Sie sind auf Seite 1von 105

5.

Fracture Rock Properties


Affecting Reservoir Performance

1
Naturally Fractured Reservoirs
can be Characterized as a
Dual-Porosity/Dual Permeability System

2
Classification of Naturally Fractured Reservoirs Based
on the Contribution of Fractures to Total Reservoir
Porosity and Permeability

3
Fractured Reservoir Porosity

ФFR = ФF + ФM
Fracture porosity is usually less than 1% of total porosity.

Fracture porosity depends on


•  Fracture Morphology (open, partially-filled, or filled)
•  Fracture Size (aperture, length, and height)
•  Fracture Network Density (geometry and spacing)
•  Effective Stress (σ – PP)

4
Fracture and Matrix Porosity
Communication and Interaction

•  Cross-flow in a two porosity system (fractures and


matrix) is essential in determining reservoir drainage
area and recovery factor.

•  Natural fractures may have uninhibited cross flow or


inhibited cross flow between the matrix porosity and
fracture.

•  Type of cross flow depends on fracture morphology.

5
Reserves

φTOTAL = φMATRIX + φFRACTURE


φRECOVERABLE = aφMATRIX + bφFRACTURE

Fracture & matrix porosity values exhibit different behavior with pressure
since fractures are highly compressible compared to the matrix. This
compressibility difference may be used to determine the quantities of
hydrocarbon by a material balance calculations.

6
Fractured Reservoir Permeability

KFR = KF + KM

Fracture Permeability Depends on


•  Fracture Morphology (open, partially-filled, or filled)
•  Fracture Size (hydraulic aperture, length, and height)
•  Fracture Network Density (geometry and spacing)
•  Fracture Network Interconnectivity (geometry and spacing)
•  Effective Stress (σ – PP)
•  Fluid-Flow Communication Between Matrix and Fractures

7
Petrophysical Properties Required for
Evaluating Naturally Fractured Reservoirs

•  Fracture Porosity
•  Fracture Permeability
•  Fracture Compressibility
•  Fluid Saturations within the Fractures
•  Recovery Factor Expected from the Fracture System

The key to economically producing NFR lies in:


1. Evaluating recoverable reserves as a function of well costs.
2. Predicting optimum well locations & well performance with time
under a variety of potential completion & development scenarios.
Primary recovery patterns could be inappropriate for secondary
8
recovery
Fracture Porosity (φf) and
Matrix Porosity (φr) Are Basically
Different Quantities

9
The Basic Relationships Used to
Calculate Fracture and Matrix Porosity

Vp e
φr = × 100 φf = × 100
Vb D+e

where
ϕr = matrix porosity
ϕf = fracture porosity
Vp = Volume pores (other than fractures)
Vb = Bulk volume
D = Average spacing between fractures
e = average effective width of fractures 10
Important Differences Between Fracture
and Matrix Porosity

•  Fracture porosity is very scale dependant, while matrix


porosity is not.

•  While fracture porosity is very small, it is usually highly


interconnected and does, therefore, have a much more
dramatic effect on permeability than matrix porosity.

•  Fractures compress or reduce porosity and permeability


much more readily than the matrix with increasing stress.
Accordingly, fracture porosity and permeability is much
more sensitive to reservoir depletion.

11
Fracture Porosity Determination is Scale Dependent

A B

Fracture

ϕf determinations are dependent on the size & positioning of the sample, 12


while ϕm determinations are not.
Calculated Fracture Porosity is a Function of
Fracture Spacing and Fracture Width

Variation in fracture
spacing can have a
dramatic effect on
ϕf. The combined
effect of both
fracture width &
spacing on ϕf is
shown here. A good
qualitative feeling
for the effect of
outcrop or core
observations of
fracture spacing at
an assumed fracture
width, or vice versa,
can be derived from
these diagrams.
13
Fracture porosity is usually low but can be important in
specific reservoirs with large vertical and areal extent.

- Always less than 2.0%

- Excluding small zones less than 1.0%

- General less than 0.1%

14
While Fracture Porosity φf is Small in Magnitude

Ÿ  φf is very effective porosity.

Ÿ  Possible large volume due to large drainage


area and reservoir thickness.

Ÿ  Assume recovery great.

ϕf has a recovery factor generally assumed to be greater than that of


the matrix, in part due to the assumed low Sw in smooth, open
fractures. The potential for substantial reservoir volume residing in
fracture systems even of 1 % ϕf is high, depending on reservoir
thickness & drainage area.
15
Relative Contribution of Fracture Porosity to Total
Reservoir Porosity Decreases with Increasing Matrix Porosity

16
Fracture porosity is very difficult number to calculate.
Estimates can be made by:
•  Core Analysis

•  kf/φf Relationship (Using Dual Porosity or DFN Models)

•  Field Determinations (Surface Outcrop Analogs)

•  Logs There is no direct method of calculating ϕf from well logs. Several


log suites have been developed to detect natural fracture systems, but
none can calculate ϕf directly.

•  Multiple Well Tests (Using Dual Porosity or DFN Models)


Pressure transient analyses probably give the most accurate
estimates of ϕf.

As many methods as possible should be used.

17
Fracture Porosity (φf)

Logs and well testing will not tell us how φf is


distributed &, therefore, will not tell how it changes
with depth & depletion.

φf is difficult to calculate & even more difficult to


interpret once determined.

Interpretation of the significance of φf depends on the


type of fractured reservoir.
18
Well-fracture intersections are considered from a probabilistic perspective. Vertical wells could
intersects few vertical fractures.
Fracture Azimuth = 96o ± 12o
Near horizontal core in the Paludal sandstones. Natural fractures are abundant in the sandstone
reservoir intervals, natural fractures are not present in mudstones.
Fracture Azimuth = 102o ± 10o
Fracture occurrence in subhorizontal core from the SHCT-1 well in Cozzette Sandstone. A characteristic of
21
east-striking fractures is that they are arranged in swarms defined by areas of close fracture spacing. Swarm
widths range from 2 inches to more than 160 ft. Fracture spacing within swarms is variable.
Calculate Fracture Porosity from Core

By subtracting the consistently low ϕm (taken from the average of


the unfractured samples) from the whole core samples containing
fractures, an estimate of ϕf is made. This method of calculating ϕf
is, of course, fraught with scale & sampling problems.

22
Fracture Spacing in MWX SHCT-1 Well
Mesaverde Formation, Rulison Field, Colorado

Individual Fractures Fracture Swarms


Minimum Maximum Average Minimum Maximum Average
(m) (m) (m) (m) (m) (m)

Paludal: < 0.03 1.83 0.854 0.762 2.378 1.463


2 Lenticular Sandstones

Cozette: < 0.03 5.34 0.915 1.220 5.336 2.744


Blanket Sandstone

23
Homework: Calculate Fracture Porosity from
Dual-Porosity Sheet Model Using Core Data

Aperture Spacing Porosity


(cm) (cm) (percent)

Paludal: 0.1 85.4


2 Lenticular Sandstones

Cozette: 0.1 91.5


Blanket Sandstone

One Parallel Fracture Set with Equal Spacing:


Фf = e/d x 100
Where
Porosity - Φf is in %
Aperture - e is in cm
Spacing - d is in cm 24
Homework: Calculate Fracture Porosity from
Scanlines from Outcrop Fracture Network Map

Фf = e/d x 100
Фf = fracture porosty (%)
B
e = fracture aperture (cm) A
d = fracture spacing (cm)
(scanline length / C
# fractures)

For calculation
D
e = 0.1 cm, 0.01 cm
Scanlines
A & D = 10 m
B & C = 30 m

25
Hydraulic Fracture Width

Experimental Fracture Widths at 10,000 ft (Simulated Depth)


(Number of Samples Not Statistically Significant)
Medium to Coarse Grained Sandstone 10-2 cm
Fine to Medium Sandstone and Crystalline Carbonates 10-2 cm
Siltstones 10-2 cm
Shales (Textural Term) 10-2 cm
Chalks (Compositional and Texture Term) 10-2 cm - 10-6 cm

Some Published Natural Fracture Widths


Noorishad and Others (1971) 3.0 x 10-2 cm
Ohnishi and Goodman (1974) 1.3-2.5 x 10-2 cm
Sharp and Others (1972) 1.0–5.0 x 10-2 cm
Snow (1968a) 5.0 x 10-3 cm
Snow (1968b) 0.5-1.5 x 10-2 cm
Van Golf-Racht (1982) 1.0-4.0 x 10-3 cm
26
Wilson and Whiterspoon (1970) 2.5 x 10-2 cm (mean)
Fracture Length, Lateral Distribution and
Interconnectivity Can Not be Determined from Cores

Fracture Azimuth = 102o ± 10o

27
DFN Model Development
Using Core Data and Geologic Analog Based on Outcrop
Fracture Network Map for Mesaverde Tight-Gas Sandstone
Intersection/termination
frequencies not only are
useful for estimating MWX - 3
fracture length
distributions; they also SHCT–1 Deviated Wellbore Azimuth
are essential to properly
model flow & network
appearance. Flow
simulation work
demonstrated that it was
insufficient to merely MWX - 1
know the fracture length
distributions. It is 0 feet 25

imperative also to control


intersection/termination 0 meters 8

frequencies, because a
network's connectivity MWX - 2
affects flow rates. 28
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

29
Calculated Fracture Porosity from Outcrop Analog of
Regional Fracture Network in Mesaverde Sandstone

Fracture Azimuth = 100o ± 11o


Average Fracture Porosity from 20 Scanlines = 0.0154% ± 0.005%
with Fracture Aperture = 0.01 cm (100 µm)
30
Fracture Network is Dominated by Short Fractures
Fracture Aperture May Decrease with Decreasing Fracture Length
45
Regional Fractures
40 Mesaverde Sandstone
Piceance Basin, Colorado
35
Number of Fractures

30

25

20
293 Fractures
15

10

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

31
Calculation of Fracture Porosity from
Outcrop Fracture Network Map

Фf = e/d × 100
Фf = fracture porosity in %
e = fracture aperture (cm)
d = fracture spacing (cm)
(scanline length / Scanline
number of fractures)

32
Block Schematic of Subparallel, Poorly Connected
Regional Extension Fractures in Heterogeneous,
Tight-Gas Sandstone Formation

Modeling connectivity is an objective of the fracture


network modeling effort.
33
Dual Porosity Models
2-ϕ models assume that fractures
separating matrix blocks are
symmetrically repetitive, then we
could have slabs, where flow is 1D,
columns, with 2D flow, & cubes,
where flow is 3D. Reiss calculated
the ϕf & kf of slabs, columns, &
cubes in terms of fracture spacing
& fracture aperture. k of a single
fracture when a fluid flows
between 2 parallel slabs with a
given aperture is given by a2/12.
Combined km & kf depends on the
dimensions of the matrix blocks,
thickness of the fracture, # of sides
of the matrix blocks in contact with
the fracture, & the direction of flow.
34
a is fracture spacing (cm) ---- b is fracture aperture (µm)
Reiss (1980) has worked out the relationship between ϕ & k for several cases
Calculated Porosity for Fracture Network with
1. One Parallel Set with Equal Spacing
Фf = 0.01 (e/d)

2. Two Orthogonal Sets with Equal Spacing


Фf = 0.01 (2e/d)

3. Three Orthogonal Sets with Equal Spacing


Фf = 0.01 (3e/d)

Where fracture porosity – Фf is in %


aperture – e is in microns
spacing – d is in cm

36
0.25

Parallel Set of Fractures with Equal Spacing


0.2
Fracture Porosity (percent)

0.15

0.1

0.05
Fracture Aperture = 0.2 mm
0.05 mm 0.1 mm
0
0 25 50 75 100 125 150 175 200 225
Fracture Spacing (cm)
Using eqn for one parallel set, we have for slabs that the ϕf in terms of
aperture & spacing is given by this figure
0.25
Fracture Network of Two Orthogonal
Fracture Sets with Equal Spacing
0.2
Fracture Porosity (percent)

0.15

0.1

0.05
Fracture Aperture = 0.2 mm

0.05 mm 0.1 mm
0
0 50 100 150 200 250 300 350 400 450
Fracture Spacing (cm)
Using eqn for 2 orthogonal sets, we have for columns that the ϕf in terms of
aperture & spacing is given by this figure
0.25
Fracture Network of Three Orthogonal
Fracture Sets with Equal Spacing
0.2
Fracture Porosity (percent)

0.15

0.1

0.05

Fracture Aperture = 0.2 mm


0.05 mm 0.1 mm
0
0 50 100 150 200 250 300 350 400 450 500 550 600 650
Fracture Spacing (cm)
Using eqn for 3 orthogonal sets, we have for cubes that the ϕf in terms of
aperture & spacing is given by this figure
2.5
Fracture Network of Three Orthogonal
Fracture Sets with Equal Spacing
2
Fracture Porosity (percent)

1.5

0.5

Fracture Aperture = 0.05 mm 0.075 0.1 mm 0.125 mm


0 mm

0 2 4 6 8 10 12 14 16
Fracture Spacing (cm)

Using eqn for 3 orthogonal sets, we have for cubes that the ϕf in terms of
aperture & smaller spacing is given by this figure
Calculation of Fracture Porosity from
Outcrop Regional-Fracture Network Map

0 feet 25

0 meters 8

Фf = 0.01 e/d
e = fracture aperture (100 µm)
d = fracture spacing (cm)
(scanline-length / number of fractures)
Фf = Total Fracture Length X Fracture Aperture/Area
Mean = 0.0154 % Constant Aperture (100 µm) = 0.0148 %
STD = ± 0.0045 % Variable Aperture (fn of length) = 0.0136 %
(e = 50 to 200 µm)
Range
Max = 0.0226 %
* Both methods assume fractures are vertical and strata
Min = 0.0098 %
bound with constant height equal to bed thickness. Both
methods yield similar results.
Permeability

•  Equations for Fluid Flow


•  Matrix Permeability
•  Fracture Permeability (function of fracture
width, fracture spacing)

•  Fracture Network Permeability (function


of fracture density & connectivity of fracture
network)

42
Fluid-Flow in a Dual-Porosity
Naturally-Fractured Reservoir

Production Well

Matrix Recharge
-- Darcy’s Law

Fracture Flow
-- Cubic Law

q= e3 /(12 d) ; e= aperture, d= spacing 43


Equations for Fluid Flow– Rock Matrix

Darcy (1856)

dh
Q = KA
dl

where:
Q = Flow Rate
K = Hydraulic Conductivity (k/µ for single-phase flow)
A = Cross-Sectional Area
dh/dl = Head Gradient (pressure gradient)
44
Equations for Fluid Flow - Fractures
In an attempt to model fractures, the parallel-plate theory of flow
was developed.
Flow in this theory is assumed to occur between two smooth
parallel plates separated by a distance, e.
The basic equation is:
e dh ρ g
3
Q
=
A 12 D dl µ
where:
D = Fracture spacing, the average distance between parallel
regularly spaced fractures.

This equation is valid for single-phase, newtonian, laminar flow in


planar fractures with small overall changes in width, e.
45
Approach Used to Model Fluid Flow in a
Naturally Fractured Reservoir

Each of the two qualitative relations describe only a portion


of the total flow through a fractured, porous rock.

•  Darcy’s equation is valid.

•  Parallel-Plate theory (cubic law) for the fractures,


kf = e3/ (12 D)

46
Equations for Fluid Flow in a
Naturally Fractured Reservoir
A simple approach to determine the total flow is to combine these
equations (Parsons, 1966):
3 2
e cos α
k fr = k r +
12 D
where:
kfr = Permeability of the fracture plus intact rock system
kr = Permeability of the intact rock matrix
e = Fracture aperture
D = Fracture spacing between one set of parallel equally spaced
fractures

α = Angle between the axis of the pressure gradient and the


fracture planes.
This is approx. because the kr contributes to the flow in the source term but not to the primary flow,
but even if the matrix contributes & they have parallel flow, kfr would be an arithmetic average.
Assumptions for Equations for Fluid Flow in a
Naturally Fractured Reservoir

This equation assumes that flow is laminar between


smooth, nonmoving, parallel plates, that fluid flow across
any fracture/matrix surface does not alter the flow of either
system, and that the fractures are open, homogeneous with
respect to orientation, width, and spacing.

48
Matrix Permeability

•  Effect of Porosity

•  Effect of Stress

- Hydrostatic Stress

- Deviatoric Stress

49
Porosity vs Log Permeability

In theory there are regions in this plot of permeability vs. porosity


for each type of rock.
Effect of Depth of
Burial on Porosity and
Permeability

The increase in depth, at which


pore pressure is hydrostatic, is
associated with a uniform
increase in the vertical effective
stress & a corresponding
decrease in porosity &
permeability.

51
Routine Laboratory Compression Tests

52
Matrix Permeability Decreases with
Increasing Hydrostatic Stress
Lab tests where
hydrostatic
confining pressure
is applied. km
decreases when
net confining
pressure
increases. Initial
km of the samples
varies between
0.04 md to 191 md
& reflects the
variation of initial
ϕm, grain packing,
clay abundance &
cementation.
km loss is higher
for the lower initial
km values.

53
Stress-Dependent Matrix Permeability is
Inelastic

54
Uniaxial Strain Test

S1

S3

S3 = KoS1
Ko is Uniaxial Strain
Compaction Coefficient

55
Matrix Permeability as a Function of Maximum Effective
Stress for Hydrostatic and Uniaxial Strain Loading
(Nelson, 1981)

Because the state of stress


due to burial in the
subsurface is distinctly
anisotropic, the direct use
of standard lab testing
procedures in burial Permeability measured parallel
simulations is difficult if to σ1 in uniaxial strain tests.
not misleading. Estimates
of reservoir quality (ϕm &
km) are function of
maximum effective stress
for both hydrostatic &
uniaxial strain loading

56
Effect of Confining Pressure and Deviatoric Stress on
Sandstone Permeability During Triaxial Compression
Deformation to Brittle Fracturing
Deviatoric
stress is
obtained by Permeability measured
subtracting parallel to σ1

the mean
normal stress
Microcrack
from the
Dilation
normal stress
components. Elastic
Anisotropy
may be caused
by
microcracks,
Confining pressure
generated by a reduces permeability
deviatoric
stress &
predominantly
oriented 57
normal to σ3
Lab Measurements Show that the Change in Matrix
Permeability & Permeability Anisotropy with Pore
Pressure Drawdown is a Function of Stress Path
(Khan & Teufel, 2000)

Hydrostatic Uniaxial Strain Triaxial Compression


km of sandstone at 5 MPa confining pressure was 0.1 md & ϕm was 12%. Effects of
depletion for different stress path loading conditions on km measured parallel (kv) &
perpendicular (kh) to σ1 direction. For isotropic loading, R=1, km decreases by the same
magnitude in both directions with increasing effective stress as pp is reduced. After a pp
reduction of 40 MPa, km was reduced to about 40% of the initial value. As the stress path
is reduced & deviatoric stress is applied to the sample, km decreases with increasing
effective stress, but the magnitude of reduction in km is < in the direction parallel to the
maximum (σV). At a stress path of R=0, km in the σ1 direction is > than in the initial 58
conditions because of the large deviatoric stress & stress anisotropy imposed on sample.
Fracture Permeability

•  Effect of Fracture Morphology

•  Test of Cubic Law

•  Effect of Stress

•  Effect of Shear Deformation

•  Effect of Fracture Spacing and Fracture Width

59
Fracture Morphology Effects Permeability of
Natural Fractures

•  Open Fractures
•  Deformed Fractures
- Gouged Filled
- Slickensided
•  Mineral Filled Fractures
•  Vuggy Fractures

60
Whole Core Permeability of Open Fracture

61
Slickensided Shear Fracture in Mesaverde
Sandstone Core from the Rulison Field, CO

62
Whole Core Permeability of Deformed Fracture
with Slickensides

63
Cataclastic Microfault Zones in Sandstone Core

Displacement of rock masses along the fracture plane causes cataclasis or


granulation of the grains in contact across the fracture. This granulation or
cataclastic zone can be several grain diameters wide, & it reduces ϕ, grain size,
64
sorting, & k of the fracture zone.
Permeability Across Fracture Gouge Zone

If asperities are torn off & gouge


produced within the fracture,
significant permeability
reductions (2 or more orders of
magnitude) have been observed

65
Fracture Filled by Secondary Mineralization

Core plug with a fracture filled with planar quartz-rich structures66


Comparison of Vuggy Fracture and Matrix Core
Plug Permeability

Presence of low permeability quartz filled fractures explain the


67
reduced permeability values.
Fracture Permeability & Porosity are Stress Sensitive & Decrease
with Pore Pressure Depletion & Increasing Effective Stress

Pp

σN

Fracture Fracture
Aperture Permeability

Normal Stress Normal Stress


Hyperbolic Model for Normal Stress Correction of Aperture. Based on
lab testing with fractures, the influence of normal stress on kf is 68
generally hyperbolic (or exponential).
Parameters Affecting Coupled Stress-Sensitive
Deformation & Permeability of Fractures

•  Fracture Morphology
- Surface Roughness
- Filling

•  Strength and Compliance of Surface Asperities

•  Effective Normal Stress on Fracture

•  Shear Stress on Fracture

69
Surface Topography Measurements

Quantify
•  Surface
Roughness
•  Mismatch
•  Aperture

70
Effect of Normal Stress on Fracture Aperture
and Fluid Flow

71
Laboratory Test of the Cubic Law for Fracture
Flow

72
Sketch of Lab Test Showing Flow through
Fractured Core with Applied Effective Stress

73
Depth and Depletion Corrections Are Much
More Severe in ϕf & kf Than in ϕr & kr

Normalized porosity (ø/ø500) & permeability (k/k500) are shown as a function of


hydrostatic confining pressure in tests on rock matrix & fractures. As
confining pressure increases due to either increasing depth of burial or
depletion, fractures compress or reduce in ϕf & kf much more readily than the
matrix. This difference in behavior is most dramatic in relatively brittle rocks
74 &
less dramatic in rocks with a ductile matrix.
Fracture Conductivity of Vertical Fracture in
Limestone Core is Stress-Sensitive and Inelastic
Change in
conductivity due to
loading &
unloading of a
limestone, where
conductivity loss
due to burial is not
recovered upon
unloading, when
the fracture closure
occurs & the rock
goes into an
inelastic
deformation. At low
stress values, the
limestone goes into
an elastic
deformation & the
conductivity is 75
recovered.
Vertical Partially-Filled
Regional Fractures in
Mesaverde Sandstone Core
from Piceance Basin,
Colorado

76
Lab Measurements Show that Small Partially-
Filled Fractures have Higher Permeability than
Matrix Rock at Reservoir In Situ Conditions

77
Comparison of Matrix & Fracture
Stress-Dependent Permeability

Hairline Tectonic Fractures

A core with higher k at atmospheric conditions due to hairline tectonic fractures


is more dependent on confining pressure than a core with lower k & less
78
fractures.
Conductivity of Natural Fractures in Mesavede
Sandstones in the Rulison Field have a Higher Stress
Dependence with Lower Initial Fracture Conductivity

79
Conductivity of Vertical Fracture in Mudstone
Core is Stress-Dependent & Inelastic

Tests were run at constant pp cycles, both through loading & unloading. Fig
shows a factor-of-five drop in conductivity on the initial cycle. Because of80
the time dependence, the mudstone could not be used for other testing.
Fracture Conductivity is Stress-Sensitive &
Depend on Fracture Lithology and Morphology

81
Increasing Effective Normal
Stress will Reduce Fracture
Permeability to Matrix
Permeability

kf is very much reduced with the


increase of the confining
pressure, which is equivalent to
the increasing net overburden
pressure, approaching km when
the fracture is closed.

82
Effect of Stress on Chalk Permeability

Over 20 years of production from the Ekofisk field has resulted in a 21-24
MPa reduction in pp. kf is very much reduced with the increase of σv,
approaching km when the fracture is closed. 83
Effect of Shear Deformation on Fracture Permeability

kf gradually reached a maximum threshold value. Combined effects of both


asperity degradation & gouge production, which prohibited the subsequent
enlargement of mean fracture aperture, mainly caused a reduction on kf.
Shear Fracture in Laboratory Specimen

85
Deformation of Shear Fracture
Coconino Sandstone Pe = 15 MPa

Differential stress, change of volume, & k are plotted vs axial strain, for an
effective pressure of 15 MPa. The stress-strain curves are similar to those of
samples that were subjected to continuous loading at corresponding 86
effective pressure.
The Problem of Scale When Modeling
Deformation and Fluid Flow in Naturally
Fractured Rock Mass

87
Size of the sample needs to be representative to model deformation & fluid flow in NFR
Simple Idealized Fracture Networks and Matrix
Blocks for Dual Porosity Models (from Reiss, 1980)

Arrows indicate possible directions of flow


Reiss considered simple geometric fracture systems of slabs, columns,
& cubes to find the relationships between ϕf, kf, frequency, & flow
direction 88
a is fracture spacing (cm) ---- b is fracture aperture (µm)
from Reiss (1980)
Calculate Fracture Permeability from Core

90
Fracture Azimuth = 102o ± 10o
Average Fracture Spacing = 91.5 cm
91
Calculate Permeability of Fracture Network
of One Parallel Set with Equal Spacing

Kf = 8.33 x 10- 4 (e2 Фf)

Where fracture permeability – Kf is in Darcys


porosity – Фf is in percent
aperture – e is in microns
spacing – d is in cm

92
Homework: Calculate Porosity & Permeability from
Dual-Porosity Sheet Model Using Core Data
Fracture
Aperture Spacing Porosity
Permeability
(µm) (cm) (percent)
(mD)

200 91.5

100 91.5

50 91.5

25 91.5

One Parallel Fracture Set with Equal Spacing:


Фf = 0.01 (e/d) Kf = 8.33×10-4 (e2 Фf)
Where
Porosity - Φf is in %, Aperture - e is in microns, Spacing - d is in cm, Kf is in Darcy
7000
Parallel Set of Fractures
6000 with Equal Spacing
Fracture Permeability (md)

5000 Fracture Aperture = 0.2 mm

4000

3000

2000
0.1 mm

1000
0.05 mm
0
0 0.05 0.1 0.15 0.2 0.25
Fracture Porosity (percent)

For slabs, if we keep the fracture aperture constant, changing the ϕf, i.e.
changing the fracture spacing, the kf curves are given by these lines.
Calculated Fracture Porosity & Permeability from
Outcrop Analog of Regional Fracture Network of
Mesaverde Sandstone Using Dual Porosity Sheet Model

Fracture Azimuth = 100o ± 11o

Average Fracture Porosity from 20 Scanlines = 0.0154% ± 0.005%


with Fracture Aperture = 0.01 cm (100 µm)

Average Fracture Permeability = 128.28 mD ± 41.65 mD 95


Fracture Network is Dominated by Short Fractures
Fracture Aperture May Decrease with Decreasing Fracture Length
45
Regional Fractures
40 Mesaverde Sandstone
Piceance Basin, Colorado
35
Number of Fractures

30

25

20
293 Fractures
15

10

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40
Fracture Length (m)
Thus, we have a distribution of fracture apertures & therefore we will have a
distribution of kf.
Orthogonal Fracture Network in Sandstone

97
Calculated Permeability for Fracture Network
of Two Orthogonal Sets with Equally Spacing

Kf = 1.04 (d2 Фf3)

Kf = 4.16 x 10- 4 (e2 Фf)

Where fracture permeability – Kf is in Darcys


porosity – Фf is in percent
aperture – e is in microns
spacing – d is in cm

Using columns, with parallel flow to the columns, to compute kf

98
3500
Fracture Network of Two Orthogonal
3000 Fracture Sets with Equal Spacing
Fracture Permeability (md)

2500
Fracture Aperture = 0.2 mm

2000

1500

1000
0.1 mm

500
0.05 mm
0
0 0.05 0.1 0.15 0.2 0.25
Fracture Porosity (percent)
For columns, with parallel flow to the columns, if we keep the fracture
aperture constant, changing ϕf, i.e. changing the fracture spacing, the kf
curves are given by these lines.
Calculated Permeability for Fracture Network
of Three Orthogonal Sets with Equal Spacing

Kf = 0.62 (d2 Фf3)

Kf = 5.55 x 10- 4 (e2 Фf)

Where fracture permeability – Kf is in Darcys


porosity – Фf is in percent
aperture – e is in microns
spacing – d is in cm

Using cubes to compute kf


100
5000
Fracture Network of Three Orthogonal
4500 Fracture Sets with Equal Spacing
4000
Fracture Permeability (md)

3500
Fracture Aperture = 0.2 mm
3000

2500

2000

1500
0.1 mm
1000

500
0.05 mm
0
0 0.05 0.1 0.15 0.2 0.25
Fracture Porosity (percent)

For cubes, if we keep the fracture aperture constant, changing ϕf, i.e.
changing the fracture spacing, kf curves are given by these lines.
20

18
Fracture Network of Three Orthogonal
Fracture Sets with Equal Spacing
Fracture Permeability (Darcies)

16 0.125 mm

14

12

10 0.10 mm

6
0.075 mm
4

2
Fracture Aperture = 0.05 mm

0
0 0.5 1 1.5 2 2.5
Fracture Porosity (percent)

102
For higher ϕf range
Fracture Porosity & Permeability is Greatest
in Thin Chert Beds in the Monterey Formation

103
Modification of Dual Porosity Sheet Model using
Fractals to Scale Fracture Spacing & Aperture
Since it is
common to
have a fractal
distributions
of spacings &
apertures, I.e.
power laws, it
would be
better to use
these
distributions
to compute k
at different
scales, which
will help to
explain why
we have
different k
using cores & 104
well test data.
Comparison of Core to Well Test Permeabilities
in the Mesaverde Formations in Rulison Field

FLUVIAL E

FLUVIAL B
CORE (RESTORED STATE)
COASTAL YELLOW
WELL TEST
COASTAL RED
PALUDAL 3&4
PALUDAL 2

UPPER COZZETE
Anisotropic
LOWER COZZETE
CORCORAN

PERMEABILITY 0.1 1 10 100 1000 MICRODARCIES


Plot of lab-measured km & well-test permeabilities for different sandstone types.
Well-test k are one or more orders of magnitude higher than km, indicating that a
significant component of k is to the result of open fractures in the field.

Das könnte Ihnen auch gefallen