You are on page 1of 13

PAPER 2009-064

A New Model for


Reservoirs with Discrete
Fracture System
F. ZENG, G. ZHAO
University of Regina

This paper is accepted for the Proceedings of the Canadian International Petroleum Conference (CIPC) 2009, Calgary,
Alberta, Canada, 16‐18 June 2009. This paper will be considered for publication in Petroleum Society journals.
Publication rights are reserved. This is a pre‐print and subject to correction.

Abstract
fracture distribution are the main factors affecting the pressure
Dual-porosity and dual-permeability naturally fractured response. In some particular situations the fracture-dominated
reservoir models assume that the fractures in the reservoir are fluid flow region in the pressure derivative curve may present
connected with each other and distributed uniformly. However, two villages, which has been met in some field cases. The model
in some cases, the reservoir characteristics exhibits discrete provides with a tool for identifying the fracture pattern in a
fracture systems, which means that the fractures might be specific reservoir. Also, this model can be applied for
unconnected and their distribution is not uniform. In this work, optimization design of tight gas reservoir development.
a new computational model is developed to compute the
transient pressure behavior for reservoirs with discrete fracture
system. This computational model is based on Laplace Introduction
transform. The fluid flow in the fracture system and reservoir
are computed separately and flux and pressure equivalent Classically, the fractured reservoir is modeled with Dual-
conditions in Laplace space are applied in the fracture wall to porosity model(1) or dual-permeability model. Those models
couple the fluid flow in both systems. assume that the fractures are connected with each other and
distributed uniformly. The dual-porosity model also assumes
The results suggest that the pressure response in a reservoir that the fluid is produced from the fractures which are
with a discrete fracture system has three flow regions: fluid intersected with the well. However, in some cases, the reservoir
flow nearby the wellbore, fracture-dominated fluid flow and characteristics exhibit discrete fracture system, which means
fluid flow beyond the fracture. The fracture orientation (i.e. the that the fractures may be unconnected and their distribution is
distance between the fracture and the well), fracture not uniform. Such kind of reservoir system was illustrated by
parameters (fracture conductivity and non-Darcy effects) and Gao et al.(2), as shown in Figure 1. Gao et al.(2) also pointed out
that the chance for a vertical well to intersect a discrete natural
ci (x D ) =
−1 [
⎧⎪ 2 cosh (x Di − x D ) u i / Cη −( x − x )
+ e Di D
] ui / Cη
⎫⎪
fracture is extremely small, since natural fractures in a reservoir ⎨ 2( x − x ) u /C ⎬
ui ⎪⎩ e Di Di −1 i η − 1 ⎪⎭
tends to be vertical. Another type of discrete fracture system is C fD

artificial fracture system in tight gas reservoirs. To obtain
economical production rate, most of wells in a tight gas
reservoir are hydraulically fractured. Therefore, the whole Cη
reservoir looks like an artificial discrete fracture system. di = −
In this work, it is assumed that the discrete fracture system C fD u i
has the following characteristics,
1.The fractures in the reservoir are discrete and not
q fDi
connected with each other;
q *fDi =
2.Each fracture can be described with its orientation,
geometry and diffusivity;
δi
3.The well is not intersected with any fracture; and,
4.The fluid flow in fracture system obeys Forchheimer q ND
q NDi −1 =
*
equation and the fluid flow in matrix system is Darcy
flow. δ x Di −1
,
Zeng and Zhao(3) presented a model for non-Darcy flow in
hydraulic fractures. For a reservoir with only one fracture
existed, if the fracture is close enough to the well, then the q ND
q NDi
*
=
system is similar to the system with a hydraulically fracture
well. This work applied the similar algorithm for hydraulically
δ x Di
,
fractured well described by Zeng and Zhao(3) to study the
transient pressure behavior in an infinite reservoir with discrete Especially, for both ends of each fracture we have,

0 = q NDN = 0
fracture system.
q ND
* *

Model and Algorithm The pressure behavior in the matrix is a function of the flow
The comprehensive reservoir system with discrete fracture rate at the well and the flow rates at the fracture wall of each
distribution is first viewed as two sub-systems, the fracture fracture segment in each fracture, i.e.,
system and the matrix system (Figure 2a). Each fracture is
further discretized into several independent segments (Figure p D (x Di , j , y Di , j , u ) =
2b), each one with a uniform flux from reservoir to fracture, qfi Nf
(
q wD a 0 (x Di , j , y Di , j , u ) + ∑ ∑ q fDi0 , j0 aij x Di , j , y Di , j , x Di0 , j0 , y Di0 , j0 , u )
N
in m3/(s.m), and two node fluxes, qNi-1 and qNi in m3/s (Figure
i0 =1 j0 =1
2c). Then the analytical solutions in Laplace domain in matrix
system and in each fracture segment are derived, which results where x Di0 , j0 and y Di0 , j0 give the coordinates of source at
in a linear-format solution with variables of qfi, qNi-1 and qNi in
Laplace domain. After that, the pressure equivalent and flux the segment j0 of fracture i0. x Di , j and y Di , j are the
continuity conditions in Laplace domain are applied at each
coordinates of the location we are interested in its pressure, and,
interface between the fracture system and matrix system and
between any two adjacent fracture segments to combine these
solutions into a linear equation system. Finally, fluxes at the
interfaces are obtained by solving the linear equation system,
a 0 (x Di , j , y Di , j , u ) = K 0 ⎛⎜

(x Di , j
2
+ y Di , j u ⎞⎟
2


)
and the bottom-hole pressure can be computed.
For each segment of each fracture, we view that the segment
has two given flux boundary condition (as shown in Figure 2c). (
aij x Di , j , y Di , j , x Di0 , j0 , y Di0 , j0 , u = ∫ K 0 d u dl ) L
( )
Specifically, since the boundary conditions on the both ends of
each fracture in the discrete fracture system should be closed L is the linear segment starting from x Di , j −1 , y Di , j −1 and ( )
( )
boundary condition, the flux on the both ends of each fracture 0 0 0 0

should be zero. Appendix A derives the solution for each ending at xDi , j , yDi , j , d is the distance between any point at
this linear segment and (x , y Di , j ) .
0 0 0 0

fracture segment. The analytical solution in each segment is,


Di , j

Applying the coupling conditions(4), we can get a linear


p fDi ( x D , u i ) = bi ( x D )q NDi
*
−1 + ci ( x D )q NDi + d i q fDi
* *
, equation system. The coefficient matrix is illustrated in Figure
3. Specific algorithm for sparse matrix may accelerate solving
where the linear equation system; however, in this study LU
decomposition method is employed. Solving the linear equation
system yields the fluxes qfD and qND in Laplace domain. Finally
bi ( x D ) =
1 [ ]
⎧⎪ 2 cosh ( x D − x Di −1 ) ui / Cη −( x − x
+ e D Di −1
) ui / Cη
⎫⎪ the well bottom-hole pressure in Laplace domain can be
⎨ 2( x − x ) u /C ⎬ computed with matrix pressure equation.
ui ⎪⎩ e Di Di −1 i η − 1 ⎪⎭
C fD

2
Results for a System with Only One Effects of the distance between the well
Fracture and fracture
The real system with only one fracture is hard to meet in The effect of the distance between the well and fracture, d, is
realistic field development. However, if a test is run in a vertical presented in Figure 7. The value of d is taken as 0.01Lf, 0.05Lf,
well interfered by a neighboring hydraulically fractured well, 0.1Lf, 0.2Lf and 1.0Lf, respectively. The pressure response and
the recorded pressure behavior from this test may behave like its derivative are sensitive to the value of d. A smaller value of
the results discussed here. In this discussion, the effects of the d leads to lower pressure drops and an earlier and lower dip in
Forchheimer number in the fracture and the distance between the derivative curve. A larger value of d has weaker effects on
the well and fracture on the transient pressure behavior and flux the pressure responses. When d = 1.0Lf, the effect of the fracture
distribution are presented. The parameters for basis case are CfD on the pressure response is too slight to be observed.
= 20, Cη = 1.0×105. Figure 8 gives the flux distribution profile along the fracture
under different values of d. The location of the fracture does
significantly affect the flux distribution. The closer the fracture
Effects of Forchheimer number in the and the well are, the more fluid flowing from the reservoir to
fracture the well by passing the segments which have the shorter
distance with the well. The larger the value of d, the more
Figure 4 shows the effects of Forchheimer number in the
uniforms the flux distribution. When d = 1Lf, the fluid flows to
fracture on the transient pressure type curve. Three flow
the well through the fracture becomes very small.
regions, fluid flow nearby the wellbore, fracture-dominated
The approximate linear relationship between the ratio of the
fluid flow and fluid flow beyond the fracture, are observed. The
fluid flows from the reservoir to the well by passing the fracture
first regions of these four sets of type curves in this figure have
and the log value of d is further presented in Figure 9, which
the exactly same pressure responses. During this period the
suggests the fluid ratio declines very fast when the value of d
pressure propagation has not approached to the fracture.
increases. When d = 0.05Lf, more than 50% fluid directly flows
Therefore, the pressure response in this region is not affected by
to the well rather than through the fracture; while when d =
the fracture, but formation characteristics around the wellbore
1.0Lf, only 3.4% fluid flows through the fracture.
area. In the second region, the existed fracture causes a
The results discussed are applicable in tight gas reservoir
downward dip in the pressure derivative curve. Similar to the
development, where most of wells are hydraulically fractured to
observation in hydraulically fractured wells, the larger the
obtain economic production rate. Besides the hydraulic
Forchheimer number, the larger the pressure drop and the lower
fracturing design, another important factor affecting the
the dip in the derivative curve. Until tD = 100, the effect of the
economic benefit of tight gas reservoir development is well
fracture on the pressure derivative disappears, which means the
spacing. In this discussion, it is found that the well spacing and
formation characteristics beyond the fracture dominate the
fracture length are correlated, which means the well spacing
pressure responses in the third region.
should be a function of fracture length. Therefore, it is
Figure 5 further discusses the effect of non-Darcy flow on
important that the hydraulic fracturing design should be made in
the flux distribution along the fracture. A positive value of the
the initial reservoir development planning along with the well
flux qfD means the fluid flowing from the reservoir to fracture;
spacing design to obtain optimal depletion strategy, instead of
and a negative value suggests the fluid flowing from the
making such design after wells have been drilled.
fracture to the reservoir. As shown in this figure, the fluid flows
from the reservoir to the fracture at the two wings of the
fracture, then, flows from the fracture to the reservoir in the
middle segments of the fracture, where the shortest distance
Results for a System with Multiple
from the well to the fracture is, and finally flows from the Fractures
reservoir to the vertical well. The larger the value of
The transient pressure behavior in a system with multi-
Forchheimer number in the fracture, FND,F, more uniform the
fractures depends on the fracture distribution and the values of
flux distribution and the shorter the fracture segments with a
the fracture parameters. In this discussion, it is assumed that all
negative flux.
fractures in a system have the same parameters including
Different from hydraulically fractured wells, where almost
fracture length and CfD, Cη and FND,F values. However, this
100% of fluid flows from the reservoir to the fracture first, and
model is capable of dealing with fractures with different
then flows to the wellbore through the fracture, in a discrete
fracture length and CfD, Cη and FND,F values.
fracture system, one part of fluid flows to the well by passing
the fracture, and the other part of fluid directly flows from the
reservoir to the well without passing the fracture. Figure 6 gives Effect of fracture distribution
the relationship between the ratio of the fluid flowing to the
wellbore by passing the fracture and FND,F, and presents that the Figures 10 and 11 represent type curves for two different
more severe non-Darcy flow, the less the fluid flowing to the fracture distribution scenarios. The fracture distribution scenario
wellbore by passing the fracture. In the particular cases, when is shown in the same figure, accordingly. For each scenario, the
the value of FND,F is up to 1000.0, the ratio is 61.8%, while for type curves for different fracture number are discussed. For
Darcy flow (FND,F = 0.0) the ratio is 81.6%; the difference is example, in Figure 10, the pressure and its derivative for the
19.8%. cases which consider fracture 1, fractures 1 through 2, fractures
1 through 3, fractures 1 through 4, fractures 1 through 6,
fractures 1 through 8, and all fractures are represented,
respectively. Generally, the more the fractures, the larger the
pressure drops. On derivative curves, downwards dips are
observed. When only fractures 1 and 2 exist in the reservoir,
one dip is observed. When more fractures are added at the

3
locations far away from the well, the second dip appears and the fracturing design should be made in the initial reservoir
more fractures, the deeper the second downwards dip. However, development planning along with the well spacing design to
in this scenario, more fractures (for example over 3 fractures) obtain optimal depletion strategy, instead of making such
do not lead to more downwards dips necessarily. design after wells have been drilled.
The results for the other scenario are discussed in Figure 11.
For this scenario, only one dip is observed on the derivative
curves. The increase of the fracture number does not lead to Acknowledgements
more downward dips. More fractures just cause smaller pressure
derivative values at the downwards dip. Adding fractures 7 and The financial support from Natural Sciences and
8 into the system does not affect the pressure response too Engineering Research Council (NSERC) is highly appreciated.
much, which is similar in scenario 1, where adding fractures 9
and 10 has little effect on the pressure behavior. Checking the
location of those fractures, we can find out that the fractures 9 NOMENCLATURE
and 10 in scenario 1 and fractures 7 and 8 in scenario 2 are B = formation volume factor, [m3/ m3]
almost located beyond 2 times the fracture length far away from Ct = total compressibility, [pa-1]
the well. CD = dimensionless wellbore storage, dimensionless
The above results suggest that the fractures locating beyond CfD = fracture conductivity, dimensionless
2Lf far away from the wellbore are almost not helpful to Cη = fracture diffusivity, dimensionless
provide better conductive conduit to the well production. FND,F = Forchheimer number for non-Darcy flow in
Applying this observation in a tight or low-permeability gas fracture, dimensionless
reservoir, the similar conclusion can be made to the observation h = net-pay thickness, [m]
in Section 2, which is the well space and fracturing stimulation I0 = Modified Bessel function of zero order
should be optimally designed to maximize the production from k = reservoir permeability, [m2]
the whole reservoir. kf = fracture permeability, [m2]
K0 = Modified Bessel function of zero order
Effect of fracture parameters M = molecular mass, [g/mol]
p = pressure, [pa]
The effects of fracture parameters CfD, Cη, FND,F on transient pi = initial reservoir pressure, [pa]
pressure behavior are presented in Figure 12. The fracture qf = flux from reservoir to fracture along fracture,
scenario used here is scenario 1 with 10 fractures shown in the [m3/(s.m)]
figure 10. Type curves under four sets of parameter qN = flux at the node, [m3/s]
combinations are presented, curve 1 with CfD = 20, Cη = Qsc = surface flowrate, [m3/s]
1.0×103, FND,F = 0, curve 2 with CfD = 20, Cη = 1.0×105, FND,F = t = time, [s]
0, curve 3 with CfD = 2, Cη = 1.0×105, FND,F = 0, and curve 4 wf = fracture width, [m]
with CfD = 20, Cη = 1.0×105, FND,F = 100. Comparison of curves xf = half fracture length, [m]
1 and 2 suggests that the fracture diffusivity, Cη, has slight u = Laplace variable
effect on the transient pressure behavior, while the effect of v = velocity, [m/s]
fracture conductivity, CfD, or non-Darcy flow, FND,F, on the type β = inertial factor, [1/m]
curve is significant through comparing the curves 2 and 3, or μ = viscosity, [pa.s]
the curves 2 and 4. A lower CfD value or a larger FND,F value ρ = density, [kg/m3]
leads to larger pressure drops and more shallow pressure φ = porosity, fraction
derivative dips. Subscripts
D = dimensionless
i = node number or segment number
Conclusion
In this chapter, the results for non-Darcy flow in a reservoir
with discrete fracture system are discussed. The results can be REFERENCES
practically applied to tight gas reservoir development and some 1. WARREN, J.E. and ROOT, P.J., The Behavior of
naturally fractured reservoir with sparse and unconnected Naturally Fractured Reservoir; SPE Journal, September
fractures. The following observations are made, 1968.
1. The pressure responses in a reservoir with discrete fracture 2. GAO, G., EVANS, R.D. and CHANG, M., Pressure-
system have three flow regions, fluid flow nearby the Transient Behavoir for a Horizontal Well Intersecting
wellbore, fracture-dominated fluid flow, and fluid flow Multiple Random Discrete Fractures; prepared for
beyond the fracture. presented at the SPE 6th Annual Technical Conference
2. The pressure responses in a reservoir with discrete fracture and Exhibition held in New Orleans, LA, U.S.A., 25-28
system are significantly affected by the fracture orientation September 1994.
(i.e. the distance between the fracture and the well), fracture 3. ZENG, F. and ZHAO, G., The Optimal Hydraulic
parameters (CfD and FND,F) and fracture distribution. Fracture Geometry Under Non-Darcy Flow Effects;
3. The ratio of fluid flowing through the fracture to the paper SPE114285 presented at the CIPC/SPE Gas
wellbore is a function of the fracture orientation and the Technology Symposium 2008 Joint Conference held in
fracture parameters. Calgary, Alberta, Canada, 16–19 June 2008.
4. The observations suggest that in tight gas reservoir 4. ZENG, F. and ZHAO, G., Semianalytical Model for
development the well spacing and fracture length are Reservoirs With Forchheimer's Non-Darcy Flow; SPE
correlated. Therefore, it is important that the hydraulic Reservoir Evaluation and Engineering, April 2008.

4
Appendices Finally, the solution can be written as

The mathematical model in segment i ( x Di −1 ≤ x D ≤ x Di ) is p fDi ( xD , ui ) = q NDi


*
−1 • Ps ( x D , x Di −1 , ui ) − q NDi • Ps ( x D , x Di , ui )
*
..............(A.7)

− q *fDi
∂ 2 p fD q fDi (t D ) 1 ∂p fD C fD ui
− =
∂x D2 δ i C fD Cη δ i ∂t D
, .................................................(A.1)
For no-flow boundary segments, 0 ≤ x′D ≤ Δx D , with a
with initial condition, source at x ′D 0 , we have (van Kruysdijk 1988)
p fD ( x D , t D = 0 ) = 0
......................................................................(A.2)
Ps ( x ′D , x ′D 0 , u i ) =
1 [
⎧⎪ 2 cosh ( x ′D − x ′D 0 ) u i / Cη ]
and boundary conditions at xDi-1 and xDi, 2
ui

⎪⎩ (
exp 2Δx D u i / Cη )

∂p fD 1 ⎛ q ND ⎞
[
+ exp − x ′D − x ′D 0 u i / Cη ]
∂x D
= ⎜ ⎟
C fD ⎝ δ ⎠ x +
[
2 cosh ( x ′D + x ′D 0 ) u i / Cη ]
xDi −1 Di −1
, (
exp 2Δx D u i / Cη )
[
+ exp − x ′D + x ′D 0 u i / Cη ]} .............. (A.8)
∂p fD 1 ⎛ q ND ⎞
=− ⎜ ⎟
∂x D x Di
C fD ⎝ δ ⎠ x The solution for the no-flow boundary segment
Di
..........................................................(A.3) xDi −1 ≤ xD ≤ xDi with a source located at x D 0 = x Di −1 can be
obtained through Eq. (A.8). We let
Letting
τ Di = ∫ δ i (ς )dς
tD

0
x ′D = x D − x Di −1 , x ′D 0 = x D 0 − x Di −1 = x Di −1 − x Di −1 .(A.9)
,
⎛q ⎞
q *NDi −1 = ⎜ ND ⎟ Substituting Eq. (A.9) into Eq. (A.8) yields
⎝ δ ⎠ xDi −1 ,
Ps ( x D , x Di −1 , u i ) = ...........(A.10)
q * ⎛q ⎞
= ⎜ ND ⎟ 1 [
⎧⎪ 2 cosh ( x D − x Di −1 ) u i / Cη ]
+ exp − x D − x Di −1 u i / Cη [ ]
NDi
⎝ δ ⎠ xDi , ui

(
⎪⎩ exp 2( x Di − x Di −1 ) u i / Cη )

and doing Laplace transform yields
Similarly, the solution for the same segment with a source
∂ 2 p fD q *fDi 1 located at x D 0 = x Di can be obtained by letting
− = u i p fD
∂x 2
D C fD Cη x ≤ x D ≤ x Di
, Di −1 ................(A.4)
x ′D = x D − x Di , x ′D 0 = x D 0 − x Di = x Di − x Di ....(A.11)

with left and right boundary conditions, Substituting Eq. (A.11) into Eq. (A.8) yields

Ps (xD , xDi , ui ) =
∂p fD
=
1 *
q NDi −1 1 [
⎧⎪ 2 cosh ( xDi − xD ) ui / Cη
+ exp − xDi − xD ui / Cη
] [ ]
.....(A.12)

∂x D x Di −1
C fD ui

⎪⎩ exp 2ΔxD ui / Cη ( )
, Cη

Thus, the solution to the mathematical model is


∂p fD 1 *
=− q NDi p fDi ( x D , u i )
∂x D xDi
C fD
= bi ( x D )q NDi −1 + c i ( x D )q NDi + d i q fDi
............................................................(A.5) , ...................(A.13)
* *

The solution to the above system can be obtained with


source function, Where

∂p fD
p fDi (x D , x D 0 , u i ) = • Ps (x D , x D 0 , u i )
∂x D x D = xD 0
.........................(A.6)

5
bi ( x D ) =
1 [
⎧⎪ 2 cosh ( x D − x Di −1 ) u i / Cη ] −( x − x
+ e D Di −1
) ui / Cη
⎫⎪
⎨ 2 ( x Di − xDi −1 ) ui / Cη ⎬
ui ⎪⎩ e −1 ⎪⎭
C fD

ci ( x D ) =
−1 [
⎧⎪ 2 cosh (x Di − x D ) u i / Cη ] −( x − x )
+ e Di D
ui / Cη
⎫⎪
⎨ 2 ( x Di − x Di −1 ) ui / Cη

ui ⎪⎩ e −1 ⎪⎭
C fD


di = −
C fD u i

6
Figure 1 A discrete fracture system generated by Monte Carlo Simulation (Courtesy of Gao et al., 1994)

7
Fracture Nf

Fracture 2
well
Fracture 1

Fracture 3

(a) Discrete fracture system

0 1 2 3 … N-1 N
1 2 3 … N-1 N

(b) Discretizing each fracture into segments

q NDi , j −1 i, j q NDi , j

qfDi,j

(c) flux at segment j of fracture i

Figure 2 Discretizing the comprehensive discrete fracture system

8
Figure 3 Coefficient matrix structure for modeling non-Darcy flow in a discrete fracture system with nf = 3

10
CfD = 20
Cη = 1e5
dfw = 0.01Lf
1, 2, 3, 4

1
pwD, dpwD/dlntD

4
3
0.1 2 1, FND,F = 0
1
2, FND,F = 100
3, FND,F = 500
Lf
4, FND,F = 1000
dfw

0.01
1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 1e+1 1e+2
tD

Figure 4 Effect of FND,F on the pressure response in the system with single fracture

9
0.1

1, 2, 3, 4 1, 2, 3, 4

0.0
1, 2, 3, 4 1, 2, 3, 4

-0.1
CfD = 20
qfD

Cη = 1e5
-0.2 dfw = 0.01Lf

1, FND,F = 0
2, FND,F = 100
-0.3 Lf 3, FND,F = 500
4, FND,F = 1000
dfw
1, 2, 3, 4
-0.4
0.0 0.2 0.4 0.6 0.8 1.0
xfD

Figure 5 Effect of FND,F on flux distribution along the fracture in the system with single fracture

1.0
CfD = 20
Cη = 1e5
ratio of fluid flowing through the fracture

dfw = 0.01Lf
0.9

0.8

0.7

0.6 Lf

dfw

0.5
0 200 400 600 800 1000
FND,F

Figure 6 Effect of FND,F on ratio of fluid flowing through the fracture in the system with single fracture

10
10
CfD = 20
Cη = 1e5
FND,F = 500
1, 2, 3, 4, 5

1, dfw = 0.01Lf

pwD, dpwD/dlntD
2, dfw = 0.05Lf
3, dfw = 0.1Lf
4, dfw = 0.2Lf
1
5, dfw = 1.0Lf

5
4
3
2
1
Lf

dfw

0.1
1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 1e+1 1e+2
tD

Figure 7 Effect of the distance between the fracture and the well, dfw, on the pressure response in the system with single fracture

0.1
1, 2, 3, 4, 5
1, 2, 3, 4, 5

0.0

CfD = 20
-0.1 Cη = 1e5
qfD

FND,F = 500

1, dfw = 0.01Lf
-0.2 2, dfw = 0.05Lf
3, dfw = 0.1Lf
4, dfw = 0.2Lf
5, dfw = 1.0Lf
-0.3
Lf

dfw
1, 2, 3, 4, 5
-0.4
0.0 0.2 0.4 0.6 0.8 1.0
xfD

Figure 8 Effect of the distance between the fracture and the well, dfw, on flux distribution along the fracture in the system with single
fracture

11
1.0
CfD = 20
ratio of fluid flowing through the fracture Cη = 1e5
FND,F = 500
0.8

0.6

0.4

0.2 Lf

dfw

0.0
0.001 0.01 0.1 1 10
dfw / Lf

Figure 9 Effect of the distance between the fracture and the well, dfw, on ratio of fluid flowing through the fracture in the system with
single fracture

100 8
3
1, fracture 1
2
8
7
2, fractures 1~2
1
5 3, fractures 1~3
3
1
0
4, fractures 1~4
-3 -2 -12 0 1 2 3 5, fractures 1~6
10 4
-1
6 6, fractures 1~8 6
9
-2 7, fractures 1~10 1
10
-3
dpwD / dlntD

pwD

1 7 4
dpwD/dlntD pwD 1

7
0.1 2

CfD = 20, Cη = 105, FND,F = 0


0.01 0
1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 1e+1 1e+2
tD

Figure 10 Effect of the fracture number on the transient pressure behavior (pattern a)
12
1000 10
2.5
2
7
1.5 CfD=20, Cη=105, FND,F=0
100 1
5
3 0.5 8
0
1
-1.5 -0.5-0.5 20.5 1.5
4
-1
10 -1.5 6
-2
6
dpwD / dlntD
8 -2.5

pwD
1 dpwD/dlntD

4
0.1

fractures 1~2
pwD 2
0.01 fractures 1~4
fractures 1~6
fractures 1~8
0.001 0
1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 1e+1 1e+2
tD
Figure 11 Effect of the fracture number on the transient pressure behavior (pattern b)

10 7
3

2
8
7
1
5 6
3
1
0
-3 -2 -12 0 1 2 3
4
-1
6
9 5
-2
1 10
-3
1, 2, 3, 4
dpwD / dlntD

4
pwD

0.1
1, 2, 3, 4 2
1, CfD = 20, Cη = 1e3, FND,F = 0
2, 1, 3, 4
2, CfD = 20, Cη = 1e5, FND,F = 0
1
3, CfD = 2, Cη = 1e5, FND,F = 0
4, CfD = 20, Cη = 1e5, FND,F = 100
0.01 0
1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 1e+1 1e+2
tD

Figure 12 Effects of CfD, Cη, FND,F on the transient pressure behavior in pattern a.

13

You might also like