You are on page 1of 10

ELSEVIER

PETROLEUM SCIENCE & ENGINEERING


Journal of Petroleum Science and Engineering 13 ( 1995) I69- 178

Matrix-fracture transfer shape factors for dual-porosity simulators


K.T. Lim, K. Aziz *
Stunford University, Depurtment ofPetroleum Engineering, Green Earth Sciences Research bldg., Stanford, CA 943052220, Received 17 November 1994; accepted 22 November 1994 USA

Abstract
Matrix-fracture transfer shape factors for dual-porosity simulation of naturally-fracturedreservoirs were derived by combining analytical solutions of pressure diffusion for various flow geometries. The resulting equations describing the matrix-fracture flow were cast in a form similar to those proposed by Barenblatt et al. and Warren and Root , but without making the pseudosteady state assumption. Shape factors representing one, two and three parallel sets of fractures were obtained. A generalized

matrix-fracture transfer fnncticn for a non-isotropic, rectangular matrix block was also derived. Fine-grid single-porosity and one-block dual-porosity models were used to verify the results. The results of fine-grid and one-block dual-porosity models were
in good agreement. The shape factors derived in this paper are consistent with observations porosity simulation studies. made in several publications on dual-

1. Introduction
CT=

4N(N+ 2) L2

The concept of treating a naturally-fractured reservoir as a dual-porosity medium was introduced by Barenblatt et al. ( 1960) and later by Warren and Root ( 1963) more than three decades ago. In both papers the transfer per unit bulk volume between the matrix and the fracture was assumed to take place under pseudo-steady state conditions. The resulting transfer rate per unit bulk volume has the form:

(2)

(1)
where u was defined as a parameter characteristic of the fractured rock. has the dimension of reciprocal area, known as the shape factor. Warren obtained the following expression:
* Corresponding author. Warren and Root ( 1963) Barenblatt et al. ( 1960). 0920-4105/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved SSDlO920-41OS(95)00010-0

where N is the number of sets of fractures ( 1, 2 or 3). For cubic matrix blocks having a fracture spacing of L, u has the value of 12/L2, 32/L2, 60/L2 for one, two and three sets of fractures, respectively. The application of the shape factor in numerical simulation was introduced by Kazemi et al. ( 1976). Using a finite-difference formulation for the flow between the matrix and the fracture, they showed that for a threedimensional case:

representing the The parameter u and is commonly and Root (1963)

L,;

L- 7

L- z

For Lx = Ly = L, = L, u has a value of 12/L* for three sets of fractures. For one and two sets of fractures, the values of u are 4/L2 and 8/L, respectively. The shape factors proposed by Kazemi et al. are used in a number of commercially available simulators (Firoozabadi and Thomas, 1990). Ueda et al. (1989) pointed out that

170

K.T. Lirn, K. Aziz / Journal of Perroleum Science and Engineering 13 (1995) 169-178

the application of Eq. 3 is equivalent to assuming a linear pressure gradient between the center of a matrixblock and the fracture. The shape factors described in the preceding appear to be based only on the geometry of the system, and do not account for the fact that a pressure gradient exists in the matrix. Coats ( 1989) included pseudo-steady state matrix-fracture diffusion in his derivation of a matrix-fracture transfer function. He obtained shape factors exactly twice those of Kazemi et al. Thomas et al. (1983) studied various fine-grid single-porosity and single-block dual-porosity models. For a three-dimensional oil-water model, with near unit mobility ratio, they concluded that a good match between the two models was obtained with a shape factors of 251L2. Ueda et al. ( 1989) investigated the shape factors for one- and two-dimensional flows. They introduced a multiplication factor to the Kazemi et al. type shape factors. Results of dual-porosity models were compared to those obtained from fine-grid models and experiments. They concluded that Kazemis shape factors need to be adjusted by factors of at least 2 for onedimensional flow. They also concluded hueristically that the Kazemi et al. shape factor needs to be multiplied by a factor of 3 for two-dimensional flow. This paper presents the derivations of new shape factors by combining the geometrical aspects of the system with analytical solutions of the pressure diffusion equation for flow between the matrix and the fracture. It will be shown that the shape factors derived in this manner are consistent with the observations made by Thomas et al. ( 1983) and Ueda et al. ( 1989). The derivation for the case with one set of fractures is presented in the following. Derivations for systems with two and three sets of fractures follow the same principle and are presented in the Appendix.

, Matrix

x=-l

Fig. I. Schematic of a matrix block surrounded by one set of fractures.

be shown that the matrix pressure distribution satisfies the following partial differential equation (Matthews and Russell, 1967) : aP _=-- k azp at +~_Lc~ a.2 (4)

If the matrix is initially at a uniform pressurep, and the pressure at the matrix-fracture boundary is maintained constant at pn the solution of Eq. 4 can be expressed as follows:

p$?y_~
I I n=,(2n+

8
l)%?

- (2n+ exp [

1)%7%t

W,L2

1
(5)

2. Derivation In Fig. 1, a matrix block surrounded by a set of parallel fractures is considered. The fractures have infinite lateral extent such that the matrix block resembles a slab of thickness L, which is also the fracture spacing. The flow from the matrix to the fracture is one-dimensional, and perpendicular to the fracture plane. Assuming that the flow in the matrix obeys Darcys law, it can

A detailed derivation is presented in the Appendix. In a dual-porosity simulation model, all mass flow is from the matrix to the fracture, or vice versa. Therefore, the rate of matrix-fracture transfer can be related to the rate of mass accumulation in the matrix as follows:
4=

_p&&

at

Taking the partial derivative of Eq. 5 with respect to time, t, and substituting it in Eq. 6 will result in an equation for the rate of matrix-fracture transfer as a function of matrix and fracture pressures. The form of

K.T. Litn. K. Aziz /Journal of Petroleum Science and Engineering 13 (1995) 169-178

171

dix for details). equation is:

The resulting matrix-fracture

transfer

(8)
Eq. 8 has a form similar to Eq. 1, but the assumption of pseudo-steady state was not made in its derivation. Comparing the two equations, it is easy to see that the shape factor for a system with one set of fractures is: r? a=,=-L 9.87 L2 (9)

0.0 I-

I
1

2 rr2kt&c,L

Fig. 2. Analytical solution of diffusion in a plane and its approximation.

Eq. 5, however, suggests that differentiation with respect to f will not eliminate the time dependence of the function. The dependence on t in the form of an infinite summation series makes the resulting expression unsuitable for implementation in reservoir simulators. Instead, the analytical solution (Eq. 5) is approximated by a simple exponential-type function as follows:

Prn_~-()~1
Pf-Pi

exp

(_ 1
dkt
-

The derivations of shape factors for systems with two and three sets of fractures follow the same principle and are presented in the Appendix. Two separate approaches are presented. The first approach involved the assumptions that a bar-shaped matrix block formed by two sets of fractures can be represented by a cylinder, and that a cube formed by three sets of fractures can be represented by a sphere. In both cases, the volumes of the matrix blocks and the corresponcling cylinder or sphere are the same. The resulting shape factors are 18. 17/L2 for two sets of fractures and 25.67/L2 for three sets of fractures. The second approach used the Newman product of dimensionless solutions for diffusion in planes (see Appendix for details). The matrix-fracture transfer function for systems with anisotropic, rectangular matrix blocks is shown to be:

h-G2

(7) By defining an equivalent isotropic permeability kat, 1937) : k= (kJ&)

(10)
(Mus-

This approximation is obtained by taking only the first term in the infinite summation series in Eq. 5. Fig. 2 shows the plots of Eq. 5 and Eq. 7. The vertical axis is a dimensionless pressure and the horizontal axis a dimensionless time. The approximate solution is in excellent agreement with the analytical solution except for dimensionless time of less than 0.1. For typical reservoir values, this translates to a time of less than 1 second. The departure from the origin is not important as it is eliminated from the final form of the equation (see the Appendix for details). The important parameter is the coefficient oft in the exponent, which reflects the rate of transient pressure change. Differentiating Eq. 7 with respect to t and substituting into Eq. 6 enables the elimination of the time parameter (see Appen-

(11) matrix

an equivalent shape factor for an anisotropic block can be defined as:

(12)
For an isotropic (k, = kv= k,) , rectangular block, the shape factor reduces to: matrix

(13)

172

K. T. Lim, K. Aziz /Journal of Petroleum Science and Engineering I3 (I 995) 169-I 78

Table I Summary of the constant crL as defined by various authors. For this study: (a) fromEq. 13, (b) cylindrical approximation, (c) spherical approximation Sets of fractures Warren and Root (1963) 12 32 60 Kazemi et al. (1976) Coats (1989) This study

third fine-grid model is a cube of size L/2 with fractures on three sides, representing one-eight of a cube of size L surrounded by three sets of fractures. Fig. 3 and Table 2 summarize the model descriptions. As a comparison, the dual-porosity model also used shape factors of Warren and Root, Kazemi et al. and Coats. Plots of cumulative production against time for the case with one set of fractures are shown in Fig. 4. / /

1 2 3

4 8 12

8 16 24

7? 18.17b, 2+a) 25.67. 36

, /_
I

For a system with uniform fracture spacing L, the shape factors for two and three sets of fractures computed using Eq. 13 are 2$lL2 and 37?lL2, respectively. The result for one set of fractures is identical to that derived in the preceding 3. Table 1 summarizes the shape factors as proposed by Warren and Root, Kazemi et al. and Coats, and those derived in this paper. These shape factors are used in the following analyses. 3. Verification Three separate fine-grid single-porosity, and one single-block dual-porosity simulation models were used to verify the shape factors derived in this paper. The dual-porosity model has one gridblock of 10 ft on all sides, and is governed by the matrix-fracture transfer function based on the number of sets of fractures and the fracture spacing. In this case the fracture spacing, L, is also 10 ft. The single-porosity models have one discrete gridblock representing the fracture with the rest of the gridblocks representing the matrix. The singleporosity model for the case with one set of fractures is a half-plane with a thickness of L/2, with the fracture on one side. The case with two sets of fractures is represented by a prism-shaped matrix equivalent to one quarter of a cube, with the fracture on one side. The The work leading to the derivation of Eq. 13 was performed independently by the authors in 1993 and early 1994. It was brought to the authors attention, after Eq. 13 was derved, that Chang ( 1993) and Kazemi and Gilman ( 1993) have derived the same expression. Also, results consistent with Eq. 13 were obtained by Barkve ( 1992) and Zimmerman et al. ( 1993). Although the assumptions and approach as used in arriving at the shape factors are slightly different in this and other studies (Lim, 1995), it is reassuring that all the results are consistent.
y

_N=l

N=2

F\----T-----\ \ 1 \

I- x \L____L-__J
\

4;\I

I
I

I I

Fig. 3. Schematic of single-porosity models. Shaded areas represent fracture planes. Fewer than actual number of gridblocks am shown for clarity.

K.T. Lim, K. Aziz / Journal of Petroleum Science and Engineering 13 (1995) 169-178
Table 2 Data for single-porosity Fine grid, N = 1 Grid dimensions Grid spacing Fine grid, N = 2 Grid dimensions Grid spacing

173

and dual-porosity

models

11X1X1 Ax=2.445,

Ay=Az=
11x1x1

1.28, 0.64, 0.32, 0.16,0.08, loft

0.04, 0.02, 0.01, 2x0.005

ft

Ax = same as Ax for N = 1
Ay=2.445,6.17, 8.09, 9.05, 9.53.9.77, 9.89, 9.95, 9.98, 2x9.995 ft

AZ= 10 ft
Fine grid, N = 3 Grid dimensions Grid spacing Dual porosity Grid dimensions Grid spacing Common data for all models Matrix porosity Matrix permeability Initial pressure Initial temperature Fracture pressure Fluid properties are correlated from steam 6X6X6 A~=A~=Az=0.01,0.02,0.08, 1X1X1

0.32, 1.28, 3.3Oft

Ax=Ay=Az=lOft
0.05 0.001 mD 1000 psia 600F 500 psia for t > 0 table

60

this paper are for single-phase superheated steam flow. The single-phase compressed water flow cases were also analyzed, with essentially the same observations. In general, the results of dual-porosity models with shape factors derived in this paper are in good agree60

0 0 __._--_

-_--

Fine Grid, 11 blocks Fine Grid, 21 blocks 1 block DP. Waren-Root, o=12fL* 1block DP, This work, a=n2/L2 1 block DP, Coats, o&/L2 1 block DP, Kazemi et al., 0=4/L
Ill

II/

III

81

11

1,

tll
18 20
0

8 10 12 Time (day)

14

16

Fig. 4. Results for system with one set of fractures.

______-

Gridblock size sensitivity for the fine-grid model is also indicated. The results of the single-porosity models have been scaled to enable a valid comparison. The results for two and three sets of fractures are shown in Figs. 5 and 6, respectively. All the results presented in

Fine Grid, 11 blocks 1 block DP, Warren-Root, 0=32/L* 1 block DP, This work, a=2x2/L2 1 block DP. This work, o=18.17/L2 1 block DP, Coats, 0=16/L 1 block DP. Kazemi et al.. 0=8iL*
I
A

luu

4
Time (day)

10

Fig. 5. Results for system with two sets of fractures

174
60

K.T. Lirn, K. Aziz / Journal of Petroleum Science and Engineering 13 (1995) 169-I 78

, I

3D Fine 1 block 1 block 1 block 1 block 1 block

Grid, 216 blocks DP, Warren-Root, a=60/L* DP, This work, 0=3&L DP, This work. a=2567iL DP. Coats, a=24/L2 DP, Kazemi, 0=12/L*

4
Time (day)

Eq. 8 is capable of describing the transient behavior of the matrix fracture flow because the matrix pressure, Pm, and the fracture pressure, pr are both time dependent. Proper selection of time step size during numerical simulation should capture the transient behavior. The shape factors derived in this paper are consistent with the observations made by authors cited in the introduction. The factor of 25/L* used by Thomas et al. ( 1983) for the three-dimensional model is in close agreement with the factors of 25.67/L2 and 3$lL2 derived in this paper. Ueda et al. ( 1989) stated that for one-dimensional flow the Kazemi et al. shape factor (4/L2) needs to be multiplied by a factor of at least 2. This observation is consistent with the factor of d/L2 for one set of fractures. For a system with two set of fractures, Ueda et al. argued hueristically that the Kazemi et al. shape factor ( 8/L2) needs to be modified by a factor of 3. The analysis in this study showed that it should be a factor of about 2.5.

Fig. 6. Results for system with three sets of fractures.

ment with those of the single-porosity models. Calculations using Warren and Root shape factors tend to overestimate the rate of recovery. Results obtained using Kazemi et al. shape factors underestimated the recovery rate significantly.

5. Conclusions From the preceding discussions, the following conclusions can be made: ( 1) The matrix-fracture transfer shape factor for dual-porosity simulators is a parameter governed both by the geometry of the system, and the physics of mass transfer and pressure gradients in the matrix. (2) A method was presented for the derivation of the shape factors by approximating analytical solutions of pressure diffusion equations for various geometries of the system, without making the pseudo-steady state assumption. (3) The shape factors derived in this paper were verified using single-porosity single-phase flow models. These shape factors are applicable to all single-phase flow problems and two-phase flow problems with near unit mobility ratios. (4) The results obtained in this paper are consistent with observations made in the cited literatures on the need to modify the Kazemi et al. type shape factors for matching with finegrid model and experimental results.

4. Discussions

The results presented in the preceding section show that the correct shape factors are obtained by coupling the geometry and the physics of pressure diffusion from the matrix to the fracture. The shape factors are applicable over all times. The single term approximation of the analytical solution introduced a small error at early times, but the error became negligible at late times (see Fig. 2). Although the form of Eq. 8 is similar to Eq. 1, pseudo-steady assumption was not made in its derivation. The assumption of pseudo-steady state was the premise of the matrix-fracture transfer function proposed by Barenblatt et al. ( 1960) and Warren and Root ( 1963). Shape factors derived by Warren and Root (1963), Kazemi et al. (1976) and Coats (1989) all followed the pseudo-steady state assumption. As shown in the derivation presented in the preceding, the pseudo-steady state assumption is not necessary when applied to dual-porosity simulation.

6. Nomenclature C Fluid compressibility, ct Total compressibility, k Permeability, [ L2] [LT2/M] [ LT*/M]

K.T. Lim, K. Aziz/Joumal

of Petroleum Science and Engineering 13 (1995) 169-178

175

L Fracture spacing, [L J M Mass, [M] N Number of sets of fractures P Pressure, [M/LT*] 4 Mass rate per unit bulk volume, [ M/TL] r Radial dimension variable, [L] R Outer radius of cylinder or sphere [L] t Time, [T] X Linear dimension variable, [L] u Shape factor, [ 1/L2] Viscosity, [M/LT] P P Density, [ M/L31 6.1. Subscripts i f m x y z Initial condition Fracture Matrix x-coordinate direction y-coordinate direction z-coordinate direction

where the superscript O refers to a reference state. The solutions of Eq. Al for various geometries are presented as follows: 1.1.1. One set offractures Consider a matrix block surrounded by a set of parallel fractures as shown in Fig. 1. The flow from the matrix to the fracture is one dimensional and perpendicular to the fracture plane. Eq. Al for a one-dimensional, linear system is: dP -=-- k dp at &_Lc,a,? (A3)

which is the equation for diffusion in a plane. The following boundary conditions apply: PPi* PPf, p=pn -L/21x<L/2 x= -L/2, x=+L/2, t>O t>O t=O (A41 (As) (A61

Acknowledgements Funding for this work was provided by the Stanford University Reservoir Simulation Industrial Affiliates (SUPRI-B) program.

The boundary conditions represent a step change in the fracture pressure, pr, at the initial time, which is held constant. The analytical solution can be found in Crank ( 1975). If M, denotes the total amount of mass which has entered the system, i.e., the matrix, at time t and M, the corresponding quantity after infinite time, the solution can be expressed as:

Appendix 1 1.1. Derivation of matrix-fracture functions for various geometries transfer

(2n+ l)*dkt 4/&*

1
(A7)

The flow of fluids of small and constant compressibility in porous media is governed by the pressure diffusivity equation (Matthews and Russell, 1967) :

Because Eq. A7 is written on a unit volume basis, the left-hand side can be substituted by the ratio of density increment at time t compared to the initial state and the expected increment at infinite time: Mt -=M, Pm-Pi
Pf-Pi

Ipv .

at

( 1
kg
4Wl
=pu

(AlI

(A8) Eqs. A2, A7 and A8 yields: - (2nf


1)27? exp

The solution depends on the geometry and initial and boundary conditions. The assumption of a small and constant compressibility fluid implies that the density of the fluid at a given pressure can be calculated as follows:
P(P)

Combining

1)dkt

4Wtt2

(A9) where pi is the average pressure in the matrix at time

+c(P-P)l

(AZ)

176

K.T. Lbn, K. Aziz /Journal of Petroleum Science and Engineering I3 (1995) 169-l 78

t. Eq. A9 can be approximated Fig. 2) : $=l-0.81exp


1 i

by the expression

(see

dkt &-%L2

1
dkt

(AlO)

This is equivalent to taking the first term in the infinite summation series. Differentiation of Eq. A10 with respect to t yields: 1 pf-pi a& ~=~81 r?k

exp [ ~~C,L2

(All)

(Al21
0.2 )

using Eq. AlO. Simplification

of the equation yields: (Al3)


0.0 Ii] 0.0

a% -=. ,+,:tPm-,,f)
at CI
2

0.2

0.4 WwRZ

0.6

0.6

1.0

Substituting Eq. Al3 into Eq. 6 results in the following matrix-fracture transfer function: (A14)

Fig. 7. Analytical solution of diffusion in a cylinder and its approximation.

Using arguments similar to those for the case with one set of fractures, the analytical solution can be expressed as follows (Crank, 1975): (Al9)

1.1.2. Two sets offractures In this case the matrix block is surrounded by two sets of perpendicular fractures, with both sets of fractures having a spacing L. Pressure diffusion from the matrix to the fracture can be closely approximated by that of a cylindrical matrix block. The equivalent radius, R, of the cylinder is the radius that would yield the same volume as a bar with a L XL square-shaped cross-section. The volume is used as a basis of equating the two geometries due to the considerations given when deriving Eq. A8. In this case: R = 0.564L The pressure diffusion geometry is: equation (Eq. Al) (A15) for radial

where the rr,s are the roots of: J,(Ra,) =0 (A20)

where J,,(x) is the Bessel function of the first kind of order zero. The analytical solution is approximated by taking only the first summation term in Eq. A19: s=l-0.69exp

[ 1
5.78kt dwR=

(421)

The analytical solution for cylindrical diffusion and its approximations are plotted in Fig. 7. Differentiating Eq. A21 with respect to t, and substituting in Eq. 6 yields:

(416) The initial and boundary p=pi, p=pf, O<rlR r=R, t=O t>O conditions are: (AI7) (A18)

,2$~(,,

-p,)

+z+-m_pf)

(A=)

1.1.3. Three sets offractures A matrix block surrounded by three sets of fractures with uniform spacing L has the shape of a cube. The

K.T. Lirn, K. Aziz/Journal of Petroleum Science and Engineering13 (199.5)169478

111

pressure diffusion from the matrix to the fracture can be closely approximated by that of a spherical matrix block. The equivalent radius, R, of the sphere is the one that would yield the same volume as a cube with a length of L on all sides. In this case:
R = 0.62OL

(~23) (Eq. A 1) for spherical

The pressure diffusion equation geometry is:

(~24) The initial and boundary


p=pi, p=pr, O<r<R r=R, t=O t>O

conditions

are: (A25)
(A261

0.2
r o Analytical Solution Fvst Term only ___ 1

0.0 0
0

2 k&/@pqR

The analytical solution can be expressed (Crank, 1975; Zimmerman et al., 1992) :

as follows Fig. 8. Analytical solution of diffusion in a sphere and its approximation.

~427)

Pi -Pi _ The analytical solution is again approximated only the first summation term in Eq. A27: _ s=l-0.61exp
I

by taking

Pf_Pi

1-+2_y
(A281
a=op=oy=o

l
(2a+ 1)2(2p+ 1)2(2y+ 1)2

The analytical solution for diffusion in a sphere and its approximations are plotted in Fig. 8. Differentiating Eq. A28 with respect to t, and substituting in Eq. 6 yields:
4=j+Pln-Pl.)

(A30) By taking only the first terms in each of the summation series (similar to the derivation of solution for diffusion in a plane), Eq. A30 becomes:

I? pk
P

=-- L2

25.67 pk _

El.(Pm--Pf)

(~29)

1.1.4.

Anisotropic,

rectangular

matrix blocks

matrix block of dimensions directional permeabilities k,, ky and k,. If the system is initially at pressure pi and is exposed to a constant fracture pressure pf, the analytical solution can be derived by taking the Newman product of the dimensionless solution for diffusion in a plane (Eq. A7) as follows (Holman, 1990; Abbazadeh, pers. commun., 1994) :
Lx, Lx and L, with corresponding

Consider

a rectangular

(A31) Differentiation of Eq. A3 1 with respect to t yields: (~32) Substituting Eq. A32 into Eq. 6 yields the final form of the matrix-fracture transfer function:

178

K.T. Lim, K. Aziz / Journal

of Petroleum Science and Engineering 13 (1995) 169-I 78 *Conversion (A33) References factor is exact.

By defining an equivalent kat, 1937) : #c= ( k&kz) 3 and rewriting Eq. A33 as:

isotropic permeability

(MusBarenblatt, G.I., Zheltov, 1u.P. and Kochina, I.N., 1960. Basic concepts on the theory of seepage of homogeneous liquids in fissured rocks. PrikladnayaMatematikai Mekhanika, Akad. Nauk, SSSR, 24(5): 852-864. Barkve, T., 1992. Generation of pseudo data for dual-porosity simulation Unsolicited SPE Paper 22531. Chang, M., 1993. Deriving the shape factor of a fractured rock matrix. Technical Report NIPER-696 (DE930001 70), NIPER, Bartlesville Okla. Coats, K.H., 1989. Implicit compositional simulation of singleporosity and dual-porosity reservoirs. SPE 18427, presented at 10th SPE Symp. Reservoir Simulation, Houston, Tex., Feb. 6-8. Crank, J., 1975. The Mathematics of Diffusion, 2nd ed. Claredon Press, Oxford, pp. 44-103. Firoozabadi, A. and Thomas, L.K., 1990.6th SPE Comparative Solution Project: Dual porosity simulators. J. Pet. Technol., pp. 42 (6): 710-763. Holman, J.P., 1990. Heat Transfer, 7thed. McGraw-Hill, New York, N.Y., pp. 160-164. Kazemi, H. and Gilman, J.R., 1993. Flow and contaminant transport in fractured rock, Academic Press, Orlando, Fla., pp. 267-323. Kazemi, H., Merrill, L., Porterheld, K. and Zeman, P., 1976. Numerical simulation of water-oil flow in naturally fractured reservoirs. Sot. Pet. Eng. J., 16(6): 317-326. Lim, K.T., 1995. Simulation of fractured reservoir with applications to geothermal reservoir. Ph.D. dissertation, Stanford Univ., Stanford, Calif., pp. 117-l 18. Matthews, C.S. and Russell, D.G., 1967. Pressure buildup and flow tests in wells. SPE Monogr., Vol. I: 2-17. Muskat, M., 1937. Reprinted 1982. The Flow of Homogeneous Fluids Through Porous Media. IHRDC, Boston, Mass., pp. 225227. Thomas, L.K., Dixon, T.N. and Pierson, R.G., 1983. Fractured reservoir simulation. Sot. Pet. Eng. J., 23( 1): 42-54. Ueda, Y., Murata, S., Watanabe, Y. and Funatsu, K., 1989. Investigation of the shape factor used in the dual-porosity reservoir simulator. SPE 19469 presented at SPE Asia-Pacific Conf., Sydney, Australia, Sept. 13-15. Warren, J.E. and Root, P.J., 1963. The behaviorofnaturally fractured reservoirs. Sot. Pet. Eng. J.. 3(3): 245-255. Zimmerman, R.W., Chen, G. and Bodvarsson, G.S., 1992. A dualporosity model with an improved coupling term. Presented at 17th Stanford Geothermal Reservoir Engineering Workshop, Stanford, Calif., Jan. 29-3 1. Zimmerman, R.W., Chen, G., Hagdu, T. and Bodvarsson, G.S., 1993. A numerical dual-porosity model with semianlytical treatment of fracture/matrix flow. Water Resour. Res., 29( 7): 2127-2137.

(A34)

(A35) an equivalent shape factor for an anisotropic block can be inferred from Eq. A35 as: matrix

(A36) For an isotropic (k,= ky = k,), rectangular block, the shape factor reduces to: u= nJ(;+++;) matrix

(A37)

The shape factor for two and three sets of fractures calculated using Eq. A37 differs slightly from those presented earlier, due mainly to the different methods used in approximating the analytical solution. The is logical because a cube has a higher surface area to volume ratio compared to a sphere of the same volume. The approximate form as shown in Eqs. A33 and A37 was chosen to enable implementation in a system which is not necessarily cubic or isotropic.

1.2. SI metric conversion factors

ftX3.048*

E-Ol=m E - 01 = kg

lb x 4535,924

psi X 6.894,757 E + 00 = kPa mDX9.869,233 (F-32)/1.8=C E-01 =prn

You might also like