6.

Naturally-Fractured Reservoir
Models and Fluid-Flow Simulation

1
 Factors to Consider When Applying Conceptual
Models for Simulation of Fluid Flow in
Naturally Fractured Reservoirs

•  Geology of the fractured rock

•  Scale of interest
•  Purpose of the method and model being applied

2
Fluid-Flow Simulation Models for
Naturally-Fractured Reservoirs
1.  Equivalent Continuum Models
2.  Dual-Porosity Models
3.  Discrete Fracture Network Models
- Deterministic
- Stochastic

3
 Equivalent Continuum Simulation Models

•  In a conventional equivalent continuum model, the

heterogeneity of the fractured rock is modeled by using a
limited number of regions, each with uniform properties.
•  Individual fractures are not explicitly treated in the model,
except when they exist on a scale large enough to be
considered a separate fluid flow unit (e.g., an areally extensive
fractured zone).
•  At the scale of interest, hydraulic properties of the rock mass
are treated by coefficients, such as permeability and effective
porosity, that express the volume-averaged behavior of many
fractures.
•  Flow through the fracture network is calculated in many
directions and is used to derive the equivalent permeability
ellipse or permeability tensor.

4
 Equivalent Continuum Model for
Naturally Fractured Reservoirs

EQUIVALENT
REAL FRACTURED ANISOTROPIC, ELASTIC
MEDIUM NON-FRACTURED MEDIUM
Properties: Properties:
Fracture statistics of Elastic compliance and
apertures, lengths, permeability tensors
Replacing the
orientations, and stiffness fractured medium
by an
anisotropic,
elastic non-
fractured medium

5
 Dual Porosity Models
•  Dual porosity models consider fluid flow and transport in both
the connected fractures and the matrix blocks.
•  For a rock mass with large porous matrix blocks between the
conducting fractures, dual porosity models have been used to
account fro the release of fluid from storage in the matrix
blocks into the fracture network.
•  The geometry of the fracture network is idealized to the extent
that it can be represented by a small number of geometric
parameters (e.g., the average dimension of the matrix block).
•  The rock is characterized by two overlapping continuua, and
both are treated as porous media.

6
 Dual Porosity Models (Continued)
•  The key hydraulic properties are
- effective permeability and porosity of the fracture network
- matrix porosity and permeability
- storage coefficients for the fractures and matrix blocks.
•  The equation expressing fluid flow through the fracture network
contains source terms to account for flow from the matrix to adjacent
fractures.
•  A second set of equations describes the fluid flow in the matrix
blocks.
•  Drainage into the fracture network depends on the geometry of the
fracture-matrix interface and hydraulic diffusivity of the matrix.
•  Usually, a uniform geometry and block length are adopted for large
regions of the flow domain.
•  Some guidance in choosing a representative value for block length
can be determined from knowledge of fracture spacing.

7
 Interactions in a Dual Porosity
Naturally Fractured Reservoir Model

Fluids exchange (crossflow) between

fractures and matrix blocks
SOLID
PHASE
FRACTURES
(Secondary Porosity)
MATRIX BLOCKS
(Primary Porosity)
8
Fluid-Flow in a Dual-Porosity
Naturally-Fractured Reservoir

Production Well

Matrix Recharge
-- Darcy’s Law

Fracture Flow
-- Cubic Law

9
Matrix and Fracture Fluid-Flow Equations

Darcy’s Law
k Δp
Q = −A
µ LX
A

Cubic Law
b 3 Δp
Q = − LY
12 µ LX b

10
Idealizations of a Fractured Reservoir Adopted in a
Dual Porosity Fluid Flow Simulation Model

11
 Limitations of Dual Porosity Models
•  The conceptual basis of dual porosity is appealing because of
its simplicity.

•  However, there are limits to the prediction capabilities of dual

porosity models:
- Over-regularization of the geometry of the fracture
network.
- Good parameter estimates are difficult to obtain.

12
Simple Idealized Fracture Networks and Matrix
Blocks for Dual Porosity Models (from Reiss, 1980)

Arrows indicate possible directions of flow

13
from Reiss (1980)
 Relationship Between Fracture
Permeability kf, Fracture Porosity
φf, Fracture Width b, and Matrix
Block Size a, for Sheet Fracture
Network Geometry, Dual-Porosity
Model (from Reiss, 1980)
 Relationship Between Fracture
Permeability kf, Fracture Porosity
φf, Fracture Width b and Matrix
Block Size a, for Match Stick
Fracture Network, Geometry Dual
Porosity Model (from Reiss, 1980)
 Relationship Between Fracture
Permeability kf, Fracture Porosity
φf, Fracture Width b and Matrix
Block Size a, for Cube Fracture
Network Geometry with Two
Effective Fracture Planes, Dual
Porosity Model (from Reiss, 1980)

17
 Relationship Between Fracture
Permeability kf, Fracture Porosity
φf, Fracture Width b and Matrix
Block Size a, for Cube Fracture
Network Geometry, Dual Porosity
Model (from Reiss, 1980)

18
Matrix – Fracture Fluid Transfer Parameter

Fracture a Fracture

σ1 Matrix σ2

a
Smaller Surface Area Larger Surface Area
Hinders Fluid Transfer σ1 < σ 2 Eases Fluid Transfer

⎛ 1 11 11 ⎞⎞ km
σ = 4⎜ 2 + 22 ++ 22 ⎟⎟ Tmf = σ ∗ ∗ ( Pm − Pf )
⎝a aa aa ⎠⎠ µ
19
Shape Factor as a Function of Fracturing Spacing for Network
of Three Orthogonal Fractures of Equal Spacing

20
Gas Rates Obtained from Dual Porosity Model for
Different Values of Fracture Spacing

10 acre model with single well in center with Pi = 3600 psi and constrained Pbh =1000 psi
21
 Cumulative Gas Production Obtained from Dual
Porosity Model for Different Values of Fracture Spacing

10 acre model with single well in center with Pi = 3600 psi and constrained Pbh =1000 psi
22

ECLIPSE Dual Porosity Model Input Data
Dry Gas Reservoir Properties
φm = 0.2 φf = 0.001
km = 1 mD kf = 100 mD

Lx = 500 ft
Ly = 1000 ft
h = 40 ft Lx
γ = 0.626
Pi = 3500 psi
BHP = 2000 psi Ly
23
Gas Production Rate Increases as Fracture Aperture
Increases with Constant Matrix Block Size

b Increases
a = 100 cm
φm = 0.2
km = 1mD

24
Reservoir Pressure Depletion Increases as Fracture Aperture
Increases with Constant Matrix Block Size

a = 100 cm

φm = 0.2
b Increases
km = 1 mD

25
Gas Production Rate Increases as Matrix Block Size Decreases
with Constant Fracture Aperture

a Decreases

φm = 0.2
km = 1 mD b = 100 microns

26
Reservoir Pressure Depletion Increases as Matrix Block Size
Decreases with Constant Fracture Aperture

a Decreases

b = 100 microns

φm = 0.2
km = 1 mD

27
Gas Production Rate Increases as Fracture Porosity
Increases with Constant Fracture Permeability

kf = 500 mD φm = 0.2
km = 1 mD

φf Increases

28
 Reservoir Pressure Depletion Increases as Fracture
Permeability Increases with Constant Fracture Porosity

kf = 500 mD

φf Increases
φm = 0.2
km = 1 mD

29
 Gas Production Rate Increases as Fracture
Permeability Increases with Constant Fracture Porosity

φf = 1.00E-5 φm = 0.2
km = 1 mD

Kf Increases

30
 Reservoir Pressure Depletion Increases as Fracture
Permeability Increases with Constant Fracture Porosity

φf = 1.00E-5

φm = 0.2 Kf Increases
km = 1 mD

31
Comparison of Cumulative Gas Production for Single
(Equivalent Continuum) and Dual Porosity Models

32
 Stochastic Discrete Fracture Network Models

•  Fracture network geometry is characterized by statistical

descriptions of fracture orientation, location, areal extent, and
transmissivity for each of the fracture sets in the rock mass.
•  From this statistical model it is possible to generate multiple
realizations of a fracture network and to solve the fluid flow
through each network.
•  Each realization is one possible representation of the fracture
network in the rock mass from which the geometric properties
were mapped.
•  For each realization a large-order system of equations is solved
to determine the distribution of hydraulic head at all points in
the fracture network.
•  Monte Carlo simulation techniques quantify the magnitude of
the uncertainties in model predictions.

33
Estimation of Model Parameters for Statistical
Models of Fracture Networks
Field data must be collected to estimate the parameters of the
stochastic model used to represent the fracture network
geometry. The following data are normally required:
–  Fracture location as observed in boreholes or along scanlines in
surface outcrops.
–  Fracture orientations from borehole logs, core, or surface
outcrops.
–  Fracture trace length in surface outcrop.
–  The percentage of fractures that terminate against other fractures
as a function of orientation, as obtained from surface outcrops.
–  Estimates of the transmissitivity of individual fractures from
laboratory or field hydraulic tests.

34
 Single Porosity Simulation Models

•  Single porosity models consider flow and transport only in the

open, connected fractures of the rock mass. Matrix is assumed
to be impermeable.
•  When using the model a grid is superimposed on the flow
domain and values of permeability are assigned to each grid
block.
•  Estimation of permeability involves synthesis of laboratory and
in situ measurements, model calibration, and model evaluation.
•  The rock mass is usually represented by a limited number of
hydrological units, each with homogeneous properties.
•  Hydrological units may be isotropic or anisotropic.
•  Because fractures typically occur in sets having preferred
orientations, large scale permeabilities may be anisotropic.

35
 Numerically Generated Fracture Network and Flow
Through the Network Is Calculated in many Directions
and Is Used to Derive the Equivalent Permeability Ellipse

36
Examples of Single Realizations for Three Fracture
Systems with Differing Fracture Densities

37
Three Dimensional Orthogonal Fracture Model

38
Baecher Disk Model of a Fracture System

39
 FRACGEN / NFFLOW

•  Specifically designed to simulate dry gas flow in low

permeability, fractured reservoirs
•  FRACGEN: discrete fracture network generator
–  fracture realization from statistical inputs (termination/
intersection frequencies, length and orientation variation, etc)
–  user defined fracture locations/frequencies can be used to
condition realization
–  fractures are vertical and strata-bound
–  Multiple fracture sets may be defined on multiple layers
–  Several fracture models control swarming characteristics

40
 NFRAC / NFFLOW (cont.)
•  NFFLOW: flow simulator
–  flow through fractures by cubic law
–  fractures recharged by matrix by Darcy flow

41
 Overview of NFFLOW Operation
Initializes simulator:
• reads PVT data
Subroutine: • reads fracture network and determines its geometry
• determines acute angle between each fracture and direction of maximum
INIT principal stress
• computes fracture transmissibilities and matrix block transmissibilities and
volumes.

Controls iteration process required to obtain solution for each time step:
• forms pressure equation, calls linear solver routine, updates potential
• calculates rate at which matrix recharges fracture network, calculates average
STEP
matrix pressure in the formation
• calculates pvt properties for each grid block at each iteration
• computes the well gas flow rates in pressure-controlled wells
• calculates component of effective normal stress acting on each fracture
• calculates amount of fracture closure (or dilation) due to change in
effective normal stress (due to change in gas pressure)
• computes new fracture stiffness
• decreases (or increases) fracture aperture
• re-calculates fracture transmissibilities

Controls the time stepping of the program:

• computes real gas pseudopressure function at the end of a time step
CONTRL • outputs simulation results in existing wells
• outputs simulation results in reservoir nodes
42
Fluid-Flow in a Dual-Porosity
Naturally-Fractured Reservoir

Production Well

Matrix Recharge
-- Darcy’s Law

Fracture Flow
-- Cubic Law

43
Discrete Fracture Network
Characterization and
Fluid-Flow Simulation

44
NFFLOW Simulation Showing Pressure Depletion
for Fractures Near Producing Well

330 Days of
Production

45
 Pressure Decline Curve for
NFFLOW Simulation
No Stress Sensitivity

3850

3800

3750

3700
Pw f (psi)

3650

3600

3550

3500
0 50 100 150 200 250 300 350
Time (days)

46
 Results of NFFLOW Showing Distribution of
Pressure in Fracture Network after 630 Day of
Production in Tight-Gas Sandstone Reservoir

47
NFFLOW History Match Simulation

48
 Upscaling from Local DFN Model to Reservoir Model

Discrete Fracture Network

Equivalent Reservoir
Volume
Kxx

Kyy

Calculated Fracture Porosity

and Permeability for DFN Volume

X
Equivalent Porous Media
(Continuum)
49
Upscaling from DFN Model to Reservoir Model

Discrete Fracture Reservoir Continuum Model

Network Model

50
Upscaling of DFN Model to Reservoir Continuum Model

51
Locations of 100 m2 Area Subsectors
in 1 km2 Study Area
22

52
 Calculated Fracture Porosity & Permeability
for Layer 1 in Subsector 22

DFN Model Calculations

Fracture Aperture = 100 µm

Φf = 0.0072%

KV = 2.26 mD

KHmax = 2.83 mD

KHmin = 0.58 mD

53
DFN Output to Reservoir Fluid-Flow Simulation

Fracture Permeability

54
 Field Examples of Fracture
Characterization and Fluid-Flow
Simulation Applications to Naturally-
Fractured Tight-Gas Reservoirs

55
Rulison Field in the Piceance Basin,
Colorado

Naturally-Fractured Tight-Gas
Sandstone Reservoirs
Mesaverde Formation

Porosity : 6 – 8 %
Permeability : 0.001 mD (Matrix)
0.01 – 50 mD (Reservoir)
Pore Pressure : 4350 psi (30 MPa)
Overburden Stress : 7000 psi (48 MPa)

56
Vertical Partially-Filled Extension Fracture
in Mesaverde Core, Rulison Field

57
 Orientations of Natural Fractures in Core of
Mesaverde Sandstone in Rulison Field,
Colorado
N

W E

Fractures N = 62

S
58
Regional Fracture Network Map of
Mesaverde Sandstone Outcrop

0 feet 25

0 meters 8

59
 Fracture Network is Dominated
by Short Fractures
45
Regional Fractures
40 Mesaverde Sandstone
Piceance Basin, Colorado
35
Number of Fractures

30

25

20

15

10

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40
Fracture Length (m)

60
 Fracture Network is Dominated
by Short Fractures
100
90
80
Cummulative Percent

70
60
50
40
30 Regional Fractures
Mesaverde Sandstone
20 Piceance Basin, Colorado
10
0
0 5 10 15 20 25 30 35 40
Fracture Length (m)

61
Orientation and Distribution of Fractures in Outcrop
Fracture Network Map Relative to Well Locations
in Mesaverde Pilot

MWX - 3

SHCT–1 Deviated Wellbore Azimuth

MWX - 1

0 feet 25

0 meters 8

MWX - 2
62
 Fracture Distribution of Mesaverde
Core and Outcrop Scanlines

Core, SHCT - 1
Natural Fractures

85o
0 10 20

feet
Scanline 1
Outcrop
Scanline 2

Scanline 3

63
Applying Fracture Characterization to
Reservoir Fluid-Flow Modeling

0 feet 25
Warren and Root (1963)
0 meters 8

64
 Regional Fracture Reservoir Models

Dual - Porosity Deterministic - Discrete

Sheet Model Fracture Model

kmin

Fixed Geometry with Non - Uniform Geometry with

Continuous Fractures Variable Fracture Orientation,
having Equal Spacing Length, Aperture, Intensity, and
and Constant Aperture Interconnectivity

65
 Fluid-Flow Simulation Indicates
Reservoir Permeability is Anisotropic

Kmin = 0.57 mD

Kmax = 86 mD

0 feet 25

Kmatrix = 0.001 mD 0 meters 8

66
Effect of Fracture Length and Hydraulic Aperture
on Deterministic Fracture-Network Model
Fluid-Flow Simulations

kHmax kHmin kEffective

(mD) (mD) (mD)

All Fractures
Aperture = 0.1 mm 95.3 2.59 15.71

All Fractures
Aperture = 0.05 mm 5.7 0.16 0.96

Fracture Lengths > 10 m

Aperture = 0.1 mm 39.2 0.02 0.59

Fracture Apertures are

Function of Length 86.0 0.57 7.01

67
 Model Comparisons of
Calculated Horizontal Permeability

kHmax kHmin kEffective

MODEL (mD) (mD) (mD)

Deterministic Model with

Constant Aperture 95.3 2.59 15.71

Deterministic Model with

Variable Aperture 86.0 0.57 7.01

Sheet Model 128.2 0.001 0.36

68
Comparison of Fluid-Flow Models to Well-Interference Test
of Horizontal Permeability in the Cozzette Sandstone
in Mesaverde Pilot, Rulison Field

kHmax kHmin kEffective

(mD) (mD) (mD)

Deterministic Model with

Constant Aperture
95.3 2.59 15.71

Deterministic Model with

Variable Aperture 86.0 0.57 7.01

Sheet Model 128.2 0.001 0.36

Well Interference Test 50 1.0 7.07

69
Development of Stochastic DFN Model
of Cozzette Sandstone in Mesaverde Pilot
•  Horizontal Core Data
Fracture Strike and Dip
Fracture Distribution (Spacing and Clusters)

•  Outcrop Data
Fracture Strike and Dip
Fracture Length Distribution
Fracture Length vs Orientation
Number and Type of Fracture Intersections
Fracture Spacing vs Length
Fracture Cluster Distribution

70
3-D View of Stochastic DFN Model of Cozzette
Sandstone at MWX Site in the Rulison Field

100 ×100 × 4 m

71
 Map View of Stochastic DFN Model for Cozzette
Sandstone at MWX Site in the Rulison Field and
Calculated Fracture Porosity & Permeability

DFN Model Calculations

Variable Fracture Aperture
(25 µm to 100 µm)

Φf = 0.012 %
KHmax = 58.61 mD

KHmin = 0.79 mD

KHeff = 6.80 mD

100 × 100 × 4 m
72
Comparisons of Calculated Horizontal Permeability
of DFN Models to Well Interference Test

kHmax kHmin kEffective

(mD) (mD) (mD)

Stochastic Model with

Variable Aperture 58.6 0.79 6.80

Deterministic Model with

Variable Aperture 86.0 0.57 7.01

Well Interference Test 50.0 1.0 7.07

73
 Discrete Fracture Network
Fluid Flow Model

Effect of Fracture Interconnectivity

on Reservoir Productivity

74
Reservoir Simulation Study of Discrete Fracture Network
Models of a Naturally-Fractured Tight Gas Reservoir

75
Study Involved 30 Simulations of Discrete Fracture
Networks That Are Statistically Similar

76
Production Profiles of the Single Well Case Show Flow
Rates Vary by an Order of Magnitude

77
Difference in Well Productivity is Not Related to Total
Fracture Network Connectivity of Study Area

78
Comparison of Pressure Maps for Two Single Well Simulations:
Well A Produced 5 Times more Gas than Well B Because of Better Local
Communication with Regional Fracture Network

79
 Comparison of Detail Maps of Fracture Network
Surrounding Wells Shows Greater Local Fracture
Connectivity in the Higher Producing Well A than Well B

80
Direct Relationship Between Gas Production and Local Fracture
Connectivity and Fracture Intersections Near the Well

81
Five-Spot Well Pattern for Simulation Study

82
Production Profiles from 30 Trails of the Five-Well Case:
Total Production was Divided by Five to Yield a Per-Well Average
Cumulative Production

83
Histograms Comparing Cumulative Production (per-well
average) for Simulations Using 1, 2, 3, 4, and 5 Wells

84
 Inverse Relationship Between the Number of Wells
Producing from a Regional Fracture Network and the
Standard Deviation of Cumulative Production

85
Anisotropic Discrete Fracture Network with
a Dominant North-South Orientation

86
Effect of Fracture Anisotropy on Production
of Horizontal Wells

87
 Integrating 3-D Geological and Fluid Flow
Models in the Naturally-Fractured Ekofisk Field
Using a Continuum Model Approach

88
Ekofisk Platform,
North Sea

89
 North Sea Map
Showing Location
of Ekofisk Field

90
 EKOFISK FIELD, NORTH SEA
NATURALLY FRACTURED CHALK
RESERVOIRS

Porosity : 25 - 48 %

Permeability : 1 – 5 md (Matrix)
1 – 160 md (Reservoir)

Oil in Place: 6.9 Billion Barrels

Pore Pressure : 7100 psi (49 MPa)

SALT Overburden Stress : 9100 psi (63 MPa)

91
Shear Fractures in Ekofisk Chalk

92
93
Reservoir Characterization Process

Reservoir Architecture
Hydrocarbon Pore Volume

Heterogeneity
Anisotropy

Upscaling

Fluid Behaviour
Rock-Fluid Interaction

Figure 2. Reservoir Characterization process. To describe the distribution of pore volume, an architecture must first be built in three
dimensions that contstrains the rock-fluid system being described.

94
 Upscaling Geologic Hydrocarbon Model to
Reservoir Flow Model
RESERVOIR FLOW
MODEL

UPSCALING

HIGH RESOLUTION
HYDROCARBON MODEL

DOWNSCALING
ING
CAL
S
UP

LOW RESOLUTION
HYDROCARBON MODEL

Fig. 7.--Schematic illustrating upscaling-downscaling loop between geological and flow model.
95
3°10 ' 3°15 '
Ekofisk Fault Model Development

1991 1995

Figure 3. Comparison of the fault model developed through the Ekofisk Reservoir Characterization project to a fault model developed through
a previous effort in 1991.

96
Porosity Distribution Map Development

Ekofisk Reservoir Characterization

4
1

01
6
1

1 2
8 4 2
2 6
0
W
1 2 W

2
W
W
2

0
W
W
6

1
W 2 8

6
W

2
41 3 0 3

1
2
0

2 2
2
2

8
W

4 2
W

W W
W W W
W

W
W

4 3
3 W
W
6 W
W
W
3 W

8
3

4
4
8 W
1
6

2
W
3 W

6
3 4
W

2
6 W
W

2
3

0
3

3 8
G
G W

3
W W

4
W
G
G W
W
3 W
2 W
3 G

8 2
3 38
8

G
1

4
4

26
W
0

W
2 W
8

W W
W W 3
2
2

W
W 8
6

36
4

4
2

1
2

2 W
W
0
W
2

8
4
W W
W W

2
W
2

2 2

1
3

2 G
6 3

6
G

4
3
W W

02
4 W
2 W

8
1
2
4 3 2

2
3
W
W
6 2

W
W W

3
W

0
8

6
2

2
W
W W
3

W
0

4
2
2
2
3 0

3
62
4

0 2

0 2
2 4

3
2

0 2
6

2
2 2

81
2 2
4
8
2

2
4

62
0

0 2
2

8 2

Old New
2

0
0

Figure 3 Ekofisk Formation Porosity Comparison

Figure 4. Map of one of the Ekofisk Formation layers comparing the ERC porosity distribution with a previous characterization effort. 97
 2 Pebble Floatstone Category 3
3 Burrowed Homogenous Chalk Category 2
4 Laminated Chalk Category 2

Fractures in Ekofisk Chalk5

6
Burrowed Laminated Chalk
Burrowed Argillaceous Chalk
Category 1
Category 1

(a) (b) (c) (d)

98
Fig. 1.--Photographs of fracture types from cores from the Ekofisk field: (a) tectonic fractures, (b) stylolite-assocaited
Fracture Intensity by Fracture Type for Different
Petrofacies Groups in theandEkofisk Formation
Fig. 1.--Photographs of fracture types from cores from the Ekofisk field: (a)
fractures, (c) slump fractures, (d) healed fractures. Scales are in 5-cm i

1.2 2.5
Heale d
1 Stylolite
2
NUMBER OF FRACTURES PER FOOT

NUMBER OF FRACTURES PER FOOT

Tectonic
0.8 Irregu lar
Slu mp 1.5
0.6
1
0.4

0.2 0.5

0 0
Category 1 Category 3
Category 2
PETROFACIES GROU P

Fig. 2.--Fracture intensity by petrofacies group for the Ekofisk Fig.

99 3.--Fr
formation. (a) formation
Procedure for Distributing Effective Permeability

100
Flow Chart for Calculating Effective
Reservoir Permeability

Quartz
Formation

Petrofacies
Porosity Fracture Effective
Intensity Permeability
Fracture Type

Structural Location Gaussian Curvature

Match Well Test

Data - Geostatistics

101
Upscaling Geologic Hydrocarbon Model to
Reservoir Flow Model
RESERVOIR FLOW
MODEL

UPSCALING

HIGH RESOLUTION
HYDROCARBON MODEL

DOWNSCALING
ING
CAL
S
UP

LOW RESOLUTION
HYDROCARBON MODEL

Fig. 7.--Schematic illustrating upscaling-downscaling loop between geological and flow model.
102
3°10 ' 3°15 '
 Ekofisk Reservoir Permeability Anisotropy
•  Fluid flow is governed by the distribution, orientation, and
interconnectivity of the natural fractures and fault system.
•  Over 300 fault planes mapped, majority of large faults defined
in the fluid flow model.
•  Faults control local pressure and/or fluid flow direction.
•  Natural fractures affect reservoir permeability anisotropy.
•  Fracture density was used as an additional basis for
determining the value of anisotropy.
•  Areas of higher fracture intensity were assigned higher values
of anisotropy and vice versa.
•  Anisotropy values ranged form 10:1 to 3:1.
•  History match required tuning of the anisotropy values, but the
relationship to fracture density was maintained throughout.

103
 DOWNSCALING
G
LIN
SCA
UP

Maps of Faults and Associated Effective

LOW RESOLUTION
HYDROCARBON MODEL
Permeability for Ekofisk Formation
Fig. 7.--Schematic illustrating upscaling-downscaling loop between geological and flow model.

3°1 0' 3°1 5'

6270000
56°34'
56°32'

6265000
56°30'

6260000

51 000 0 51 500 0

KM 0 1 2 KM

Fig. 8.--Fault map drawn on top Ekofisk. X direction modifier template downscaled from fluid flow model to
geological model reflecting major fault patterns shown in plot on the left.
104
Permeability Anisotropy Map for Ekofisk Fluid Flow Model

Brown areas represent the fault network

and are the higher values of anisotropy.

Blue areas reflect the regional trend

105
Effective Permeability Distribution Map for
Ekofisk Formation

106
 Total Field GOR Match for Ekofisk

Green line reflects actual and red line reflects simulated response.
107
Total Field Watercut Match for Ekofisk

Red line reflects actual and blue line reflects simulated response.
108
 Water-Oil Relative Permeability Curves
for Low and High Fracture Intensity Regions in Ekofisk

109
 Gas-Oil Relative Permeability Curves
for Low and High Fracture Intensity Regions in Ekofisk

110
 Reservoir Management

Structure & Facies

Rock Properties & Fluids
Reservoir Description

Fluids/Pressures/Production
Grid Refinement
Core & Logs
Residual/Pseudos
Seismic
Data Collection Flow Model Development

Stimulations Production & Reserve Performance Evaluation

Recompletions Enhancement

Waterflood & IOR

Reservoir & Well Forecasts
Infill & Flank Drilling
Integrated Long Term Forecasts

Figure 9. Reserves optimization cycle as associated with the Ekofisk Reservoir Characaterization project.

111