (Wilhelm and Morgan, 1986),
an area adjacent to and northwest of the Marietta
basin. Surface and subsurface fractures in the Bravo Conclusions
dome area form bimodal patterns oriented west-
The appearance of a systematic change in the ori-
northwest and northeast (Gustavson and Budnik,
entation of stress-induced or stress-related elliptical
1985; Gustavson and Finley, 1985). Mapped linea-
wellbore enlargements among some closely spaced
ments in the Slick Hills area also formed bimodal
drill holes located within a structurally well-defined
trends and were oriented northwest and northeast
area suggests a local variation in the regional hor-
(Wilhelm and Morgan, 1986). From the limited data
izontal stress pattern. Detailed knowledge of local
available, the degree of influence fractures or joint
variations in the regional state of stress is useful
sets may have on the reorientation or rotation of
when evaluating the seismic potential of an area or
stress trajectories within the Marietta basin or Bravo
when local stress conditions influence the success of
dome study areas is not known. However, the area
a borehole drilling or a wellbore fracturing program.
between subparallel basement faults in these two
Questions of this nature can be addressed through
areas may have experienced intense fracturing his-
a detailed and systematic analysis of stress-related
tories resulting in a variety of fracture orientations.
wellbore features.
Fractures or joint sets between subparallel faults
having orientations aligned with the local orientation
of SHm2iX could aid in the development of elliptical References
enlargements.
From the above discussion, it is not possible to Amadei, B., Swolfs, H.S. and Savage, W.Z., 1988. Gravity-
induced stresses in stratified rock masses. Rock Mech. Rock
state conclusively that the observed apparent varia-
Eng., 21: 1-20.
tion in the calculated mean stress orientations of any Bell, J.S. and Gough, D.I., 1979. Northeast-southwest compres-
individual well is a direct indication of stress trajec- sive stress in Alberta evidence from oil wells. Earth Planet.
tory rotation within the two study areas. However, Sei. Lett., 45: 475-482.
in light of the apparent observed shift in the ellip- Bell, J.S. and Gough, D.I., 1982. The use of borehole breakouts
in the study of crustal stress. In: M.D. Zoback and B.C. Haim-
tical long-axis orientation of wellbore enlargements,
son (Editors), Proceedings of Workshop XVII Workshop
it does seem reasonable to speculate on the possibil- on Hydraulic Fracturing Stress Measurements. U.S. Geol.
ity of stress trajectory rotation given the location of Surv., Open-File Rep., 82-1575: 530-557.
drill holes with respect to mapped subparallel faults Borjeson, R.W. and Lamb, T.J., 1987. Geotechnical borehole
(Fig. 4). A bimodal-orthogonal distribution of well- testing report Holtzclaw No. 1 well (PD-10), Palo Duro basin.
bore enlargements would be expected in the Bravo Stones and Webster Engineering Corporation, No. 312.
Brown, R.O., 1978. Application of fracture identification logs
dome area where the maximum principal stress (\) in the Cretaceous of north Louisiana and Mississippi. Gulf
is vertical (Stone and Webster Engineering Corp., Coast Assoc. Geol. Soc, Trans., 28: 75-91.
1983) and thus, where wellbore enlargements as- Budnik, R.T., 1987. Late Miocene reactivation of Ancestral
sociated with drilling-induced hydraulic fracturing Rocky Mountain structures in the Texas Panhandle a
(Healy et al., 1984; Stock et al., 1985) would be more response to basin and range extension. Geology, 15: 163-166.
Budnik, R.T and Smith, D., 1987. Late Miocene reactivation of
likely to form. A similar bimodal-orthogonal distri-
Ancestral Rocky Mountain structures in the Texas Panhandle
bution of wellbore enlargement orientations would a response to basin and range extension. Geology, 15:
not necessarily be expected for data from the Ma- 163-166.
rietta basin. The Marietta basin is located within Butler, K.R., 1980. A structural analysis of the Cambro-
the Midcontinent stress province, a region domi- Ordovician strata on the north flank of the Wichita Moun-
nated by east west compressional tectonics (Zoback tains, Oklahoma (abstract). Geol. Soc. Am., Abstracts with
Programs, 12: 1-2.
and Zoback, 1980). However, suspected strain accu- Cox, J.W., 1970. The high resolution dipmeter reveals dip-related
mulation of east-northeast-striking subparallel faults borehole and formation characteristics. Soc. Prof. Well Log
528 R.L. Dart and H.S. Swolfs

Analysis, 11th Annu. Logging Symp., May 3-6, 1970, pp. and R.H. Shaver (Editors), Modern and Ancient Geosynclinal
1-25. Sedimentation. Soc. Econ. Paleontol. Mineral., Spec. Publ.,
Dart, R.L., 1989. Horizontal stresses from wellbore breakouts 19: 38-55.
and lithologies associated with their formation, Oklahoma Hudson, J.A. and Cooling, CM., 1988. In situ rock studies and
and Texas Panhandle. Okla. Geol. Surv. Circ, 90: 97-120. then measurement. In the U.K. Part 1, the current states
Dart, R.L. and Zoback, M.L., 1989. Well bore-breakout stress of knowledge. Int. J. Rock Mech., Mining Sei. Geomechan.,
analysis within the Continental United States. Log Anal. J., Abstr., 25(6): 363-370.
30: 12-25. Johnson, K.S., Branson, C.C., Curtis, N.M., Jr., Ham, W.E.,
Donovan, R.N., Sanderson, D J . , Marchini, D., 1982. An analysis Marcher, M.V. and Roberts, J.F, 1972. Geology and earth
of structures resulting from left-lateral strike-slip movement resources of Oklahoma, an atlas of maps and cross sections.
between the Wichita Mountains and Anadarko Basin, south- Okla. Geol. Surv. Educ. Publ., 1: 8.
western Oklahoma. Geol. Soc. Am., Abstracts with Programs, Johnson, K.S. and Denison, R.E., 1973. Igneous geology of the
14: 476. Wichita Mountains and economic geology of Permian rocks
Dutton, S.P., Goldstein, A.G. and Ruppel, S.C., 1982. Petroleum in southwest Oklahoma. Geol. Soc. Am., Guidebook for Field
potential of the Palo Duro Basin, Texas Panhandle. Texas Trip, 6: 35.
Bur. Econ. Geol., Rep. Invest., 123: 80. Johnson, K.S., Bruchfield, M.R. and Harrison, W.E., 1984.
Frederickson, E.A. and Redman, R.H., 1965. Geology of Love Guidebook for Arbuckle Formation field trip, southern Okla-
County, in Geology and petroleum of Love County, Okla- homa. Okla. Geol. Surv. Spec. Publ., 84-1: 23.
homa. Okla. Geol. Surv. Circ, 63: 7-47. Jordan, L., 1967. Geology of Oklahoma a summary. Okla.
Gough, D.I. and Bell, J.S., 1981. Stress orientations from oil- Geol. Surv., Geol. Notes, 27: 215-229.
well fractures in Alberta and Texas. Can. J. Earth Sei., 18: Mastin, L.G., 1984. The development of borehole breakouts in
638-645. sandstone. M.S. Thesis, Stanford University, Stanford, Calif.,
Gough, D.I. and Bell, J.S., 1982. Stress orientation from borehole 101 pp.
wall fractures with examples from Colorado, east Texas and McLean, R. and Sterns, D.W., 1983. Fault analysis in Wichita
northern Canada. Can. J. Earth Sei., 19: 1358-1370. Mountains. Am. Assoc. Pet. Geol., Bull., 67: 511-512 (ab-
Gustavson, T.C. and Budnik, R.T., 1985. Structural influences on stract).
geomorphic processes and physiographic features, Texas Pan- McGookey, D.A. and Goldstein, A.G., 1982. Structural influence
handle technical issues in siting a nuclear-waste repository. on deposition and deformation at the northwest margin of the
Geology, 13: 173-176. Palo Duro Basin, in Geology and geohydrology of the Palo
Gustavson, T.C. and Finley, R.J., 1985. Late Cenozoic geo- Duro Basin, Texas Panhandle a report on the progress of
morphic evolution of the Texas Panhandle and northeastern nuclear waste isolation feasibility studies. Texas Bur. Econ.
New Mexico case studies of structural control on regional Geol. Circ, 82-7: 28-37.
drainage development. Texas Bur. Econ. Geol., Rep. Invest., Morin, R.H., Anderson, R.N. and Barton, CA., 1989. Analysis
148: 42. and interpretation of the borehole televiewer log informa-
Haimson, B.C. and Herrick, CG., 1985. In situ stress evaluation tion on the state of stress and the lithostratigraphy at hole
from borehole breakouts, experimental studies. Proc. 26th 504B. Proc. Ocean Drilling Program, Scientific Results, Vol.
U.S. Symp. on Rock Mechanics, Rapid City, S.D., 1985, 2: Ill, College Station, Texas.
1207-1218. Plumb, R.A. and Hickman, S.H., 1985. Stress-induced borehole
Ham, WE., Denison, R.E. and Merritt, CA., 1964. Basement elongation; a comparison between the four-arm dipmeter and
rocks and structural evolution of southern Oklahoma. Okla. the borehole televiewer in the Auburn geothermal well. J.
Geol. Surv. Bull., 95: 302. Geophys. Res., 90(B7): 5513-5521.
Harlton, B.H., 1963. Frontal Wichita Fault system of south- Presley, M.W, 1980. Upper Permian salt-bearing stratigraphic
western Oklahoma. Am. Assoc. Pet. Geol., Bull., 47: 1552- units, in geology and geohydrology of the Palo Duro Basin,
1580. Texas Panhandle a report on the progress of nuclear waste
Healy, J.H., Hickman, S.H., Zoback, M.D. and Ellis, W.L., 1984. isolation feasibility studies. Texas Bur. Econ. Geol. Circ, 80-7:
Report on televiewer log and stress measurements in core 33-40.
hole USW-G1, Nevada Test Site, December 13-22, 1981. U.S. Rodgers, D.A., 1984. Analysis of pull-apart basin development
Geol. Surv., Open-File Rep., 84-15: 47. produced by en echelon strike-slip faults In: A.G. Sylvester
Hempton, M.R. and Neher, K., 1986. Experimental fracture, (compiler), Wrench Fault Tectonics. Am. Assoc. Pet. Geol.,
strain and subsidence patterns over en echelon strike-slip Reprint Sen, 28: 345-360.
faults implications for the structural evolution of pull- Segall, P. and Pollard, D.D., 1980. Mechanics of discontinuous
apart basins. J. Struct. Geol., 8: 597-605. faults. J. Geophys. Res., 85(B8): 4337-4350.
Henry, G.E., 1968. Recent development in the Marietta basin. Stock, J.M., Healy, J.H., Hickman, S.H. and Zoback, M.D.,
In: J.S. Wendell (Editor), Basins of the Southwest, 1. Amer- 1985. Hydraulic fracturing stress measurements at Yucca
ican Association of Petroleum Geologists, Tulsa, Okla., pp. Mountain, Nevada and relationship to the regional stress
71-78. field. J. Geophys. Res., 90(B10): 8691-8706.
Hickman, S.H., Healy, J.H. and Zoback, M.D., 1985. In situ Stone and Webster Engineering Corporation, 1983. Area geo-
stress, natural fracture distribution and borehole enlargement logical characterization report for the Palo Duro and Dalhart
in the Auburn geothermal well Auburn, New York. J. basins, Texas. Tech. Rep. DOE/CH/10140-1, 429 pp.
Geophys. Res., 90: 5497-5512. Teufel, L.W., 1985. Insights into the relationship between well
Hooker, V.B. and Johnson, C.F., 1969. Near-surface horizontal bore breakouts, natural fractures and in situ stress. Proc. 26th
stresses, including the effects of rock anisotrophy. U.S. Bur. U.S. Symp.on Rock Mechanics, Rapid City, S.D., 1985, 2:
Mines, Rep. Invest., 7654: 18. 1199-1206.
Hoffman, P., Dewey, J.F. and Burke, K., 1974. Aulacogens Tomlinson, L.W. and McBee, W, Jr., 1959. Pennsylvania sed-
and their genetic relation to geosynclines and a Proterozoic iments and orogenies of Ardmore district, Oklahoma. In:
example from Great Slave Lake, Canada. In: R.H. Dott, Jr. Petroleum Geology of Southern Oklahoma, 2. American As-
Subparallel faults and horizontal-stress orientations: in-situ stresses inferred from elliptical wellbore enlargements 529

sociation of Petroleum Geologists, Tulsa, Okla., pp. 3-52. thermatic-mapper data for lineament analysis of the Slick
von Schonfeldt, H.A., Kehle, R.O. and Gray, K.E., 1973. Map- Hills area. In: R.N. Donovan (Editor), The Slick Hills of
ping of stress field in the upper Earth's crust of the U.S.. U.S. Southwestern Oklahoma Fragments of an Aulacogen.
Geol. Surv., Final Tech. Rep., Grant 14-08-0001-1222, Reston, Okla. Geol. Surv., Guidebook, 24: 35-39.
Va., 78 pp. Zoback, M.D., Moss, Daniel, Mastin, R.L. and Anderson, R.N.,
Wickham, J.S., 1978. The southern Oklahoma aulacogen In: J.S. 1985. Wellbore breakouts and in-situ stress. J. Geophys. Res.,
Wickham and R.E. Denison (Editors), Structural Style of the 90(B7): 5523-5530.
Arbuckle Region. Univ. Okla. Guidebook for Geol. Soc. Am. Zoback, M.L. and Zoback, M., 1980. State of stress in the
South-Central Annu. Meet. Field Trip, 3: 9-41. conterminous United States. J. Geophys. Res., 85(B121):
Wilhelm, S.J. and Morgan, K.M., 1986. Utilization of Lansat 6113-6156.

R.L. DART Branch of Geologic Risk Assessment, U S. Geological Survey, Box 25046, MS 966, Denver Federal Center, Denver, CO 80225,
U.S.A.
H.S. SWOLFS U.S. Geological Survey, Box 25046, MS 966, Denver Federal Center, Denver, CO 80225, U.S.A.
 531

References index

Aagaard, P., 323, 340, 517 Baird, R.A., 80

Aanstad, M.K., 381, 385, 394 Baker, D.M., 490, 492
Aarseth, I., 195 Baker, P.A., 345, 352
Adgestein, T, 251 Bakhtar, K., 288
Aga, O., I l l , 120 Baiberg, L, 371, 373, 380
Agdestein, T., 53, 219, 302, 394, 467 Balgimbayev, M.B., 119
Airy, B.G., 187,194 Ballard, J.F., 219
Akers, W H , 267 Balling, N., 17, 111,184
Allemand, P., 219 Balling, N.P., 7, 11,16
Allen, J.R.L., 509,527 Bally, A.W., 469, 478
Allen, U.S., 328, 337, 340 Bandis, S., 288
Alvarez, F., 77, 80 Bandis, S.C., 278, 288, 366, 379, 380
Amadei, B., 525, 527 Barber, D.J., 327, 340
Anda, E., 185,195 Barka, A.A., 84, 90
Andereassen, K., 185 Barnett, J.A.M., 293, 299, 302
Andersen, B.G., 189,190,195 Barr, D., 244, 247, 251, 297, 302, 392, 395, 465, 467
Andersen, C , 2, 3, 4, 9,16,17, 202, 216, 220, 439, 440, 441, 455, Barrientos, S.E., 290, 303
456, 498, 499, 506 Barton, CA., 528
Anderson, E.M., 85, 89, 389, 391, 394 Barton, N., 278, 279, 280, 281, 282, 288
Anderson, H., 303 Barton, P, 1, 3, 11,16, 17, 43, 46, 53, 54, 224, 229, 294, 302, 457,
Anderson, M.J., 305, 306, 323 463, 465, 467
Anderson, R.N., 436, 528 Bas, T., 479
Andersson, J., 365, 371, 373, 380 Baturin, O.G., 166,184
Anderton, R., 220, 479 Beach, A., 43, 44, 53, 199, 200, 205, 217, 219, 246, 251, 457, 459,
Andreassen, K., 162,195 462, 466, 467
Andresen, A., 483, 487, 488, 489, 490, 492, 493 Bear, J., 306, 323
Andrews, J.R., 208, 219 Beaumont, C , 5, 15,16
Angelier, J., 85, 87, 89 Belfield, W.C., 345, 352
Angelier, T., 263, 265, 267 Bell, J.S., 519, 527, 528,
Angevine, C.L., 327, 340 Benvenuto, G.L., 477, 478
Antonsen, H.P., 168,184 Berg, R.R., 314, 323, 325, 326, 328, 335, 336, 340
Antonsen, P., 184 Bergan, M , 213, 219, 504, 506
Aram, R.B., 155,162 Bergendahl, E., 110, 114,119, 122,130
Archie, G.E., 313, 323 Berggren, W.A., 188,195
Argand, E., 469, 478 Bergh, S., 492
Armstrong, G., 81 Bergh, S.G., 489, 492
Arnaud, H., 479 Berglund, L.T., 153,161,162,185
Arnow, J.A, 106 Berglund, T, 91,106,162,185
Atkinson, B.K., 327, 340 Bergsaker, E.I., 131
Auguston, J., 106 Bergvik, A., 492
Augustson, J., 161 Bertram, G.T., 49, 53, 219, 226, 228, 229, 272, 276, 294, 297, 299,
Austrheim, H., 59, 80,184
302
Avedik, E, 80
Beskow, B., 107,131, 165
Avrov, V.P., 118,119
Bessis, F., 45, 53, 323, 456
Axen, G.J., 105,107, 244, 247, 251
Biddle, K.T., 213, 219
Aydin, A., 389, 394
Biggins, D., 304
Billaux, D.M., 365, 371, 377, 379, 380
Bachman, S.B., 352 Bird, J.M., 16
Backer-Owe, K., 492 Bird, T., 53, 219, 467
Bckstrm, S.A., 483, 492 Birkenmajer, K., 482, 483, 492
Badley, M , 220 BIRPS Group, 54
Badley, M.E., 44, 45, 49, 52, 53, 205, 219, 248, 251, 300, 302, 304, Birtles, R., 293, 302
385, 394, 457, 458, 462, 463, 465, 467 Bishop, CD., 152
Bailey, C.C., 413, 419 Bishop, R.S., 117,119
532 References index

Bjaerke, T., 492 Calcagnile, G., 6,16

Bj0rlykke, K., 312, 313, 323, 326, 340, 511, 512, 515, 517 Caline, B.P.R, 355, 363
Black, R , 239, 241 Cameron, C.S., 251
Blair, D.G., 43, 53 Campbell, C.J., 205, 219
Blake, G.H., 349, 352 Cao, S., 427, 437, 439, 455
Bleil, U., 184,195 Carey, E., 85, 89
Blenkinsop, T.G., 340 Carman, P.C., 314, 323
Blundell, DJ., 220 Cartier, E.G., 80, 395
Blystad, P., 55, 57, 78, 80 Cartwright, J.A., 495, 498, 506
Bodine, J.H., 229 Caselli, F , 76, 80, 381, 392, 395
B0e, R , 87, 89, 111, 184 Casey, M., 479
B0en, F., 55, 56, 59, 80 Cassan, J.P., 326, 342
Boles, J.R., 515, 517 Cathles, L.M., 175,176, 183,184,190, 191,195
Boote, D.R.D, 300, 302 Celerier, B., 60, 80,129,131
Borjeson, R.W., 522, 525, 527 Chadwick, R.A., 94,106
Borradaile, G.J., 327, 340, 511, 517 Challinor, A., 483, 492
Bott, M., 85, 89 Chapman, T.G., 306, 323
Boulton, G.S., 189,195 Charlaix, E., 325, 340
Boussinesq, J., 306, 323 Chatellier, J.-Y., 424, 437
Bowen, A.N., 82 Cheadle, M.J., 94,106, 200, 219, 220
Bowen, J.M., 297, 302 Chen, W.P., 294, 302
Bowen, W, 43, 53 Chenet, P.Y., 7,13,17, 323, 456
Bowin, C, 222, 229 Chermette, J.C., 299, 302
Boyer, S.E., 469, 474, 476, 478, 490, 492 Chiarelli, A., 5,16, 323, 456
Childs, C , 299, 302, 355, 359, 363
Brace, W.F., 293, 302, 327, 340
Chiles, J.P., 355, 363
Brand, P.J., 363
Chilingar, G.V., 314, 323
Branson, C.C., 528
Choubey, V, 278, 280, 281, 288
Braun, J., 5,15,16
Choukroune, P., 240, 241, 251
Brekke, H., 59, 76, 80, 111, 130, 169,184, 381, 385, 392, 395
Christianson, M., 288
Brewer, J.A., 200, 219, 385, 395
Christie, P.A.F, 1, 3, 17, 43, 46, 53, 54, 59, 64, 81, 155, 162, 111,
Brewster, J., 397, 399, 407
111, 224, 229, 254, 256, 267, 111, 276, 398, 407, 421, 426,
Brink, A.H., 128,130
437, 461, 467
Brix, M., 237, 240
Christie-Blick, H.N., 119
Brock, W.G., 337, 340
Chroston, P.N., 75, 80
Broks, TM., 110,119
Chryssanthakis, P., 288
Brooks, S.G., 75, 80
Clark, D.L., 188,195
Brown, A., 202, 213, 219 Cloos, E., 237, 240
Brown, D.A., 399, 407 Cobbold, PR., 241, 250, 251
Brown, L.D., 106 Cochran, J.R., 7,16, 52, 53
Brown, R.O., 519, 527 Cockerham, R.S., 337, 340
Brown, S., 208, 219, 220, 297, 302 Coffield, D.Q., 246, 251
Brown, S.R., 379, 380 Coffin, M.F, 424, 437
Bruce, C.H., 265, 267 Cohen, J.M., 303
Bruchfield, M.R., 528 Cohen, M.J., 307, 323
Bruhn, R.L., 300, 302 Cohen, S.C., 293, 302
Brun, J.P., 197, 219, 239, 240, 241, 251 Coleman, M., 517
Brnier, B., 85, 89 Colletta, B., 263, 265, 267
Buchan, S.H., 483, 492 Collins, C.D.N., 251
Bucher, K., 184 Collinson, J.D., 517
Bucher-Nurminen, K., 59, 80 Coney, P., 81, 380, 467
Buchner, F., 515, 517 Conford, C , 444, 455
Buckingham, J., 304 Connell, E.R., 303
Budding, M.C., 357, 363 Conort, A., 78, 80
Budnik, R.T., 522, 527, 527, 528 Cook, FA., 395
Bukovics, C , 57, 66, 80, 307, 311, 323, 381, 385, 388, 392, 395 Cooling, CM., 525, 528
Burchfiel, B.C., 105, 107, 113, 119, 129, 131, 235, 239, 241, 244, Cooper, B.A., 16
247, 251, 265, 267, 296, 304, 467 Cooper, D.M., 303
Burgers. W.F.J., 16 Cooper, J.A., 361, 363
Burgess, C.F., i07,152 Cooper, W.G., 217, 220
Burke, K., 528 Coppersmith, K.J., 300, 303
Burton, W.C., 94,107 Cornford, C , 17
Butenko, G., 184 Cowan, D., 341
Butler, K.R., 521, 526, 527 Coward, M.P., 9,16, 52, 53,197, 200, 219, 220, 479
Butler, R.W.H., 469, 470, 472, 473, 474, 478 Cox, A.V., 130
Byerlee, J.D., 327, 340 Cox, J.W., 519, 527
References index 533

Cox, S.E, 327, 333, 340 Eastman, D., 474, 479

Crossley, R., 141,151 Eaton, J.P., 337, 340
Crouch, J.K., 344, 352 Ebdon, C.C., 220
Crowell,J.C, 392, 394,395 Ediriweera, K.K., 49, 53, 251
Cundall, P.A., 278, 281, 288, 380 Edrich, S.P., 276
Curtis, C , 517 Edvardsen, A., 130
Curtis, N.M. Jr., 528 Edwards, E.B., 353
Cutbill, J.L., 106, 483, 492, 493 Egan, S.S., 294, 303, 506
Egeberg, T, 53, 219, 467
Dahl, S.O., 185,195 Eggen, S., 80, 81
Dahleen, W.K., 352 Eidvin, T., 158, 159, 161,162, 163, 166,184
Dahlstrom, C.D.A., 269, 276, 426, 437, 469, 478, 490, 492 Eikeland, T.A., 492, 493
Dalland, A., 55, 80,152,162,166,184, 323, 395 Eiken, O., 492
Dalimann, W.K., 489, 492 Einstein, H.H., 379, 380
Dalimeyer, R.D., 59, 80 Ekern, O.F., 130
Dangerfield, J.A., 398, 399, 407 Eldholm, O., 55, 56, 57, 59, 60, 62, 68, 69, 71, 75, 76, 77, 78, 80,
Darot, M., 341 80, 81, 82, 102,106, 158,162, 388, 395, 483, 492, 492
Darracott, B., 437 Ellenor, D.W., 295, 302, 311, 323
Dart, R.L., 519, 522, 525, 528 Ellingsen, K.L., 195
Davies, E.J., 462, 467 Elliot, D., 490, 492
Davis, G.A., 113,119 Elliott, D., 469, 474, 476, 478
Davis, G.H., 81, 380,467 Ellis, P.G., 238, 239, 241, 247, 250, 251
Davy, P., 241,251 Ellis, W.L., 528
Day, G.A., 4,16 Elsworth, D.A., 365, 380
de Charpal, O., 459, 467 Elverh0i, A., 158,162, 169,184, 189,195, 492, 493
de Marsily, G., 342 Emmons, W.H., 303
De Geer, G., 187,195, 481, 489, 492 Enfield, M.A., 200, 220
De Graciansky, P.C., 479 Engelder, J.T., 337, 340
De'Ath, N.G., 208, 219, 224, 229, 462, 467 England, P.C., 59, 80
Dean, J.M., 220 England, W.A., 305, 323
DeBuyl, M., 425, 428, 437 Erichsen, T, 205, 220, 243, 251
Deegan, C.E., 41, 42, 409, 419 Espitalie, J., 424, 437
Dembicki, H. Jr., 305, 306, 323 Etheridge, M.A., 327, 333, 337, 340, Ml, 479
Denison, R.E., 520, 528 Evans, C.E., 407
Derksen, S.J., 107 Eynon, G., 3,16
Derksen, S.T., 152
Dershowitz, W.S., 379, 380 Faber, E., 78, 80
Desprairies, A., 80 Faerseth, R.B., 25, 42, 53, 80, 89, 106, 110, 112, 119, 130, , 161,
Dewey, J.F., 1,16, 59, 80, 219, 222, 229, 384, 395, 469, 478, 528 162,184, 243, 251, 467
Dieterich, J.H., 85, 90 Fairbanks, R.G., 188,195
Dingwall, R.G., 199, 205, 219, 220, 385, 395 Faleide, J.E., 489, 492
Dixon, TN., 407 Faleide, J.L, 81, 91, 94, 99,106, 109, 112, 114,119, 121, 122,130,
Dodson, C.R., 305, 323 133,151, 158, 159,162, 395, 492
Doery, R., 288 Falvey, D.A., 6,16, 251, 437
Donath, F.A., 233, 240 Farrell, H.E., 407
Donovan, R.N., 521, 526, 528 Fast, C.R., 517
Dore\ A.G., 24, 32, 42,199, 202, 220, 409, 420 Faugere, E., 237, 240
Dostman, HJ., 81 Faure, J.L., 299, 302
Doudoux, B., 470, 474, 478 Fearn, P., 119
Douglass, D.N., 106, 493 Feazel, C.T, 155,162
Dow, W.G., 162 Fichler, C , 44, 52, 53
Dowdeswell, E.K., 93,106 Fichler-Fettig, C , 52, 53
Dowling, L.M., 102,106 Finley, R.J., 527, 528
Downey, M.W., 328, 340 Fischer, M., 83, 90
Dreyer,T., 517 Fisher, M.J., 440, 455
Du Plessis, A., 437 Fjeldskaar, W, 105, 106, 107, 161, 162, 166, 175, 176, 184, 188,
Du Rouchet, J.H., 305, 323 190, 191,195, 481, 493
Duddy, I.R., 162 Fleming, H.S., 166,185
Duffaud, E, 5,16 Flood, B., 488, 493
Duindam, P., 80, 385, 395 Flood, S.B., 184
Dunham, J.B., 349, 352 Flores, G., 425, 426, 428, 433, 437
Dunn, M.E., 307, 323 Foerseth, R., 395
Durand, B., 323, 456 Forster, R., 426, 437
Dutton, C.E., 187,195 Forsyth, D.W., 85, 90
Dutton, S.P., 522, 526, 528 Foscolos, A.E., 156,162
Dypvik, H., 184, 323, 340, 483, 489, 492, 493, 517 Fossen, K , 243, 244, 245, 246, 247, 248, 251, 277, 288
534 References index

Foucher, J.-R, 77, 80 Gowers, M.B., 2, 4, 16, 16, 80, 213, 220, 440, 441, 455, 495, 499,
Fowler, R.M., 300, 302 501, 506
Fowler, S.R., 82, 294, 302, 467 Graciansky, P.C., 469, 479
Frandsen, N., 455, 495, 506 Gradijan, S.X, 208, 220, 300, 302
Frank, F C , 327, 340 Graham, R.H., 57, 59, 64, 80
Frser, AX, 202, 220 Graham, S.A., 344, 353
Frser, HJ., 329, 336, 340 Grans, P.B., 267
Frederickson, E.A., 521, 528 Grantz, A., 188,195
Freeman, B., 304 Gratier, J.R, 327, 340
Freund, R., 237, 240 Gray, D.I., 161,162
Friend, RE, 106, 482, 493 Gray, D.R., 337, 341, 469, 479
Froget, C, 78, 80 Gray, K.E., 529
Froitzheim, N., 470, 479 Green, D.H., 6,16
Fchtbauer, H., 313, 323 Gregersen, O., 184
Fugelli, E.M.G., 183,184 Gregersen, S., 16
Gregory, AR., 267
Gr0nlie, A., 81, 83, 84, 85, 90
Gabrielsen, R.H., 49, 53, 55, 76, 80, 83, 89, 109, 110, 112, 115,
Groshong, R.H. Jr., 9, 14, 15,16
119, 121, 122, 123, 125, 130, 154, 162, 174, i<34, 333, 340,
Groshong, R.H., 101,106, 247, 251, 327, 340
381, 382, 394, 395, 462, 467, 507, 510, 517
Grosswald, M.G., 189,195
Gage, M.S., 199, 202, 220
Gudlaugsson, S.T., 81,106,119,130,151,162, 492
Gagnepain, X, 151, 302
Gueguen, Y., 325, 341
Gale, J.E., 278, 288
Guennoc, P., 467
Gangi, A.F., 179,184,184
Guidish, T.M., 253, 267, 421, 424, 426, 437
Gardner, L.W., 267
Guiguet, R., 327, 340
Garfunkel, Z., 303
Gnther, P.R., 162
Garing, T., 267
Gussow, W.C., 322, 323
Garrey, G.H., 303
Gustav, S.H., 300, 302
Gaskell, B.A, 493
Gustavson, T.C., 527, 528
Gawthrope, R.L., 105,106, 148, i 5 i , 221, 229, 299, 300, 303
Guyon, E., 340, 342
Gayer, R.A., 479, 482, 493
Gdula, J.E., 297, 300, 303
Hadizadeh, J., 337, 341
GECO, 441, 455
Haenel, R., 64, 82
Gee, D.G., 199, 220
Hagermann, T.H., 481, 489, 493
Geological Survey Office of Sweden, 170
Hagevang, T, 394
Gephart, J.W., 85, 90
Haimson, B.C., 519, 528
Gerdes, K., 105,106
Halbouty, M.T., 117, 118,119
Gibbs, AD., 44, 52, 53,105,106,197,198,199, 205, 219, 220, 222,
Haldorsen, H.H., 355, 363
229, 235, 240, 246, 248, 249, 251, 269, 276, 427, 437, 467
Hall, S.H., 341
Gibler, P.R., 302
Halpern, H.I., 81,152
Gibson, J.R., 269, 276, 293, 299, 302
Halstead, P.H., 43, 53, 205, 220, 390, 391, 395
Gidon, M., 479
Ham, W.E., 520, 528
Gidskehaug, A., 119
Hamar, G.R, 42, 80, 89,119, 395
Giles, M.R., 417, 419
Hambrey, M.X, 57, 81
Gillcrist, R., 470, 479
Hamilton, W, 105,106, 293, 303
Gilmour, P., 380
Hancock, P.L., 84, 85, 90, 219
Giltner, J.R, 43, 44, 51, 52, 53, 457, 462, 464, 465, 466, 467 Hanisch, J., 81
Gjelberg, I., 161
Hanken, N.M., 492
Gjelberg, J.G., 106, 111, 119, 483, 493, 494, 509, 517 Hansen, AI., 492
Gjelsvik N., 1S5 Haq, B.U., 4,16, 44, 46, 53, 15, 81
Gjessing, X, 184 Hardenbol, J.,16, 53, 81
Gjevik, B., 179, iS4 Harding, T.R, 105,106, 113,119, 265, 267, 325, 328, 341, 382, 388,
Gleadow, A.J.W., 156,162 389, 391, 392, 395, 428, 437
Glennie, K.W., 439, 455, 498, 499, 506 Hardman, R.F.R, 42
Glezen, W.H., 455 Haremo, P., 483, 485, 486, 487, 489, 491, 492, 493
Gloppen, T.G., 152,162, 380 Harker, S.D., 298, 303
Gluyas, J.G., 413, 419 Harland, W.B., 93, 94,106, 127,130, 481, 482, 483, 487, 489, 492,
Gobbett, DJ., 106 493
Gobett, D.S., 493 Harlton, B.H., 521, 526, 528
Goff, XG, 444, 455 Harper, A.S., 323, 340, 517
Golden, M., 54 Harper, T.R., 278, 280, 281, 288
Goldschmidt-Rokita, A , 75, 80 Harris, J.R, 300, 302
Goldstein, A.G., 522, 528 Harrison, W.E., 528
Gonzales, E., 341 Hart, R.D., 380
Goodland, S.W., 437 Hartnady, C.J.H., 424, 428, 433, 437
Gough, D.I., 519, 527, 528 Hrvik, L , 288
References index 535

Hastings, D.S., 78, 81, 323, 340, 517 Irwin, H., 517
Haxby, W.F., 15,16, 82 Isaacs, CM., 345, 353
Hay, J . T C , 389, 390, 391, 395 Isaksen, D., 41, 42
Hazeldine, R.S., 111, 119 Ito, K., 6,16
Heafford, A.R, 493 Iwai, K., 288
Heald, M.T., 515, 517
Healy, J.H., 525, 527, 528 Jackson, J., 296, 300, 303
Heck, R.G., 353 Jackson, J.A., 14, 16, 105, 106, 148, 151, 241, 276, 290, 293, 297,
Heiland-Hansen, W, 494 299, 300, 302, 303, 469, 479
Helle, M., 220, 251 Jackson, M.P.A., 114, 117, 118,119
Hellem, T., 293, 302 Jackson, P.A., 119,128,131
Hellinger, S.J., 43, 46, 47, 53, 221, 229, 463, 465, 467 Jacobsen, F , 2,16, 499, 506
Helwig, X, 352 Jacobsen, H.K., 25, 42
Hempton, M.R., 478, 522, 525, 526, 528 Jacobsen, H.R, 91, 94, 99, 101, 107, 109, 112, 119, 122, 131, 133,
Henden, J., 220, 251 152
Henrich, R., 184,195 Jacobsen, K.H., 42
Henriksen, E., 159,162,185 Jacobsen, V.W., 22, 39, 41, 42, 395
Henry, G.E., 520, 528 Jacquart, G., 106,119,130,151,162
Herda, H.H., 379, 380 James, A.V., 54
Hermans, L., 323 Jamieson, T.F., 187,195
Herrick, C.G., 519, 528 Jamtveit, B., 59, 81, 180,184
Heum, O.R., 311, 323 Jankhof, K., 293, 303
Hickman, S.H., 519, 528 Jansen, E., 163,184, 188,195
Hinz, K., 59, 81, 121,130 Jaritz, W, 117,119
Hirschleber, H.B., 80 Jarvis, G.T., 46, 53, 457, 458, 459, 462, 465, 467
Hitchen, K., 78, 57,385,395 Jarvis, R.J., 128,131
Hjelle, A., 488, 493 Jeans, R., 106
Hobson, G.D., 305, 306, 323 Jensen, L.N., 107, 109, 110, 114,119, 121, 122,125, 127,128,130,
Hoeppener, R., 240 162,184,185, 493
Hoffman, P., 520, 528 Jensen, P.K., 4, 7,16
H0ie, A., 517 Jensen, T.F, 455, 495, 501, 506
Hollander, N.B., 297, 302 Jenyon, M.K., 110, 117,119, 123, 128,131
Holliday, D.W., 106, 493 Johannesen, P., 9,16
Holliger, K., 462, 463, 465, 467 Johannessen, E., 152,162
Hollinger, K., 53 Johansen, H., 492, 493
Holloway, N.H., 304 Johnson, C.F., 522, 525, 528
Hollywood, J., 81,152 Johnson, G.D., 492
Holm, L , 455, 506 Johnson, G.L., 195
Holtedahl, O., 482, 493 Johnson, H.D., 208, 220
Home, P.C., 162,185, 409, 413, 420 Johnson, K.S., 520, 528
Hooker, V.B., 522, 525, 528 Johnson, R.J., 199, 205, 219, 220
Horsfield, W.T., 236, 238, 240, 483, 489, 493 Johnson, R.L., 385, 395
Horvath, F., 467 Johnstone, D.W., 437
Hospers, J., 44, 49, 52, 53 Jones, M., 42
Hossack, J.R., 59, 81, 201, 220, 269, 276, 367, 380 Jones, M.E., 327, 341
House, W M , 337, 341 Jones, R.W, 305, 323
Houseknecht, D.W., 326, 327, 341 Jordan, J.R., 155,162
Houseman, G.A., 59, 80, 106,107, 302 Jordan, L., 520, 528
Housemand, G., 151
Howard, G.C., 5 i 7 Kaiheim, J.E., 119,130,162,184
Huang, J., 106 Kalthoff, E., 240
Hubbert, M.K., 325, 326, 328, 341, 448, 455 Kamen-Kaye, M., 425, 437
Huber, M.I., 247, 249, 251 Kapustin, I.N., 119
Hudson, J.A., 365, 371, 377, 379, 380, 525, 528 Karlsson, W, 462, 467
Hughes, T.J., 189,195 Karner, G.D., 94,106, 229
Hunt, J.A., 437 Kastner, M., 345, 352
Hurst, J., 106 Katz, B.J., 155,162
Hurst, J.M., 151 Kaufman, S., 106
Hutchinson, I., 276 Kautz, S.A., 237, 239, 240, 241, 247, 250, 251
Hvoslef, S., 307, 323 Keen, C.E., 47, 54, 221, 222, 229, 457, 458, 459, 465, 466, 467
Kehle, R.O., 529
Idil, S., 53, 467 Kellogg, H.E., 481, 489, 493
lilies, J.H., 389, 395 Kendall, CG.St.C, 267, 437, 455
Ingle, J.C. Jr., 344, 353 Kennedy, G.C., 6,16, 181,184
Inglin, H.F., 357, 363 Kennett, B.L.N., 12,16
Insley, M.W, 479 Kent, P.E., 434, 437
536 References index

Khaleel, R., 365, 368, 369, 371, 379, 380 Lehner, F., 342
Kidd, W.S.F, 478 Lehner, F.K., 305, 306, 323
Kihle, R., 437 Lemoine, M., 469, 470, 479
King, G.C.P., 106,151, 293, 294, 302, 303, 355, 359, 363 Lemos, J.V., 379, 380
Kirkpatrick, S., 365, 380 Leonard, A.J., 419
Kirton, S.R., 385, 395 Lerche, I., 267, 429, 437, 439, 444, 455, 456
Kiryukhin, L.G., 119 Lervik, K.S., 49, 52, 54
Kitahara, S., 6,16 Leutz, W.K., 42
Kjemperud, A., 105,106,191,195, 302 Lewis, C.R., 254, 267
Kj0de, X, 110,112, 9 Leythaeuser, D., 305, 323
Kleinspehn, K., 494 Liest0l, O., 184
Klemperer, S.L., 44, 52, 53, 54, 457, 462, 463, 464, 465, 467 Lilleng, X, 517
Kligfield, R., 269, 274, 276, 479 Lillie, R.J., 492
Klint Jensen, P., 443, 455 Lind, E.K., 109,119,131, 493
Klomp, U . C , 310, 323 Lindgreen, H., 443, 455
Kl0vjan, O.S., 119 Lippard, S., 463, 466, 467
Knarud, R., 493 Lippard, S.J., 112,119
Knipe, R.J., 199, 220, 327, 329, 332, 334, 336, 339, 341, 474, 479, Liu, G., 463, 466, 467
511, 517 Livbjerg, F , 493
Knopoff, L., 337, 342 Livera, S.E., 297, 300, 303, 355, 363
Knutsen, S.M., 162,185 Livshits, Y.Y., 481, 493
Kodama, H., 156,162 Ljones, T.E., 217, 220
Koestler, A.G., 277, 287, 288, 333, 340, 507, 517 Ljosland, E., 251, 288
Kohlstedt, D.L., 293, 302 Llewellyn, P.G., 130
Kozak, X, 327, 341 Lloyd, G.E., 327, 329, 334, 336, 341
Krantz, R.W., 389, 392, 395 Logan, J.M., 327, 341
Kranz, R.L., 327, 341 Long, J.C.S., 365, 366, 368, 369, 371, 375, 377, 379, 380
Kringstad, L., 184,195 L0nne, 0., 184
Kristoffersen, Y., 165, 189,195 Lorben, P.M., 107
Krokan, B., 110,114,119 Lorber, P.M., 152
Krooss, B.M., 305, 323 Lovell, J.P.B., 78, 81
Krylov, N.A., 119 Lovering, J.F, 162
Kusznir, N.J., 54, 197, 220, 293, 294, 295, 303, 495, 506 L0V0, V, 169,184
Lowell, D.L., 483, 489, 493
La Pointe, P.R., 365, 371, 379, 380 Lucas, S., 341
LaBreque, J.L., 428, 437 Lucazeau, F , 13,16
Lafourcade, P., 425, 437 Lundberg, N., 341
Lake, S.D., 94,106 Lunde, G., 59, 76, 80
Lamar, D.L., 94,106, 482, 493 Lynch, H.D., 14,16
Lamb, TJ., 522, 525, 527
Lambiase, J.X, 107,152 Macdonald, C.J., 363
Lamy, J.M., 49, 54, 59, 81, 228, 229, 385, 395 Macintyre, R.M., 42
Landisman, M., 17 MacKay, T.A., 220
Langseth, M., 436 Mackenzie, AS., 323
LaPoint, P., 352 Macurda, D.B. Jr., 146,152
Larsen, B.T., 80, 81 Maddock, R.H., 341
Larsen, H.C., 57, 81 Madsen, L., 1,17
Larsen, K.-B., 162,185,195 Magara, K., 253, 254, 267, All, 437
Larsen, PH., 300, 303 Mher, C.E., 298, 303
Larsen, VB., 57, 81, 298, 299, 301, 303, 483, 493 Mher, H.D. Jr., 489, 493
Larson, R.L., 248, 251 Major, H., 483, 493
Larson, R.R., 345, 353 Makurat, A., 282, 288, 380
Larter, S.R., 323 Malod, J., 437
Last, N.C, 278, 280, 281, 288 Maltman, A.X, 511, 517
Lateltin, O., 479 Manby, G.M., 57, 81
Latil, M., 312, 323 Mandl, G., 129,131, 238, 239, 241, 342
Latouch, C , 80 Mann, D.M., 323
Laughton, A.S., 188,195 Mansinha, L., 290, 303
Lauritzen, ., 488, 493 Manum, S.B., 153,162, 168,184, 481, 492, 493
Lawver, L.A., 424, 437 Marcher, M.V., 528
Le Douaran, S., 13,16 Marchini, D., 528
LePichon,X., 60, 0,52,50 Marcussen, C., 81
Leader, M.R., 221, 229 Mareschal, J.C., 181, 184,184
Lebesbye, E., 162,185,195 Marsal, D., 323
Lee, C.K., 179,184 Marsden, G., 47, 54, 198, 220, 294, 295, 303, 506
Leeder, M.R., 105,106, 148,151, 299, 300, 303 Martin, A.K., 424, 428, 437
References index 537

Martin, J.H., 361, 363 Morton, A.C., 76, 81

Martinsen, B., 22, 42 Morton, N., 49, 54
Massie, I., 251, 288 Morton, W.H., 239, 241
Mastin, L.G., 519, 528 Moss, D., 529
Mastin, R.L., 529 Mougenot, D., 424, 426, 428, 430, 437
Matheron, G., 361, 363 Mozetic, A., 295, 302, 311, 323
Matishov, G.G., 162 Muller, D., 479
Maton, P.I., 106, 493 Mungnier, J-L., 479
Matthews, D.H., 106, 219 Murata, K.J., 345, 353
Matthews, R.K., 188,195 Murray, I.B., 120, 267
Mawer, C.K., 327, 341 Mutter, J . C , 55, 56, 59, 60, 69, 80, 81, 248, 251
McAuliffe, CD., 305, 323 Myers, W.B., 105,106, 293, 303
McBee, W. Jr., 521, 528 Myhre, A.M., 76, 80, 81, 103,106, 158,162, 395, 492
McBride, E.F., 326, 327, 342 Myhre, L., 19, 42
McBride, J.H., 106 Mykkeltveit, S., 16
McClay, K.R., 59, 81, 238, 239, 241, 247, 250, 251, 367, 380, 458,
467, 469, 479 Nagtegaal, P.J.C, 515, 517
McGeary, S., 106, 197, 200, 219, 220 Nagy, J., 184, 483, 492, 493
McGill, G.E., 237, 241 Nakayama, K., 436, 437, 439, 455
McGill, J., 107,152 Nansen, F., 153,162, 187,195
McGookey, D.A., 522, 528 Nardin, T.R., 146, 148,152
McKenzie, D.A, 1, 5, 6,17, 46, 47, 52, 53, 54, 60, 64, 81, 82, 105, Nash, D.F., 220
106, 221, 222, 226, 229, 241, 272, 276, 294, 296, 297, 299, Naterstad, J., 42
302, 303, 310, 313, 323, 424, 430, 431, 432, 433, 434, 436, Navrestad, T., 131
437, 457, 458, 459, 460, 462, 465, 467, 503, 506 Needham, D.T., 199, 220
McLean, R., 521, 526, 528 Neglia, S., 305, 323
McMechan, M.E., 469, 479 Neher, K., 522, 526, 528
McNutt, M.K., 222, 229 Nelson, K.D., 94,106
McWhae, J.R.H., 482, 493 Nelson, P.H.H., 47, 54, 59, 81, 228, 229, 385, 395
Megard-Galli, J., 479 Nelson, R.A., 237, 241, 325, 341, 399, 407, 507, 515, 577
Meinke, L., 437 Nemec, W, 507, 508, 5i 7
Meisingset, K.K., 323 Nesje, A., 169, 170, 176, 177,185, 190,195
Mercier de Lepinay, B., 478 Nesteby, H., 492
Merritt, CA., 528 Newall, M.J., 197, 220
Merzer, A.M., 237, 240 Newmark, R., 16
Miallet, N., 80 Nssen, I.R., 219
Michelsen, O., 2, 4, 11, 16,17, 441, 450, 455, 498, 499, 501, 504, Nipen, O., 53, 219, 467
506 Nogaret, E., 323, 456
Midbe, P.S., 493 Norton, I., 437
Middleton, M.F., 6,16 Norton, M.G., 50, 54, 59, 81, 367, 380, 467
Miller, E.L., 263, 267 N0ttvedt, A., 153,162, 163, 166,185, 489, 493, 494
Mills, S.J., 152 Nygaard, E., 16, 455
Milnes, A.G., 288 Nyland, B., 107, 117, 119, 163,185, 481, 493
Milton, M.J., 272, 276
Milton, N.J., 49, 53, 219, 226, 228, 229, 294, 297, 299, 302 O'Brien, J., 267
Mitchell, A.W., 219 O'Connell, R.J., 15,17, 178, 179,185
Mitchum, R.M., 54,147,152 Ockendon, J.R., 82
Mitra, S., 327, 341, 469, 474, 476, 479, 490, 493, 515, 517 Odling, N.E., 46, 54
Moe, A., 131 Oertel, G., 237, 241
Mogensen, T.E., 504, 506 Ofstad, K., 184, 395
Mokhtari, M., 75, 81 Ogle, B.A., 343, 353
M0ller, J.J., 2, 3, 4,10,17, 440, 441, 455, 495, 496, 502, 506 Ohnaka, M., 327, 341
Molnar, P., 294, 302 Olaussen, S., 153,162
Montadert, L., 467 Olausson, S., 152
Moody-Stuart, M., 482, 493 Oliver, J.E., 106
Moore, J.C, 327, 332, 341 Olsen, J.C, 16,455
Moore, S., 81 Olsen, J.S., 303
Moretti, I., 7,13,17 Opdyke, N.D., 188,195
Morgan, K.A., 195 Ormaasen, D.E., 24, 42
Morgan, K.M., 527, 529 Ormaasen, E., 205, 219
Morgan, P., 14,16 Orvin, A.K., 481, 482, 489, 493
Morin, R.H., 522, 528 Osborne, R.E., 120
Morizot, H., 220 Otha, Y., 165
Mrk, A , 483, 493 Oudin, J.L., 323, 456
Morley, C.K., 469, 479 0vreb0, O.K., 121, 121,131, 302
Mortimer, J., 302 Owen, G., 327, 341
538 References index

0xnevad, I.E., 517 Ratcliffe, N.M., 94,107

Oxtoby, N.H., 419 Rattey, R.P., 392, 395
Ravn, J.P.J., 483, 493
Pallesen, S., 176,184 Rawson, RE, 385, 395
Papadimitriou, P., 151, 302 Reading, H.G., 119
Park, R.G., 293, 303 Reches, Z., 85, 90, 237, 241, 389, 391, 392, 394, 395
Parker, J.R., 106, 481, 483, 487, 489, 492, 493 Redman, R.H., 521, 528
Parry, C C , 226, 229 Redwine, L.E., 350, 353
Parry, WT., 302 Reed, W.E., 106, 482, 493
Parson, L.M., 76, 81 Reilinger, R., 293, 303
Parsons, B., 222, 229, 437 Reither, E., 195
Patriat, P., 428, 437 Remer, J.S., 380
Payrazyan, V.V., 119 Requejo, A.G., 78, 81,147,152
Pedersen, O.E., 307, 323 Reuschle, T., 341
Pedersen, T., 60, 76, 81 Reynolds, D J . , 107,152
Pederstad, K., 243, 251 Reynolds, S.J., 232, 241
Pegrum, R.M., 81,199, 202, 205, 213, 217, 219, 220, 496, 498, 506 Rhys, G.H., 24, 42
Perkins, P.J., 106, 493 Rice, A.H.N., 110,113,119, 479
Perrin, J.E, 323, 456 Richards, P.C., 202, 220
Perry, J. Jr., 470, 477, 479 Richardsen, G., 159,162,185
Perry, R.K., 166,185 Richter, C.F., 293, 303
Petterson, O., 243, 244, 247, 249, 251, 277, 288 Richter-Bernburg, G., 128,131
Pheifer, R.N., 162 Rider, M.H., 267
Piasecki, S., 81, 395 Riis, E, 59, 76, 80,105,107,119,127,130,158,159,161,162,163,
Pickering, K.T., 149,150,152 166, 169, 183,184, 381, 385, 392, 395, 481, 493
Pickford, P.J., 267 Riley, L.A., 385, 395
Pickton, C.A.G., 130 Rindstadt, 83
Piggot, A.R., 380 Ringseth, N., 489, 490, 493
Pilaar, W.F., 342 Ringwood, A.E., 6,16
Rippon, J.H., 302
Pine, R.J., 278, 281, 288
Ritchie, J.D., 78, 81
Pisciotto, K.A., 345, 353
Roaldset, E., 185
Pittman, E.D., 325, 335, 336, 341, 507, 517
Robert, E, 341
Planke, S., 55, 56, 59, 60, 63, 68, 69, 81
Roberts, A., 54, 71, 82,131, 219, 220, 303
Plumb, R.A., 519, 528
Roberts, A.M., 292, 295, 296, 301, 303
Pollard, D.D., 522, 528
Roberts, D., 81, 83, 85, 90, 94,107,112,119,198, 220, 479
Poulsen, K.H., 341
Roberts, D.G., 467
Powell, R., 6,17
Roberts, J.E, 528
Powell, T.G., 162
Roberts, S., 299, 300, 303
Pratt, J.H., 187,195
Robinson, C , 76, 80, 381, 395, 510, 517
Precious, R.G., 342
Robinson, N.D., 57, 59, 66, 80
Presley, M.W., 522, 528
Rodgers, D.A., 522, 525, 526, 528
Prestholm, E., 507, 509, 515, 517
Roehl, P.O., 350, 353
Preston, R.M.F., 327, 341
Roeske, S., 341
Price, I., 392, 395, 419
Rognebakke, A., 220, 251
Price, J.D., 53, 219, 251, 302, 303, 394, 467
Rokoengen, K., 166,185
Price, L.C., 305, 323
Roll, A., 60, 81
Price, N.J., 350, 353
Rolle, F., 81, 395
Price, R.A., 469, 477, 478, 479 R0nnevik, H.C., 16, 59, 80, 81, 89, 91, 94, 99, 101, 107, 109, 112,
Pringle, I.R., 107
119,121, 122, 125,131, 133,152, 159,162, 395
Proffett, J.M. Jr., 239, 241
R0nningsland, T.M., 166,185
Root, 371, 380
Quigley, T.M., 323 Rose, S.C., 254, 267
Rosendahl, B.R., 105,107, 148,152
Rabinowitz, P.D., 424, 428, 437 Roux, S., 340
Rad, N.S., 288 Rowan, M.G., 269, 274, 276
Rambech Dahl, C , 53, 219, 251, 302, 394, 467 Rowley, D.B., 47, 54, 459, 465, 467
Ramberg, I.B., 83, 89, 389, 394, 395 Royden, L., 47, 54, 11, 81, 221, 222, 229, 424, 430, 431, 432, 433,
Ramberg-Moe, H., 106,161, 168,185 434, 435, 436, 437, 457, 458, 459, 465, 466, 467
Ramsay, D.M., 107 Rudkiewicz, J-L., 479
Ramsay, J.G., 247, 249, 251, 350, 353, 470, 474, 479 Rudolph, K.W., 213, 219
Ramsey, W, 187,195 Rundberg, Y., 45, 54
Ransome, F.L., 296, 303 Rundle, J.B., 293, 303
Rasmussen, A., 119 Ruppel, S.C., 528
Rasmussen, E., 162,185 Rutter, E.H., 327, 337, 340, 341, 515, 517
Rasmyshlyayev, A.A., 119 Rye, N., 185,195
References index 539

Rye-Larsen, M., 494 Simpson, R.D.H., 229

Skagen, J.I., 107,119,185, 493
Saeland, G.T., 243, 246, 251 Skarpnes, O., 42,107,119,185, 493
Sagingaliyev, A.B., 119 Skogseid, J., 55, 56, 57, 59, 60, 62, 69, 75, 76, 80, 81
Sahagian, D., 47, 54, 459, 465, 467 Skottheim, S., 220
Saigal, G.C., 511, 517 Sladen, C.P., 151
Sales, J., 185 Sleep, N., 222, 229
Sand, M., 165 Slettemark, B., 184,195
Sanderson, D.J., 528 Slevin, A., 424, 437
Sandiford, M., 6,17 Smalley, P.C., 45, 47, 54
Sangree, J.B., 152 Smith, A.G., 130
Saroglu, E, 478 Smith, D., 522, 527
Savage, W.Z., 527 Smith, D.A., 325, 326, 328, 339, 341
Sawyer, D.S., 60, 81, 221, 229 Smith, M.P., 91,106,107
Saeb0e, A., 2, 4,16, 213, 220, 440, 441, 455, 495, 498, 499, 501, 506 Smith, R.M., 54
Scarpa, R., 6,16 Smithson, S.B., 389, 395
Schaefer, R.G., 323 Smylie, D.E., 290, 303
Schamel, S., 246, 251 Smythe, D.K., 200, 219, 385, 395
Schedl, A.H., 474, 477, 478, 479 Snyder, D.B., 200, 220
Scheidegger, A.E., 368, 380 Snyder, WS., 350, 353
Schlter, H.-U., 121,130 Solheim, A., 158,162,184, 189,195
Schmidt, B.J., 162 S0rensen, K., 109, 114,119,120, 121, 122, 125, 127, 128,130
Schmidt, C.J., 470, 477, 479 S0rensen, S., 41, 42, 199, 220
Scholz, C.H., 379, 380 Soufleris, C., 151, 302
Sch0neich, H., 16 Speksnijder, A., 357, 360, 363
Schoonmaker, J., 341 Spence, G.D., 82
Schowalter, T.T., 305, 323, 325, 326, 335, 341 Spencer, A.M., 54, 55, 81, 158, 159, 162, 166, 185, 298, 299, 301,
Schrader, P., 240 303, 409, 419, 420
Schubert, G., 294, 303 Spencer, J.E., 232, 241
Schunk, D.J., 162 Stahl, W, 78, 80
Schuyleman, S.F., 208, 219, 224, 229, 462, 467 Standing, M.B., 305, 323
Schwartz, D.P., 300, 303 Stangenes, J., 131
Schwarz, H-.U., 240 Stauffer, D., 370, 373, 380
Schwarz, P.H., 42 Steckler, M.S., 6, 7, 13,17, 45, 54, 59, 72, 81, 229
Sclater, J.G., 1, 3,17, 43, 46, 47, 53, 54, 59, 64, 81, 129,131, 155, Steel, R., 162,185, 201, 220, 380, 467
162, 221, 222, 224, 229, 237, 239, 240, 241, 247, 250, 251, Steel, R.J., 53, 94,107, 111, 120, 168,185, 361, 380, 481, 482, 483,
254, 256, 267, 271, 276, 421, 436, 437, 457, 462, 467 489, 492, 493, 494, 517
Scotese, C.R., 424, 437 Steele, R.P., 220
Scott, D.L., 105,107,152 Stegena, L., 467
Scott, J., 288 Stein, R.S., 290, 291, 294, 303
Scudder, D.M., 195 Stemmerik, L., 81, 395
Scull, B.J., 41, 42, 409, 419 Sterns, D.W., 347, 353, 521, 526, 528
Segall, P., 522, 528 Stewart, D J . , 220
Segnor, A.M.C., 478 Stewart, I.J., 219
Segoufin, J., 424, 428, 437 Stock, J.M., 525, 527, 528
Seguret, M., 50, 54 Stoffa, PL., 81
Sejrup, H.P, 189,190,195 St0lan, T, 119
Seland, R., 184 Stone and Webster Engineering Corporation, 527, 528
Sellevoll, M.A., 80, 81 Storetvedt, K.M., 119
Semken, S., 467 Storli, A., 251, 288
Seni, S.J., 114, 117,118,119 Stouge, S., 441, 455
Seranne, M., 50, 54 Stow, D.A.V., 147, 148,152
Serra, S., 237, 241 Strand, K., 492
Shackleton, N.J., 188,195 Str0m, K.M., 185
Shanmugan, G., 228, 229 Stromquist, A.W, 237, 241
Shaw, N.D., 80, 395 Sturt, B.A., 94,107, 199, 220
Shay, J.T, 352 Sugden, D.E., 172,185
Shelton, J.W., 237, 241 Sundvor, E., 80,106,162, 395
Shirley, O.J., 155,162 Sunwoo, C , 355, 359, 364
Shorey, M.D., 129,131 Suppe, J., 249, 251, 469, 478, 479
Sibley, D.F., 515, 517 Surlyk, E, 57, 59, 81, 147, 148, 150,152, 385, 395
Sibson, R.H., 327, 329, 333, 337, 338, 341 Sutherland, D., 190,195
Sibuet, J.-C, 60, 80, 81 Svela, P.T., 219
Siedlecka, A., 220 Sverdrup, E., 509, 515, 517
Simpson, E.S.W., 428, 437 Swallow, J.L., 247, 251
Simpson, G.S., 243, 246, 251 Swanson, M.T., 94,107
540 References index

Sweeney, J.F., 195 Vejbaek, O.V, 1, 2, 4, 5, 9, 17, 202, 216, 220, 439, 440, 441, 455,
Swensson, E., 492 456, 495, 506
Swolfs, H.S., 527 Vendeville, B.C., 237, 238, 239, 241, 247, 250, 251
Vernon, R.H., 340
Takenouchi, S., 16 Vially, R., 44, 51, 52, 54, 463, 464, 466, 467
Talbot, C.J., 117,119,128,131 Vienot, M.E., 407
Talleraas, E., 121,131 Vierbuchen, R.C., 119
Talwani, M., 11,17, 59, 68, 71, 75, 81, 82,166,185 Vik, G., 288, 380
Tankard, A.J., 458, 459, 467 Villars, F., 476, 477, 479
Tardy, M., 478 Villemin, T.F, 355, 359, 364
Tarkard, A.J., 264, 267 Virieux, X, 80,151,302
Taylor, E., 80 Virlogeux, P., 437
Taylor, R.R, 342 Vita-Finzi, C , 106, 293, 303
Tchalenko, S., 428, 437 Vizgirda, X, 266, 267
Teufel, L.W., 407, 519, 528 Vogt, T, 482, 494
Thiede, X, 80 Vollbrecht, A., 240
Thomas, L.K., 399, 407 Vollset, X, 24, 32, 42, 80, 81, 409, 420
Thomas, W.A., 477, 479 von Herzen, R.P., 81,437
Thompson, A.F., 515, 517 von Schonfeldt, H.A., 522, 525, 529
Thompson, G.A., 129,131 Voorhoeve, H., 106,107
Thompson, S., 54 Vorren, TO., 107,119,153, 156,158,159,162,163,185, 189,195,
Thomsen, E., 81, 395, 443, 455 493
Thomsen, R.O., 443, 444, 455 Vullstad, A.A., 184
Thome, J.A., 5,10,17,12, 82
Throndsen, T, 153,162,168,184 Walker, J.M., 217, 220
Thunvik, R., 365, 371, 373, 380 Wall, V.J., 340
Tiller, G., 361, 364 Wallace, R . E , 293, 303
Tissot, B., 424, 430, 431, 437, 448, 455 Wallis, R.H., 106, 493
Tissot, B.P., 305, 314, 323 Wallis, WS., 353
Tjelland, T, 517 Walsh, J.J., 237, 241, 249, 251, 276, 299, 302, 303, 363
Todd, R., 32, 41, 42 Walters, R., 130
Tomlinson, L.W., 521, 528 Wandaas,B., 219
Tonstad, K., 41, 42 Wandas, B., 506
Torudbakken, X, 219 Wang, C.H., 327, 342
T0rudbakken, B., 80, 506 Wang, J.S.Y., 288
Torvanger, O., 394 Waples, D.W., 444, 448, 456
Toth, DX, 267, 437 Wardlow, N.C., 326, 342
Townsend, C , 94,107,119, 469, 479 Warner, M.R., 106, 219, 220
Toxwenius, B., 441, 455 Warpinski, N.R., 407
Tricart, P., 469, 470, 479 Warren, 371, 380
Trondsen, T, 481, 492, 493 Warrington, G.J., 54, 178,179,185
Trmpy, R., 469, 470, 479 Wasserburg, XG., 15,17
Trusheim, E, 111, 117, 118,120 Waterson, X, 363
Tullis, T.E., 327, 341 Watterson, X, 237, 241, 249, 251, 276, 299, 302, 303, 329, 342
Tuminas, A.C., 325, 328, 339, 341 Watts, A.B., 1, 5, 6, 10, 17, 45, 54, 59, 72, 81, 82, 221, 222, 229,
Turcotte, D.L., 16, 60, 82, 303, 327, 340 294, 304
Turner, C G , 303 Watts, N.L., 325, 326, 335, 339, 342
Watts, T.R., 462, 467
Udintsev, G., 82,185 Weber, X, 121,130
Uldall, A , 11,17 Weber, K.J., 328, 339, 342
Underhill, J.R., 327, 342 Wei, Z., 439, 456
Ungerer, P., 305, 323, 439, 456 Weidick, A., 172,185
Welbon, A.I., 469, 470, 471, 472, 474, 479
Vgnes, E., 81 Welsink, H J . , 264, 267, 458, 459, 467
Vail, PR., 16, 32, 41, 42, 46, 53, 54, 81 Weite, D., 17
Van Couvering, J.A., 188,195 Weite, D.H., 305, 314, 323, 439, 448, 455, 456
Van Den Driessche, X, 240 Weite, D.M., 430, 431, 437
Van der Molen, L, 355, 359, 364 Wernicke, B., 43, 54, 105, 107, 129, 131, 231, 235, 239, 241, 244,
van Hinte, J.E., 45, 53 247, 251, 265, 267, 296, 304, 459, 467
Van Hoorn, B., 59, 80, 385, 395 Wesson, R.L., 327, 342
Van Kuyk, A., 323 Westbrook, G.K., 82
Van Veen, P., 395 Weyl, P.K., 515, 517
Vann, I., 269, 276 Wheatley, T.X, 300, 304
Vanney, X-R., 437 White, J.C, 327, 342
Veevers,XX, 428, 437 White, N., 44, 52, 54, 199, 220, 34, 457, 459, 465, 467
References index 541

White, N.J., 221, 229, 235, 241, 269, 273, 276, 290, 293, 297, 299, Woods, E.P., 236, 241
300, 303 Woodward, N.B., 469, 479
White, R.S., 76, 82 Worsley, D., 55, 80, 94,107, 111, 120, 168,184,185, 395, 481, 482,
White, S.H., 327, 341, 342 483, 493
White, W.A., 171,172,177,185 Worzel, J.L.,77
Whitley, P.K.J., 229 Wrang, P., 81, 395, 455
Whitman, R.R., 795 Wright, P.A., 310, 323
Whitten, CA., 293, 304 Wyllie, M.R.J., 255, 256, 267
Wickham, J.S., 520, 529
Widmier, J.M., 752 Xiao, H.-B., 249, 257
Wiik, M., 208, 220, 300, 302
Wiik, V, 420 Yamashita, T., 337, 342
Wilhelm, S.J., 527, 529 Yarzab, R.F., 267, 437, 455
Wilke, S., 325, 342 Yeats, R.S., 492
Willemin, J.H., 106, 251 Yielding, G., 54, 82, 129, 131, 198, 218, 220, 292, 295, 296, 297,
Williams, D.E, 188, 795 298, 299, 303, 304, 715
Williams, G., 269, 276 Young, R., 211, 220, 304
Williams, P.F., 327, 341 Yousuf, M., 492
Williamson, P.E., 248, 257 Ykler, M.A., 5, 77, 439, 456
Wilson, C.R., 380
Wilson, J.C., 326, 327, 342 Zamin, A.S.H., 492
Wilson, T.V, 515,577 Zheng, L., 106
Wiltschko, D.V., 474, 477, 478, 479 Ziegler, P.A., 1, 2, 10, 77, 22, 24, 25, 32, 39, 42, 43, 44, 49, 54, 51,
Winsnes, T.S., 493 59, 66, 80, 82, 94,107, 202, 217, 220, 307, 311, 323, 384, 385,
Witherspoon, P.A., 288, 365, 380 395, 440, 456, 462, 467, All, 479, 498, 501, 506
Wood, R., 1, 3, 7, 11,16,17, 43, 46, 53, 54, 224, 229, 294, 302, 457, Ziegler, W.H., 284, 255
463, 465, 467 Zoback, M., 527, 529
Wood, R.J., 272, 276 Zoback, M.D., 519, 528, 529
Woodbury, H.O., 117, 118,120, 265, 266, 267 Zoback, M.L., 519, 522, 525, 527, 528, 529
Woodcock, N.H., 83, 90, 327, 342 Zoth, G., 82
Wooding, R.A., 306, 323
 543

Subject index

active normal faulting, 290 Bj0rn0ya South area, 133

active rift systems, 141 Bj0rn0ya Basin, 133, 134
Adda field, 450, 454 Bj0rn0yrenna, 163, 165
Adventdalen Group, 482, 483 block fields, 169
Airy isostacy, 71, 465 block kinematics, 239
Albatross field, 156 Blod0ks Formation, 41
allochthonous Chalk, 41 Bod0 High, 56, 68, 69, 73, 76
Alpine compression, 216 gravity anomaly, 69, 75, 76
Alta High, 99 Borah Peak earthquake 1983, 290, 291
amount of extension, 240 borehole profiling, 398
And0ya, 156 Brae, 213
anelastic strain data, 406 Brae field, 148
antithetic faults, 234 Brae Fault, 295
Appalachians, 469, 477 Brae Formation, 300
Arctic Ocean, 101 Bravo dome, 519, 521, 523, 527
Are Formation, 307, 384 breakout-data analysis, 519
sta Graben, 19, 20, 28, 41 breccia, 347
arid conditions, 22 Brent field, 297, 300
Arne structure, 450, 454 Brent Group, 202, 205, 226, 297, 355, 356
Arne-Elin Graben, 2, 3, 441, 444, 450, 454, 496 Brent Province, 208
asthenosphere, 6, 175,190, 222 Brentskardhaugen conglomerate, 482, 483
asthenospheric thermal anomaly, 76 brittle lithologies, 348
Atlantic rifting, 197 Broom Formation, 357, 361, 362
Auk Shelf area, 218 Bruce-Beryl embayment, 205
autochthonous Chalk, 41 Bryne Formation, 505
Bryne Formation, 32
back-stripping tool, 288 burial history, 444
backbone, 377
backstripping, 5, 45, 46, 59, 421, 426 Caledonian Orogeny, 91, 94, 101, 384
balancing techniques, 249 Caledonian structural grain, 84, 385, 394
Balme, 470, 471, 474, 476 Caledonides, 199, 219
Baltica, 91 California, 343
Barents Sea, 91, 99, 101, 102, 103, 105, 106, 109, 121, 133, 153, carbon isotope analysis, 511
159, 163, 167, 192, 193, 269 Carolinefjellet Formation, 482, 483
Barents Shelf, 92, 93, 101 cataclastic flow, 327
basal detachment, 274 cement seals, 325, 328
base Cretaceous unconformity, 44 cementation patterns, 325
base Zechstein fault, 412, 413 Cenozoic basin, 76
base Zechstein structural grain, 411 Cenozoic subsidence, 1
Basin and Range Province, 105, 231, 290, 300 Cenozoic tectonic subsidence, 64, 66
basin modelling, 221, 305 Central Graben, 200, 301, 439, 441, 453
numerical, 221 Central Spitsbergen Basin, 483
basin architecture, 133 Central Trough, 202, 208, 213, 217, 498
bathymetry, 299 Chapin Wash Formation, 232
bed thickness, 350 characterization of reservoir, 277
Berlevg Fault, 94, 95, 99, 101, 102 Chartreuse, 472, 474
beta factor, 44, 47, 52, 53, 62, 64, 78, 79, 458, 463, 465 chert, 344, 349, 350
Billefjorden Fault Zone, 94, 481, 483 clay smearing, 355
Billefjorden Group, 482, 483 Claymore, 301
biogenic facies, 344 co-seismic elevation changes, 290, 291, 293
Bjarmeland Platform, 121, 122 coccolith, 344
Bj0rn0ya Basin, 94, 99, 101, 112, 133, 134, 269, 270 Coffee Soil Fault, 495, 496, 498, 503, 504, 505
Bj0rn0ya Trough, 153, 156 collapse seals, 325, 328
Bj0rn0ya Island, 103 compaction, 221, 222, 253, 256
544 Subject index

compaction of sandstone, 249 early Cenozoic extension, 59

concentric geomorphic pattern (Fennoscandic), 172 early Cenozoic breakup, 56, 59
conductivity, 254 East Africa, 133, 141, 148
conductivity-displacement relationships, 279 East African Rift, 105
connectivity, 377 East Bj0rn0ya Basin, 99
continent-ocean transitional region, 66 East Greenland, 385
continental separation, between Norway and Greenland, 55, 59, East Shetland Basin, 297, 300
62 East Shetland Platform, 205, 295, 300, 301
controls of fracturing, 348 Eastern Gullfaks high, 246
controls on synrift sedimentation, 133, 141 eclogite/gabbro system, 6
Corbel Thrust Zone, 472 effective elastic thickness, 294, 295
core, 191 Egersund sub-Basin, 19, 20, 28, 41, 201, 213
core data, 399 Egersund Formation, 32, 39
Corinth earthquakes, 292 Ekofisk Formation, 41, 397, 399, 402, 403
Cormorant field, 355, 357 Ekofisk field, 397
coupled shear flow test, 282 elastic dislocation model, 290, 293
coupled stress flow behavious, 281 elastic layer, 289, 294
crestal fault system, 415 elastic thickness (EET), 222, 226
Cretaceous), 133 effective, 222, 226
critical factors, 133 elastic plate, 222
crystal plasticity, 327 Elsterian, 193
Crystallinity Index (CI), 156 emplacement of lava during breakup, 64
curved fault traces, 236 erosion, 289, 294, 295, 297
cyclic deformation in fault zones, 325 erosion from drilling parameters, 155
Etive Formation, 357, 359, 361, 362
eustasy, 187,191
Danish Central Trough, 1, 2,15, 495, 498, 501, 503, 506
excess fluid pressure, 439, 450
Dc-exponent, 155 experimental physical models, 231
decollement zones, 481 extension, 421
decompressional melting, 76 late Jurassic-early Cretaceous, 57, 59, 66
deformation model, 381, 388 extension calculations, 258
deformational textures, 507 extension estimates, 240
degree of tilt, 257 extensional detachment, 243, 247
depositional models, 133,142,143,146, 149 extensional fault trends, 142
depositional surface, 133, 142 extensional fault-block shapes, 231
depth conversion, 270 extensional faulting, 133
depth maps, 139 extensional fractures, 346
depth-dependent extension model, 44 extensional terranes, 269
detachment, 123, 128, 130
detachment fault, 231, 240 Farsund Formation, 14, 15, 443, 503, 504, 506
detachment model, 247 fault block compartmentalization, 240
deviatoric stress tensor, 83 fault block terraces, 135
dextral movement, 28 fault breccia, 84
Dia Graben, 94, 99,101 fault dip, 258
Dia High, 99 fault displacement, 507
diagenesis of clay minerals, 156 timing of, 507
diagenetic dolomite, 345 fault geometry, 83, 264
diagenetic history, 345 fault parallel microstylolites, 510
diapirism, 265 fault pattern in map view, 232, 236
diatom, 344 fault pattern in vertical section, 233, 237
differential compaction, 148 fault penetration, 421
diffusive mass transfer (DMT), 327, 330 fault plane geometry, 338
dike intrusions, onshore, 25 fault plane striae, 83
displacement vector, 286 fault reactivation, 471, 474
distribution of fault throw, 355 fault seal, 325, 326, 361, 363
distribution of seal types along fault planes 337 fault size distribution, 355, 259
dolomite, 344, 349 fault-throw distribution, 363
domino fault model, 244, 246, 247, 250, 290, 295, 296, 298, 299 fault zone diagenesis, 334
domino faults, 235 faulting, 121, 221, 253, 389, 397
downward-shallowing faults, 233 and diagenesis, 509, 512, 515
downward-steepening faults, 233 processes, 325, 326
Draugen field, 295, 299, 307, 322 faults in sandstones, 507
ductility contrast, 349 Feda Graben, 1, 2, 3, 440, 444, 450, 496, 501, 502, 504, 506
Dunlin Formation, 249 Fennoscandian Plate, 384
dynamic aquifer, 417 Fenris Graben, 56, 62, 64, 66, 68, 99, 101, 102, 105
D0nna Terrace, 56 Fenris High, 99, 101
Subject index 545

Fensfjord Formation, 213 glacial isostasy, 187

Festningen Sandstone, 482, 483 glaciation, 163,187
field development, 355 glaciers, 188
field studies, 133, 141 graben development, 495
Fingerdjupet Subbasin, 156,157 Great Glen Fault, 202, 205
Finnmark Platform, 95,156 Greece, 133, 148, 300
Finnmark East area, 94, 99,101 Greenland, 91, 93, 94, 101, 105
Fiskebank Basin, 19, 20 Gregory rift of Kenya, 141, 148
Fiskebank Formation, 41 Grensen Nose, 2, 3
fission track analysis, 156, 157, 158, 159 Gulf of Suez, 133, 141, 148
Fjerritslev Fault Zone, 199 Gullfaks field, 243
Fjerritslev Formation, 2, 28, 501, 503 Gyda field, 213
Fladen Ground Spur, 205, 300, 301
Flekkefjord Formation, 39
Halibut Horst, 301
flexural cantilever model, 294, 295, 296
halokinesis, 109, 111, 113, 121
flexural deformation, 183
halokinetic model, 109, 117
flexural rigidity, 190, 222, 294
Halten Terrace, 61, 295, 307, 381, 388, 390, 392, 510
flexural/isostatic model, 290, 293, 295
Haltenbanken, 156, 189, 192, 307, 507, 509
fluid flow, 365, 448, 449, 450
Hammerfest Basin, 94, 101, 113, 134, 140, 153, 155, 156, 157, 159
along faults, 337
hanging wall, 222, 253
fluid flow/compaction model, 439
deformation, 243, 247
fluid pressure, 350
Hardegsen unconformity, 24
folding, 346
Harstad Basin, 102, 112
footwall, 150, 222, 229, 253
Haute Savoie, 470, 476
footwall cusp, 148
heat flow, 1, 6, 12, 321, 439, 447, 457
footwall uplift, 289
Hebgen lake earthquake, 293
foreland, 469, 477
Hecla Hoek, 482
central, 470
Heidrun Field, 381, 384, 385, 388, 392
western Alps, 469, 476
Hei Graben, 56, 68
Formation Microscanner (FMS) logging 397, 405
Helvetiafjellet Formation, 482, 483, 507
Forth Approaches Basin, 202
Heno Formation, 501, 503, 505
Fosen area, 85
Heno Plateau, 1, 2, 3, 444, 454, 496, 499, 501, 504, 505, 506
Fourier transform technique, 190
Hidra Formation, 41
fractural reservoir, 343, 350
Hidra Fault Zone, 414
fracture orientation, 345
hierarchic fault throw distribution, 355, 359
fracture patterns, 367
Highland Boundary Fault, 199, 202, 217
fracture permeability, 365, 377
histories in fault zones, 325
fracture processes, 327
Hitra fault zone, 205
fractured media, 365
fractures, 345 Hitra-Snsa Fault, 84
fracturing, 397 Hod Formation, 41
Franz Josef Land, 93 Horda Platform, 43, 44, 293, 301
freeboard, 191,192 horizontal-stress orientation, 519
French Alps, 469, 470 Horn Graben, 22, 501
friction sliding, 327 Hornsund Fault, 94,103
Fr0ya High, 307 horst-and-graben, 235
Fr0ya Formation, 295 hydraulic fracturing stress, 519
Fulla Ridge, 56, 62, 64, 68, 69 hydrocarbon accumulation, 266
Fulmar Formation, 301 hydrocarbon generation, 188, 351, 439
Fulmar-Clyde area, 213 hydrocarbon-habitat modelling, 305
hydrocarbon migration, 188,192, 351
gabbro-eclogite transition, 179 modelling, 305, 306
Galley, 301 hydrocarbon potential, 421, 436
Garlock Fault, 113 hydrocarbon trapping, 188,192
Garn Formation, 307, 314, 322 hydrocarbon traps, 197, 469
gas expansion, 153, 159 hydrocarbon-water contact, 192
Gassum Formation, 28
geochemistry, 443 Iapetus Ocean, 94
geodetic data, 290 in-situ stress, 519
geodynamic modelling, 1, 6 inclined simple shear, 273
geometric analysis, 244 inclined shear, 235
geometrical restoration, 270 Inge High, 2, 3, 444, 496, 503, 504, 506
geothermal effects, 254, 261 initial subsidence, 60
Gert Ridge, 2, 3, 444, 496, 502 initial uplift, 62, 77
Gertrud Graben, l, 2, 3, 441, 444, 450, 454 496, 501, 502, 504, 506 Inner Moray Firth, 202, 217
Gipsdalen Group, 482, 483 intergrated sedimentolgical and structural analysis 355
546 Subject index

inversion, 41, 481 magnetostratigraphy, 188

Cretaceous, 413 magnitude of erosion, 167
Tertiary, 413 Magnus field, 208, 221, 224, 228, 300
isobase map, 191 Magnus Sandstone Member, 221, 228, 300
isopack maps, 138 Mandal Formation, 409, 412
isostasy, 187,191, 294, 297 Mandal High, 2, 3, 495, 496, 502, 506
isostatic adjustment, 179,183 mantle viscosity, 190
map of mean hights, 170
Janusfjellet Formation, 482, 483, 508 Marietta Basin, 519, 520, 522, 525, 527
joint compressive strength, 280 Mas0y Fault Complex, 126,127, 128
joint deformation behaviour, 277 mass balance erosion-deposition, 167,170
joint mechanical apertures, 281 mass flow units, 147
joint roughness coefficient, 280 mass flow processes, 146
Jura mountains, 472, 474 master fault complexes, 122,130
Jurassic paleogeography, 495, 503 maturation, 439, 448, 454
Jurassic sand, 496 Maud Basin, 94
Jurassic extension, 197, 202, 412 McKenzie model, 43, 44, 458
Jurassic rifting, 462 Melke Formation, 307, 311, 322
juxtaposition seals, 325, 328 mesoscopic folds, 348
Jylland, northern, 161 mesosphere, 191
metamorphic Basement, 19
microfabrics, 325
Kaapvaal Craton, 424
Mid North Sea High, 2, 3, 9, 197, 202
Kapp Toscana Group, 482, 483
Mid Norway, 169, 170, 307, 381, 507
Karoo unit, 425
Midland Valley, 199
kinematic model, 247, 392
migration, 448
Knipovich Ridge, 101
Mikkel, 307
Kristiansund-Bod0 Fault Complex, 381, 392
Miller field, 213
Kupferschiefer, 24
model tectonic heterogeneties, 277
Kval0y Basin, 94,101, 102
modelling, 305
of uplift, 176, 177, 178, 181
Lame's parameter, 190 Mohns Ridge, 101
large-scale faults, 355, 359, 360 Montery Formation, 343, 348, 349
Late Cenozoic glaciation, 158 Montrose field, 218
Late Kimmerian deformation, 381, 384, 385, 388, 389, 391, 392 M0re Basin, 197, 208
Late Permian evaporite sequence, 411 M0re-Tr0ndelag Fault Zone, 83, 84, 89, 392
Late Pliocene-Pleistocene sedimentary wedge, 163, 165, Mozambique Graben, 421, 423
Late Tertiary uplift and erosion, 163 multilayer lithospheric model, 459
Lauarm salt (Gulf of Mexico), 117 Multiple fault sets, 381
Laurentia, 91 Munin sand member, 300, 301
Laurentian Plate, 384 Muschelkalk, 25
linear joint model, 278
Ling Graben, 201 natural fracture patterns, 372
Lista Formation, 41 "negative listric" (antilistric) faults, 233
Lista Nose, 19, 20 Ness Formation, 357, 362
listric fault model, 290 Njord field, 295, 307, 322, 509, 516
listric faulting, 64 non-linear joint model, 278, 279, 280
lithosphere, 1, 6,175, 190, 222, 457 Nordkapp Basin, 94, 99, 101, 109, 111, 113, 114, 117, 118, 121,
lithospheric extension, 76 122, 125, 127, 128, 134
differential, 76 North, 99
lithospheric thickness, 14 South, 99, 101, 102, 105
loading, 221, 222 Nordkapp High, 99, 105
local inversion, 77 Nordland Ridge, 61, 69, 381, 385, 392, 393
Lofoten Islands, 75 Nordsel High, 94,112, 121, 122
log data, 397 normal stiffness, 278
Lola Formation, 501, 503, 504, 505 normal faulting, 289
Lombobo Mountains, 424 North Atlantic, 101
Lomfjorden Fault Zone, 481, 487 North Atlantic Volcanic Province, 76
Loppa High, 99, 101, 102, 105, 113, 134, 135, 269 North Caspian Depression, 118
Loppa-Nordkapp Platform, 99 North Sea, northern, 221, 462
Lost River Faults, 290, 291, 294 region, 197
Lulu structure, 450, 455 Norwegian-Danish Basin, 19, 20, 201, 202, 213
Norwegian-Greenland Sea, 158
Mads High, 2, 3, 496, 503, 504, 506 numerical simulation, 277, 306
magmatic activity, 64 Ny Friesland Orogenic phase, 482
magmatic underplating, 55, 64, 77 Nysleppen Fault Complex, 122, 127,128
Subject index 547

ODP Site 642, 76 rates of deposition, 5

oil generation, 439, 448 reactivation, 121, 481
oil migration velocity, 305, 314 reactivation of faults, 357
oil/water contact, 356 reconstruction, 426
Oklahoma, 519, 520 reflection coefficients, 255, 260
Old Red Sandstone, 94, 482 relay ramp, 300
onlapping sequence, 37 relay zone, 300
opal-CT grade diagenesis, 351 reservoir compartmentation, 355, 363
origin of joints, 346 reservoir scale, 277
Oslo Graben, 22,199 residual friction angle, 280
outcrop-scale fault, 231 resistor network method, 366
Outer Rough Basin, 2, 3, 444, 496, 504, 506 restoration, 221
oxygen isotope analysis, 188, 511 restricted water circulation, 78
0ygarden Fault Zone, 44 reverse drag, 290, 293, 294, 295
rift basins, 469
palaeobathymetry, 228 rift unconformity of Tertiary age, 64
paleic surface, 170, 171 rifting, 385
map, 171,177 rim synclines, 109
paleobathymetry, 46 original maximum salt thickness, 109
Paleocene and earliest Eocene land area 78 primary, 109
paleostress, 85 secondary, 109, 114
Paleozoic rift basin, 109, 111 Ringk0bing Fyn High, 2, 3, 4,14, 439, 440, 450, 453, 495, 496, 498,
passive continental margin, 55 501
passive margin, 469 Ringvass0y-Loppa Fault Complex, 99, 101
Roar structure, 450, 454
peripheral forebulge, 187,191
Rocky Mountains, 476
permeability, 321, 355, 361, 365, 369, 371, 377, 413, 453
R0dby Formation, 39
of faults, 515
Rogn Formation, 307
Permian extension, 411
rotated fault blocks, 253
Permian fault system, 412
Rotliegendes, 22, 498
phase boundary migration, 177, 179, 180, 183-184
Rotliegendes volcanics, 440
physical model, 236
Piper Formation, 300
planar joints, 345 Saalian, 193
planar faults, 235 Saiika High, 99, 105
plane strain, 244, 248 Sales, 470, 471, 474
pleistocene glaciation, 188 Salt Dome Basin, 2, 7
Pointe de Plate, 470, 471, 474 Salt Dome Province, 496, 501
salt facies, 24
Poisons's ratio, 190
salt pillows, 121, 122, 123
Polish-Danish trough, 37
salt wall, 411
porosity, 413, 450, 454
sand body geometries, 151
post-middle Jurassic tectonic development, 1
Santa Maria Basin, 343, 348, 351
post-rift subsidence, 44
Sassendalen Group, 482, 483
post-seismic rebound, 293
Sauda Formation, 39
Poul Formation, 503, 504
seal continuity, 336
Poul Plateau, 496, 501
seal strenght and stability, 339
pre-Jurassic structural grain, 197
seal timing, 334
pressure solution, 507, 515
seal types, 328
principal horizontal stress, 519
sealing capacity, 325, 335
principal stresses, 286
sealing faults, 416
production, 363
SECMIG, 306, 309
program RIFT, 43, 46, 457, 462
secondary migration, 187, 305
provenance, 142
section restoration, 269, 271
pull-apart, 412
sediment entry point, 141
zone, 247, 250
sediment pathways, 142
pure shear crustal extension, 43
sediment production potential, 141
pure shear extension, 457
sediment transport, 299
pure shear lithospheric extension model 43
sediment wedges, 133
PVT modelling, 153
sedimentary basin formation, 457
segregated flow, 305
Quaternary, 187 seismic character, 146
seismic data, 397
ramp-flat geometry, 233, 234 seismic facies and lithology, 133
Rannoch Formation, 357, 359, 361, 362 seismic sequence, 133
Ras Subbasin, 56, 66 seismogenic layer, 293, 294
rate of penetration (ROP), 155 seismostratigraphic interpretation, 133
548 Subject index

seismostratigraphic subdivision, 133 stylolites (stylolitization), 399, 507, 515, 517

Sele Formation, 41 sub-seismic fractures, 250
Sele High, 19, 20, 41 submarine fan depositional model, 147
self-similarity relationship, 359 subparallel faults, 519, 526
Senja Fracture Zone, 101 subsidence, 289, 299
Senja Ridge, 99,101,102, 105,156, 158,159,163 curves, 463
separation of Norway from Greenland 389 history, 2, 439, 444
sequence mapping, 143 summit level, 171
shale compaction, 155 map, 170, 171,176
shear fractures, 346 Svalbard, 93,153,154,156,159, 481,507
shear displacement, 286 Svalbardian Orogeny, 482
shear stiffness, 278 syn-rift deposisional models, 136
shear strain, 249 syn-rift megasequence, 133, 142,143
Shetland Through, 224, 228 syn-rift sediments (late Jurassic-early synsedimentary faulting),
shortening, 481 24
silica diagenesis, 344 syn-rift subsidence, 44
sill instrusions in Cretaceous sediments, 64 syn-sedimentary faulting, 507, 508
simple shear detachment model, 43 synthetic faults, 234
simulated fracture patterns, 372
Skagerrak, 161 Tail End Graben, 1, 2, 3, 7, 9,15, 439, 440, 444, 448, 453, 496, 498,
Sleipner area, 217 501, 503, 504, 505, 506
slipvector, 87 Tampen Spur, 43
slope instability, 142 Tana Low, 99
small-scale faults, 355, 359, 360, 361, 507, 508 Tarbert Formation, 357, 362
"smear" seal, 333, 337 Tau Formation, 39
Snorre block, 300, 301 tectonic backstripping, 5
Snorre field, 208, 297 tectonic fractures, 402
Sn0hvit field, 156,159,160,161 tectonic framework, 109
"soft-domino" model, 299 tectonic modelling, 43, 46
S0gne Basin, 2, 3, 440, 444, 450, 453, 496, 498, 501, 503, 506 tectonic setting, California, 343
Sognefjord Formation, 213 tectonic subsidence, 1, 44, 55, 59, 60
Sole Pit Trough, 217 tectonics, 381
S0rkapp Basin, 99 Tempelfjorden Group, 482, 483
source rock, 351 Tertiary compressional tectonicsf, 481
source capacity, 439 Tertiary tectonic events, 171
South Viking Graben, 300 Tertiary uplift, 153, 156
Soviet Arctic, 93 Tertiary West Spitsbergen Orogeny, 483
Spekk Formation, 307, 310, 311, 322 Tethyan margin, 469
spill point, 187 Tethys trough, 37
Spitsbergen, 94, 103, 153, 481, 507 Texas, 519, 520
stability of different types of seal, 325 thermal focusing, 253
Stappen High, 99, 134, 156 thermal history, 421, 428, 439, 444
Statfjord field, 44 thermal subsidence, 6, 60, 221, 224, 457
Statfjord Nord field, 208, 300 thermal uplift, 222
Stavanger Platform, 29, 39, 41 thermo-mechanical modelling, 76
steer's head geometry, 459 thick Paleocene sequence, 64
Stord Basin, 202, 213 thickness of ice sheet, 189
Storfjordrenna, 165 thick-skinned (basement involved), 481
stress considerations, 261 thin-skinned structural style, 481
stress orientation, 406, 407 Thousand Springs Valley, 290
stress systems, 83 three-dimensional strain, 381
stress trajectories, 519 three-dimensional seismic, 411
stress trajectory pattern, 519, 522 thrust belts, 469
structural controls on deposition, 133, 141, 142 thrust structures, 469
structural development, 495 Tilje Formation, 307
structural disturtion, 253 timing of fracturing, 345
structural dome (Scandinavia), 171 Tor Formation, 41, 397, 398, 403
structural inhesitance, 469 Tornquist-Teisseyre Line, 25, 197, 199, 201, 202, 213, 217
structural interpretation, 133 total tectonic subsidence, 60
structural inversion, 411 Triena Basin, 56, 66, 68, 69
structural maps, 133, 138, 139, 142 trailing-edge fault, 247
stretching model, 290 transfer zone, 140, 141, 142, 300
structural style, 197, 481 transfer fault, 416
structural traps, 475 transgressive mega cycle, 28
structurally controlled topography, 141 Triassic rifting, 462
structure, 197, 221 Troll Field, 161, 205, 293, 301, 516