See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/336367413

Using basin modelling to examine the origin of the hydrocarbons within the
deepwater Rovuma Basin of Tanzania, Mozambique and the Comoros

Conference Paper · October 2019

CITATIONS READS

3 2,760

4 authors, including:

David Boote
DBConsulting Ltd
30 PUBLICATIONS 806 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Tertiary turbidites systems Southern San Joaquin View project

East Africa petroleum systems View project

All content following this page was uploaded by David Boote on 09 October 2019.

The user has requested enhancement of the downloaded file.

 Using basin modelling to examine the origin of the hydrocarbons
within the deepwater Rovuma Basin of Tanzania, Mozambique and the Comoros
PESGB-HGS Africa E&P Conference, London, 1st-2nd October (Extended Abstract)
Mike Rego1, Andy Carr2, Nick Cameron3 and David Boote4

1Rego Exploration (Oil and Gas Consultancy Services) UK

2 Global Exploration Services Ltd.,`UK
3 GeoInsight Ltd., UK
4 DavidBoote Consulting Ltd,UK ([email protected])

INTRODUCTION
Over 200 TCF of thermally mature dry gas has been discovered offshore northern Mozambique and south-
ern Tanzania since 2009. Two distinct deep water provinces can be distinguished (Figure 1) :
Greater Mamba Province in northern Mozambique immediately east of the Rovuma fold and thrust
wedge and west of the Kerimbas Graben with 130-155TCF gas reservoired in Oligocene, Eocene and
Palaeocene deepwater sands (Figures 1 & 2).
Seagap Province in southern Tanzania aligned along the Seagap Fault with >32TCF gas and small
amounts of condensate in younger Cretaceous and Tertiary contourite-turbidite sandstones (Figures 1 &
2).

Figure 1: Rovuma Basin - simplified onshore geology and offshore EMAG2-V2 with major structural fea-
tures and oil and gas occurrences highlighted. North and South cross sections illustrated in
Figure 2 and seismic traverses (IONTZ3-2700 after McDonough et al. 2012 and IONMZ1-8700 from Dan-
forth et al. 2012) on Figures 16 and 17. Magnetic anomalies from Klimke et al. (2016)
The source of the gas is uncertain. Maturity modelling suggests candidate Permo-Triassic and Lower-
Middle Jurassic source rocks were overmature some time before charge during the late Cenozoic. Several
hypotheses have been proposed to explain the mismatch between model and observation, including
‘motelling’ gas at depth, retardation of maturation by overpressuring and a recently elevated heat flow in
response to mantle upwelling associated with East African rifting and Comoros volcanism (Sayers 2017).
This review examines the charge history of both provinces by modelling the maturity of candidate source
rocks extrapolated from regional stratigraphic control. The analysis compliments an earlier investigation by
Rego et al.(2013) entitled "Gas success along the margin of East Africa, But where is all the generated

Figure 2: Simplified regional cross sections based on offshore IONTZ3-2700 and IONMZI-8700 and
selected onshore seismic from Hudson (2011) and other sources. The sections illustrate the structural
architecture of coastal Tanzania and the Rovuma Basin with the Greater Mamba and Seagap petroleum
systems highlighted. Other hydrocarbon occurrences and key wells are also shown. Geological outcrops
supported by maturity profiles suggest the onshore basin margin has suffered significant late uplift, tilting
and unroofing of >1000m. The deep crustal interpretation of the southern cross section is based in part on

REGIONAL PETROLEUM SYSTEMS ANALYSIS

Until the recent Greater Mamba and Seagap discoveries, hydrocarbon exploration along the East African
continental margin had proved very disappointing despite numerous oil and gas shows, seeps and several
large exhumed palaeo-accumulations testifying to the presence of once prolific petroleum systems.
Candidate Source Rocks
These several petroleum systems can be tentatively correlated to organic rich source intervals ranging from
Permo-Triassic to Middle Jurassic in age (Figure 3)
 a branch of the intercontinental Karoo rift system extended south from the southern Morondava Basin of
Madagascar to the Rovuma, accommodating intra-Permian coals and carbonaceous shales (Sakoa
Formation and Lukuledi-1) capable of generating significant volumes of gas (Figures 4 & 5).
Figure 3: Tectono-stratigraphic evolution of the East African continental margin. Mega sequences modified
from Danforth et al. (2012) and McDonough et al.(2012) with a summary of key tectonic events variously
from Buscaglione et al. (1993), Jourdan et al. (2008), Michon et al. (2016), Davison and Steel (2018),
Reeves (2018) and others. Key source intervals present in coastal Tanzania and western Madagascar are
highlighted.

 the Karoo rift system suffered a period of increased extension and subsidence during the latest Permian
-early Triassic and was briefly flooded from the Paleo-Tethys in the north. Shales of this age (Middle
Sakamena Formation) are associated with two oil families – an isotopically light marine oil characteris-
tic of the giant Tsmiroro and Bemolanga palaeo-accumulations now exposed in the northern Moron-
dava and an isotopically heavier restricted marine or lacustrine oil observed in many exploratory wells
to the south (Boote & Matchette-Downes, 2009; Boote et al, 2017). Age equivalent source rocks have
not so far been encountered in the Rovuma, but regional projections from the contiguous on-trend
 Morondava Basin suggests they are probably present (Figure 6).
 Figure 4: Lower Permian (Karoo) coal and
carbonaceous shale source interval.
Sakoa Fm, southern Morondava: outcrop-
ping intra-Sakoa (100m) interval of carbona-
ceous mudstones and thin bedded coals
with sandstone. TOC content varies from 15
-70(wt%) dominated by terrestrial plant
matter with minor bacterial and algal
content. Hydrogen indices range from 80 to
280mgHC/gCorg.
Lukuledi-1, onshore Rovuma: 500m sec-
tion of interbedded sands, silts and grey-
dark grey carbonaceous shales with thin
coals towards the base. The section in-
cludes thin organic rich intervals with hydro-
gen indices ranging from 100-388 mgHC/
gCorg and coals (35-78% TOC) with indices
200-300 mgHC/gCorg.

Figure 5: Lukuledi-1, onshore Rovuma – summary of organic intervals encountered in the Lower-Middle
Permian. The middle and upper part of the Karoo (Permo-Trias) section was removed by a pre-Middle
Jurassic erosional unconformity.
Figure 6: Upper Permian/Lower Triassic source
interval.
Middle Sakamena (basal Triassic–Induan),
Morondava Basin: dark grey to black organic rich
shales (50m) passing up into grey mudstones with
thin sandstones above (2-300m +). In the north and
central part of the basin the lower unit is typically late
mature dominated by type II organic matter with
1-6% TOC and 17-307HI (original ~400 average up
to 750 mgHC/gCorg) associated with marine micro-
phytoplankton suggesting a marginal marine facies,
becoming a less organic, brackish to fresh water
sequence above. Its very light ẟ13 C character (ẟ13
Csat -33%o, ẟ13 Carom -31.5%o ) is similar to that
of the Tsmiroro-Bemonlanga bitumens.
In the central and southern part of the basin TOC
content and hydrogen indices appear somewhat
lower, associated with numerous Permo-Trias oil
shows with a rather heavier isotopic (ẟ13 Csat -27%
o, ẟ13 Carom -26%o ) signature (Boote et al.2017)
Coastal Tanzania: a time equivalent marine section
is present in the Selous Rift of Tanzania. Its organic
content is unknown but B-carotane was encountered
in a Triassic oil show at Kisangire-1 (Rifiji Basin)
similar to some of the oils found in Madagascar.
Ponte Uifundo, southern Rovuma Basin: waxy oil
from a seep at Ponte Uifundo was sourced from
organic matter dominated by plant matter of probable

Figure 7: Early Jurassic syn-rift source interval.

Nondwa Evaporite Member (late Triassic –basal
Jurassic), Mondawa Basin, Tanzania: several oil
prone marine shale intervals interbedded within the
Nondwa evaporite sequence have organic contents
ranging from 1.2-7.1% TOC and hydrogen indices of
100-900 mgHC/gCorg. The organic matter is
characterised by gammercerane and C24 tertracyclic
terpanes indicative of a restricted to hypersaline
depositional environment dominated by algal matter
with some terrestrial contribution. The shales are
commonly associated with oil shows (ẟ13 Csat
-29%o, ẟ13 Carom -26.5%o average).
Majunga offshore: seismic reflector geometries
suggests a similar evaporitic facies may be present
offshore Majunga, but it has not so far been
confirmed by a well.
 Figure 8: Toarcian-Aalenian early post-rift source
interval.
Beronono Fm (Early Toarcian-Aalenian), Majunga
Basin: early post-rift restricted marine oil prone shales
with up to 69% TOC and 790 mgHC/gCorg HI,
tentatively typed to oil shows in Bemangahazo-1 well.
The distribution of the Beronono facies appears to be
limited by pre-existing rift topography but age
equivalent shales have been encountered in wells
along the coast (Sofia-1) and offshore (Mariano-1) with
organic contents of 1.3% TOC (average) to 4.2%
(maximum).
Andafia Fm (Toarcian-Aalenian), Morondava Basin:
early post-rift restricted marine shales and marls with
thin shelly limestones and sands infilling and onlapping
pre-existing rifted block faulted topography. Late to
post-mature in the subsurface with residual late-post
mature kerogen ~ 0.6-7.65% TOC and low hydrogen
indices (20-40 mgHC/gCorg).
Mitambia Fm (Toarcian-Aalenian), coastal Tanzania:
typically represented by a nearshore facies of
transgressive marine sandy silts, muds and
subordinate limestones where observed, but
associated shales locally include oil prone

 a paroxysmal phase of extension and rifting during latest Triassic-early Jurassic is recorded by the
deposition of restricted marine evaporites in the Mandawa Basin and offshore Majunga. Oil prone
evaporitic shales (Nondwa Formation) have been encountered in the late Triassic-early Jurassic syn-rift
sequence of the Mandawa Basin typically associated with numerous oil shows (Kagya, 1996). Else-
where the equivalent sequence is represented by continental clastics, suggesting the evaporite facies
and associated source rocks might be limited to the axial part of the Jurassic rift. Its extent to the south
is unconstrained (Figure 7).
 as central and eastern Gondwana broke apart in the Toarcian, the remnant rift topography was flooded
by a marine transgression from the north accommodating organically rich late Toarcian-Aalenian,
shales locally in topographic lows. This facies is represented by the Andafia and Beronono shales in the
Morondava and Majunga Basins of western Madagascar. Where encountered onshore Tanzania, the
equivalent sequence is comprised of transgressive shallow marine sands and carbonates (Mtumbei For-
mation) but organic shales similar to those of Madagascar may be well developed further offshore
(Figure 8).
 as eastern Gondwana drifted away during the Bajocian-early Callovian, an open marine embayment
extended south from Tethys between East Africa and Madagascar, fringed by shallow marine platform
carbonates. These pass laterally into slope and basinal facies represented by the distal equivalents of
the Bemaraha Limestones (Ankarana/Mariarano Members) in western Madagascar and the Makarawe
Member of the Mtumbei/Amboni Formation (Makarawe Member) in Tanzania. This basinal facies is typi-
cally enriched in oil prone kerogen typed to oil shows and seeps by distinctive carbonate biomarkers
(Figures 9 & 10).
 as eastern Gondwana moved away from Africa during the later Jurassic, both continental margins sub-
sided and were flooded by marine shales. Subsidence slowed during the early Cretaceous perhaps in
response to its more southerly drift along the Davie transform, and delta systems began to build sea-
ward. Where penetrated, both shales and deltaic sands (Kapatimu and equivalents) typically contain
small amounts of gas prone kerogen (Kagya, 2000), increasing to 5-20% organic content locally
(Figures 11,12 & 13).
 The southerly drift of Madagascar and seafloor spreading in the Somalia deep water basin terminated
during the Aptian. Following a brief period of marginal fault reactivation/inversion, both continental mar-
gins subsided during the Middle Cretaceous and were once again blanketed by deep water marine
shales (Kagya, 2000; Mkuu, 2018). These include small amounts of dispersed gas prone kerogen, in-
creasing to 3-3.5%TOC locally in the Mvule-1 well east of Mafia Island (Figures 14 &15).
 Figure 9: Bajocian –Early Callovian early drift source
interval.
Bemaraha Fm (Bajocian –Early Callovian), Moron-
dava and Majunga Basins: organic rich slope
(Ankarana) and basin plain (Mariarana) carbonate
mudstone facies. Subsurface sections are late to post
mature with residual 2.2%TOC (estimated 3.5% aver-
age originally) and low hydrogen indices (estimated +/
- 400 mgHC/gCorg) associated with minor oil and gas
shows.
Amboni & Mtumbei Fms (Bajocian-Bathonian),
coastal Tanzania: carbonate source rocks of equiva-
lent age are also present along the western margin of
the Middle Jurassic oceanic embayment. These are
represented by the slope and basinal facies of the
Amboni Limestone (Makarawe facies = 1.5%-4% lo-
cally up to 8% TOC oil prone kerogen) and Mtumbei
limestones (>1.5% TOC), tentatively typed to carbon-
ate sourced oil seeps and shows along the coast.

Figure 10: Bemahara Facies Model, Morondava Basin: illustrating the relationship between platform car-
bonates and organic-rich basinal equivalents.
Figure 11: Upper Jurassic transgressive and early Figure 12: Lower Cretaceous highstand pro-delta
highstand shales. shales.
Antsolova Fm & Equivalent, North Morondava: Lower Cretaceous shales, coastal Morondava:
poor source potential (high-post mature shale average 0.5-3.7%TOC (locally up to 26%) hydrogen
interval with 0.7%TOC, HI 60-72 in Heloise-1) index average 100-150 locally up to 400.
Bagamoyo Shale, Songo Songo, coastal Songo Songo, Coastal Tanzania: average 0.5 -
Tanzania: 0.5-1.76%TOC, HI 33-130 1.5% TOC locally up to 10% (drifted plant material in
deltaic sands – see Figure 15)

The source potential of both late Jurassic/early Cretaceous and Middle Cretaceous shale is very poorly
constrained offshore northern Mozambique/ southern Tanzania, but regional projections suggest a rela-
tively thick section of shale with dispersed gas prone (Type III) kerogen averaging 1-3% organic content
may be present.
Petroleum Systems
Oil and gas occurrences in the coastal basins of East Africa and western Madagascar testify to the pres-
ence of both extinct and still active petroleum systems. Extinct systems are typically located along the in-
board basin margin, experiencing very significant uplift and exhumation during the Cenozoic. These are
represented by:
 North Morondava System: a basal Triassic marine sourced system responsible for the giant Tsmiroro
and Bemolanga palaeo-accumulations in the northern part of the Morondava Basin. The Triassic reser-
voirs were charged prior to Santonian volcanism and subsequently water washed and biodegraded dur-
ing intra-Cenozoic uplift and unroofing.
 South Morondava System: ubiquitous intra-Karoo oil and gas shows in the southern Morondava repre-
sent the remnants of accumulations charged by restricted marine/lacustrine basal Triassic source rocks,
subsequently dispersed during Cenozoic exhumation. The system was contiguous and on direct trend
with the Rovuma Basin prior to break-up.
 Coastal Rifiji Systems: the Wingayongo tar sand represents a Lower Cretaceous palaeo-accumulation
charged sometime in the late Cretaceous/early Tertiary. The oil was severely biodegraded during late
Cenozoic uplift and unroofing and its origin is uncertain. However, there is some rather ambiguous
Figure 13: Cretaceous well correlation, coastal Morondava highlighting the more organic prone
intervals, typically dominated by Type III/IV organic matter. Compiled from various sources.

Figure 14: Middle Cretaceous transgressive and

early highstand shales
Coastal Morondava
Serinam-1 (Albian-Cenomanian) - 0.9% TOC,
locally up to 9% TOC; HI 35-250, Type IV(II)
Saronanola (Cenomanian-Turonian) - 2-3% TOC
average, locally up to 8% TOC; HI (50 = cooked)
200-250 average, Type IV(II)

Songo Songo, coastal Tanzania (compiled from

Kagya 2000
Upper Sakura Fm. (Cenomanian-Turonian) 0.5
-1.8% (0.9% av) locally up to 11.7% TOC, HI 33-
178
Lower Sakura Fm. (Late Aptian-Albian) 0.64-
0.93%TOC (1 sample)
Figure 15: Cretaceous well correlation, coastal Tanzania, highlighting organic content of shales (and
sands). Compiled from Kagya (2000) and Mkuu (2018).

evidence suggesting it may have been a mixture of super-mature and less mature oils with a reducing
evaporitic signature, tentatively typed to Permo-Trias and syn-rift Jurassic source rocks.
Significant volumes of gas reservoired in shallow Upper Cretaceous and Triassic sandstone in the
nearby Mambukofi and Mbuyu discoveries appear to represent a different system charged late in re-
sponse to depressuring during late Cenozoic unroofing.
 Carbonate sourced systems: seeps and oil shows with a distinctive carbonate biomarker signature have
been encountered at several locations along the coast from the southern Rovuma to Pemba Island in
northern Tanzania. These represent the remnants of earlier accumulations dispersed during late Ceno-
zoic exhumation. Regional stratigraphic control points to the basinal facies of the Middle Jurassic plat-
form carbonates as the source.

Active petroleum systems are present further outboard in areas not affected by significant late uplift.These
include:
 Coastal Tanzania Fold Belt system: transpressional reactivation of deep syn-rift faults in Oligocene-early
Miocene and Pliocene to Recent time was responsible for a trend of locally very large antiformal struc-
tures along the Tanzanian coast. These now trap gas at Songo Songo, Kiliwani North, Nyuni and Mafia
Island.
 The Songo Songo Field reservoirs some 1-1.25TCF gas in Apto-Albian deltaic sandstones sealed by
plastic overpressured shale. The gas is very mature. It is associated with small amounts of condensate
and light oil, which appear to represent a fractionate released at depth by pressure reduction during up-
lift. Biomarker analysis indicates a mixed source tentatively attributed to the Toarcian-Aalenian, although
similar oils in the adjacent Nyuni field have an additional carbonate contribution. Charge was clearly very
late and probably triggered by depressuring during structural inversion and uplift. The association of gas
with low salinity formation water might reflect the co-mingled migration of gas saturated water from
 Mafia Deep-1 tested a structural anomaly on the Mafia Island anticlinal culmination, encountering sig-
nificant volumes of gas at several levels in the Cretaceous. The trapping geometry was formed by late
Cenozoic-recent inversion and charged by gas released during uplift and depressuring. Further north,
the Tundaua seep on Pemba Island may reflect late syn-tectonic leakage from some deep carbonate
sourced accumulation.
 Onshore Rovuma (Intra-Cretaceous) system: mid-Cretaceous canyon fill sandstones reservoir dry gas
at Ntorya (with traces of condensate) and Temba, onshore Rovuma. Charge timing is poorly con-
strained although pronounced regional tilting during the Cenozoic suggests it was late with gas migrat-
ing westwards from a deeply buried section along the coast and offshore.
 Onshore Rovuma (Mnazi Bay/Msimbati) system: the Mnazi/Msimbati accumulation reservoirs some
0.5TCF gas in Miocene deepwater channel sands, trapped by a large rollover culmination within the
extensional part of the Rovuma gravity slide. The gas is associated with small quantities of oil and con-
densate suggesting a largely thermogenic origin, although biogenic gas is present in the section above
and below. The source of the thermogenic gas and oil is uncertain.
 Greater Mamba deep water system: the Greater Mamba fields reservoir over 150TCF dry gas in
several deepwater basin floor fan channel and lobe sands of Oligocene, Eocene and Paleocene age
(Fletcher, 2017). The gas is held in combination footwall truncation traps formed during the late Mio-
cene, immediately in front of the Rovuma gravity slide complex. The accumulation lies updip of the Plio-
Pleistocene Karimbas Graben, where extensional boundary faults probably encouraged a rapid almost
explosive release of gas from some deep source and acted as a vertical conduit for migration into intra
-Cenozoic reservoirs above. The origin of the thermogenic gas is unknown, but considered to come
from an overmature oil prone source (Fletcher 2017).
 Seagap deep water system: the Seagap system comprises 10 gas fields aligned in a distinct trend
along the Seagap sinistral transcurrent fault. The fields range in size from 0.5 to 4.5TCF, reservoired in
slope channel and lobe sandstones of Albian to Eocene age (Sansom, 2016). The trapping style is di-
verse, dominated by structural culminations along the Seagap fault but combination/onlap trap geome-
tries are also significant. The gas is thermogenic and highly mature with minor amounts of condensate
locally. Biomarker analysis suggests a mixed terrestrial marine source tentatively typed to the Jurassic

BASIN MODELING
The maturation and expulsion history of candidate source rocks for the Greater Mamba and Seagap petro-
leum systems was reconstructed by modelling five pseudo-wells located on two published seismic tran-
sects ... ION TZ3-2700 and ION MZ1-8700 (Figures 16 & 17). The resulting burial history models provide a
robust template for assessing which source sequence or sequences were more likely to have been respon-
sible for the enormous volumes of entrapped gas. Pseudo-well Ps1 is located on the northern transect just
east of the Seagap fault and Papa gas discovery, while Ps2 to 5 are on the southern transect, together
forming a regional maturation history profile from west of the Karimbas Graben to just east of the Davie
Transform .
The stratigraphic succession at each location was extracted from published ION interpretations of the two
traverses. All but the easternmost site on MZ1-8700 overlie continental or hyper-extended continental crust.
Changing heat flows through time were estimated from a tectono-stratigraphic reconstruction of the East
African margin incorporating Permian - Early Jurassic rifting, Toarcian break-up, two phase Late Jurassic -
Early Cretaceous drift and later Cretaceous thermal subsidence followed by younger Cenozoic tectonism
with increasing structural deformation along the Karimbas Graben and Seagap fault systems into the Pleis-
tocene - Recent.
The petroleum system modelling was undertaken using Genesis software. The heat flow histories used an
extensional rift model with beta factors increasing from 1.7 at Ps-1 and Ps-2 to infinity at Ps-5 (ocean
crust). The thermal history was calibrated using measured well temperature and vitrinite reflectance
data. The petroleum generation assumed mixed oil-gas prone kerogens (I and II) in MS-1 and MS-2 and
gas-prone kerogens (III and IV) in ‘within’ MS-2, ‘within’ MS-3, base MS-4, ‘within’ MS4 and top MS-4 mega
-sequences.
The resulting geohistory analysis suggests all the candidate source rocks are currently overmature at the
Ps1, 2 & 3 locations. The Late Jurassic-Early Cretaceous gas window (Ro=1.3-3.0%) ranges from mid-
Cretaceous to late Cenozoic in age and mid-Cretaceous for the deeper Lower Jurassic interval. Both be-
come progressively cooler further outboard at Ps4 and Ps5 where Early Cretaceous sediment rests on oce-
anic crust. Peak gas expulsion for Permo-Triassic source rocks is estimated to be late Jurassic- early Cre-
taceous. These results imply that gas expelled from the deeper candidate source rocks during the Creta-
ceous would have to have been stored at depth before being released to charge post-Miocene aged traps.
However late maturity gas from the Late Jurassic-Early Cretaceous interval might have been available for
direct charge.
Figure 16: Geohistory Model PS1 (ION TZ3-2700). Ps1 is located close to the Seagap transcurrent fault and the Seagap petroleum system
(Figure 1 for location). The burial history profile for the Toarcian (base MS2) is highlighted. The model assumed a beta factor of 1.7
Figure 17: Geohistory Models Ps2, Ps3, Ps4 and Ps5 (ION MZ1 8700). PS2 is located on the western side of the Karimbas Graben immedi-
ately downdip of the Greater Mamba petroleum system. Ps2, Ps3 and Ps4 are located further outboard (Figure 1 for location). The burial
history profile for the Toarcian (base MS2) is highlighted. The models assumed a changing beta factor from 1.7 to 35 representing the change
in crustal thickness and character from west to east
CHARGE MODELS
Any robust charge model proposed for the Greater Mamba and Seagap petroleum systems must be con-
strained by several basic observations:
(1) Both systems were charged very rapidly by an enormous volume of gas.
(2) The gas is very dry with only very minor amounts of condensate
(3) Both systems are intimately linked to late (Plio-Pleistocene) extensional or transtensional faulting which
appear to have provided efficient vertical migration conduits from some deep source.
(4) The rapidity of charge and amount of entrapped gas suggests a very large volume of gas must have
been immediately available for migration, when released by late faulting.

The regional petroleum systems analysis and geohistory modelling suggests three more likely alternatives:

 Dispersed gas-prone organic matter could be present in Upper Jurassic - Lower Cretaceous pro-delta
shales directly beneath both systems. This character of this section is unknown but regional projections
suggest its organic content might range from 0.5-3.0% Type III gas prone organic matter. Such a
source may produce large volumes of gas if present over a thick interval and over an extended period
of time, perhaps enhanced by an increased rate of maturation and expulsion during the Plio-Pleistocene
in response to rapidly increasing heat flow associated with late tectonism. Nevertheless it is question-
able whether such a source would be capable of producing the enormous volumes of entrapped gas in
such a short period of time (plus an equal or even greater volume dispersed during migration).
 Expulsion from deeper more oil prone Permo-Triassic and Lower-Middle Jurassic source intervals might
have been retarded by overpressure and only released when ruptured by Plio-Pleistocene tecton-
ism. Pressures in the post mid-Ctetaceous section are typically normal (except locally in thicker more
structurally deformed shale intervals), while the existence of overpressuring in the deeper section is
unconstrained. However even if present at depth, larger volumes of liquid hydrocarbon would probably
also have been released at the same time as the gas. Although the absence of liquids does not pre-
clude this alternative, it does suggest it is less likely.
 Gas expelled from any overmature oil prone source during the Cretaceous might have been held
(‘motelled’) in Upper Karoo sandstones prior to its release in the late Cenozoic. Regional control
suggests the Karoo could have formed an effectively infinite, if low quality (fractured?) reservoir and
depending upon salinities, temperature and pressure, one capable of holding very large volumes of gas
dissolved in interstitial formation water (Figure 18). Such gas would become immediately available and
expand rapidly in response to any significant pressure drop. Because of its large expansion factor and
greater solubility in water, such gas would tend to be dry. Depressuring by late movement along the
Karimbas boundary fault (cumulative 3000m Plio-Pleistocene extension) could have exposed a very
large amount of gas saturated water at depth and pro-

Figure 18: Methane solubility in interstitial

(formation) water. The diagram is modified from
Cramer et al. (2002), Price (1979), Dore and Jensen
(1996) and others illustrating the changing solubility
of methane in low and high salinity waters with in-
creasing depth, pressure and temperature under hy-
drostatic and lithostatic regimes.
There was insufficient information to reconstruct
profiles for offshore East Africa but an approximate
range is highlighted in red for illustrative purpose.
The conceptual case illustrated, suggests a reduction
in methane solubility of 8m3/m3 in response to
changing pressure and temperature from 7 to 1kms.
The solubility of liquid hydrocarbons is lower than
methane. Consequently, assuming a sufficiently large
volume of methane saturated water, a pressure and
temperature change of this magnitude could release
enormous volumes of dry gas. While there is no di-
rect evidence, the close association between the two
deep water petroleum systems and Plio-Pleistocene
faults suggests the faulting may be responsible for
depressuring a gas saturated interval at depth
All of the Permo-Triassic and Lower Middle Jurassic candidate source rocks may have contributed in
varying amounts to this process. Several of the nearby onshore seeps have a carbonate signature support-
ing an effective Bajocian-Callovian source in the vicinity of the offshore gas provinces. Wells in the
Mandawa Basin encountered oil shows typed to the Nondwa evaporite, Toarcian-Aalenian and Middle
Jurassic source rocks. The Lukuledi well onshore Rovuma, intersected a thick Permian section with coals
and organic rich carbonaceous shales, but perhaps of most significance are the ubiquitous intra-Karoo oil
and gas shows observed in many of the wells drilled in the central and southern Morondava Basin, testify-

Figure 19: Charge model, Greater Mamba petroleum system. The conceptual model depicts migration of
deep gas vertically upwards along Plio-Pleistocene extensional faults bounding the Karimbas Graben and
then laterally via intra-Cenozoic sands to be trapped by the leading thrusts of the Rovuma Gravity Slide.
Although speculative, this finds some support from the anomalous loss of seismic definition at depth
(highlighted in red). Such an anomaly might be a response to elevated free gas concentrations
depressured by Karimbas faulting. The seismic line is from Wen et al. (2015) and is located near the
IONMZ1 8700 line (Figure 1 for location).

EXPLORATION SIGNIFICANCE
This analysis highlights two play concepts of potential significance. First, the Davie Ridge immediately east
of the Karimbas Graben is well placed to entrap gas released by late movement along the eastern
boundary fault. The Ps4 geohistory model suggests a less mature section at this location might be capable
of producing some liquids as well as gas, directly from any Jurassic source rock exposed along the
bounding fault. The potential further outboard across the Davie Transform at the Ps5 location
appears much more limited and dependent upon some unknown Cretaceous source above oceanic crust.
Second, the three inboard geohistory models suggest peak oil and gas expulsion from Permo-Triassic and
Lower-Middle Jurassic source rocks occurred during the Late Jurassic-Middle Cretaceous. Some of the oil
(and gas) released at this time, might have migrated updip towards the onshore basin margin to charge
perhaps very large accumulations. Subsequent intra-Cenozoic uplift, tilting and exhumation would have
tended to reduce trap integrity , but more robust stratigraphic geometries may have retained this early
 REFERENCES

Boote, D.R.D. and C. Matchette-Downes, 2009. Extinct, near extinct and active petroleum systems of the
East African Coastal Basins. 8th PESGB-HGS Conference on African E&P, London (presentation).
Boote, D.R.D., C.Matchette-Downes and R.Sorhabi, 2017. Permo-Triassic petroleum systems of the
East African Coastal Basins. AAPG/SEG International Conference & Exhibition, London, 15th-18th Octo-
ber (presentation).
Buscaglione, L., M.Fazzuoli, M.Chiocchini and G.Pavia, 1993. Contributions to the stratigraphy of the
early to Middle Jurassic Formations of the eastern side of the Luuq-Mandera Basin, Bay and Gedo regions,
southwestern Somalia. Geology and mineral resources of Somalia and surrounding regions. 1 st Agron. Ol-
tremare Firenze, Relaz.e Monogr. 113, 153-168.
Cramer, B., S. Schlomer and H.S.Poelchau, 2002. Uplift-related hydrocarbon accumulations: the release
of natural gas from groundwater. In: A.G.Dore, J.A.Cartwright, M.S.Stoker, J.P.Turner and N.White (eds.).
Exhumation of the North Atlantic Margin:timing, mechansisms and implications for petroleum geology. Geo-
logical Society Special Publication 196, 447-455.
Danforth, A., J.W.Granath, J.S.Gross, B.W.Horn, K-J. McDonough and E.J.Sterne, 2012. Deepwater
fans across a transform margin, offshore East Africa. GEO Expro, 9,74-78.
Davison,I. and I. Steel, 2018. Geology and hydrocarbon potential of the East African continental margin: a
review. Petroleum Geoscience 24, 57-91.
Dore,A.G. and L.N.Jensen, 1996. The impact of late Cenozoic uplift and erosion on hydrocarbon explora-
tion: offshore Norway and some other uplifted basins. Global and Planetary Change 12, 415-436.
Fletcher, T., 2017. The Windjammer Discovery: play opener for offshore Mozambique and East Africa. In:
R.K. Merrril & C.A. Sternbach (eds.). Giant fields of the decade 2000-2010. AAPG Memoir 113, 273-304.
Gilardi, M., R. Longoni and G.Spadini, 2014. Tectonics of the Mozambique margin through the integra-
tion of gravity and magnetic modelling: the Rovuma Basin Case Study. IPTC 17345, 9pp.
Hudson, W., 2011. The geological evolution of the petroleum prospective Manadawa Basin, southern
coastal Tanzania. PhD Thesis, University of Dublin. March 2011.
Jourdan,F., G.Féraud, H.Bertrand, M.K.Watkeys and P.R.Renne, 2008. The 40Ar/39Ar ages of the sill
complex of the Karoo large igneous province: implications for the Pliensbachian-Toarcian climate change.
Geochemistry, Geophysics, Geosystems 9/6, June 2008.
Kagya, M.L.N., 1996. Geochemical characteristics of Triassic petroleum source rock in the Mandawa Ba-
sin, Tanzania. Journal of Africa Earth Sciences 23/1, 73-88.
Kagya, M.L.N., 2000. Hydrocarbon potential of the deep sea off Tanzania coastal basins as indicated by
geochemistry of source rocks and oils from Songo Songa Gas Field. Petroleum geochemistry & exploration
in the Afro-Pacific Region. 5th International Conference & Exhibition, 109-116, 25-27th November 2000.
Klimke,J., D.Franke, C.Gaedicke, B.Schreckenberger, M.Schnabel, H.Stollhofen, J.Rose and
M.Chaheire, 2016. How to identify oceanic crust – evidence for complex break-up in the Mozambique
Channel off East Africa.Tectonophysics 693, 436-452.
McDonough,K-J., E.Bouanga, C.Pierard, E.J.Sterne, J.W.Granath, A.Danforth and J.S.Gross, 2012.
Submarine fan chronostratigraphy from Wheeler-Transformed ION BasinSPAN seismic data, Late Creta-
ceous-Tertiary, offshore Tanzania.PESA News Resources 45-50, August/September 2012.
Michon, L., 2016. The volcanism of the Comoros Archipelago- integrated at a regional scale. In: P. Bachel-
ery et al., (eds.). Active Volcanoes of the Southwest Indian Ocean. Chapter 21, Springer-Verlag Berlin.
Mkuu, D.E., 2018. Palynological, palynofacies, thermal maturity and burial modeling analyses of the Creta-
ceous sediments from a series of Tanzanian onshore and offshore boreholes. PhD Thesis, University of
Southampton, March 2018.
Pereira-Rego, M.C., A.D.Carr and N.R.Cameron, 2013. Gas success along the margin of East Africa, but
where is all the generated oil. Search and Discovery Article #10488 (2013)
Price, L.C., 1979. Aqueous solubility of methane at elevated pressures and temperatures. AAPG 63, 1527-
1533.
Reeves,C.V., 2018. The development of the East African margin during Jurassic and Lower Cretaceous
times: a perspective from global tectonics. Petroleum Geoscience 24, 41-56.
Sansom, P., 2016. Sequence stratigraphic scheme for the Jurassic-Neogene of coastal and offshore Tan-
zania. East Africa-from research to reserves - Geological Society Conference Abstract Volume, 33-37, 13-
15th April 2016
Sayers, N., 2016. Linking recent gas discoveries to their source kitchens, offshore Tanzania. East Africa-
from research to reserves -Geological Society Conference Abstract Volume, 76, 13-15th April 2016
Sayers, N., 2017. Impact of Tertiary tectonic activity on the petroleum system, offshore Tanzania/ Mozam-
bique, East Africa. 16th PESGB/HGS Conference Abstract Volume, 42-43, 31st August-1st September 2017.
Wen, Z., Z.Wang, C.Song, Z.He and X.Liu, 2015. Structural architecture differences and petroleum explo-
ration of passive continental margin basins in East Africa. Petroleum Exploration and Development 42/5,

View publication stats