Você está na página 1de 20

J194501 DOI: 10.

2118/194501-PA Date: 12-March-19 Stage: Page: 1 Total Pages: 20

A Semianalytical Model for Predicting


Transient Pressure Behavior of a
Hydraulically Fractured Horizontal Well
in a Naturally Fractured Reservoir With
Non-Darcy Flow and Stress-Sensitive
Permeability Effects
Liwu Jiang, University of Regina; Tongjing Liu, China University of Petroleum, Beijing; and
Daoyong Yang*, University of Regina

Summary
Non-Darcy flow and the stress-sensitivity effect are two fundamental issues that have been widely investigated in transient pressure
analysis for fractured wells. The main object of this work is to establish a semianalytical solution to quantify the combined effects of
non-Darcy flow and stress sensitivity on the transient pressure behavior for a fractured horizontal well in a naturally fractured reservoir.
More specifically, the Barree-Conway model is used to quantify the non-Darcy flow behavior in the hydraulic fractures (HFs), while the
permeability modulus is incorporated into mathematical models to take into account the stress-sensitivity effect. In this way, the result-
ing nonlinearity of the mathematical models is weakened by using Pedrosa’s transform formulation. Then a semianalytical method is
applied to solve the coupled nonlinear mathematical models by discretizing each HF into small segments. Furthermore, the pressure
response and its corresponding derivative type curve are generated to examine the combined effects of non-Darcy flow and stress sensi-
tivity. In particular, stress sensitivity in HF and natural-fracture (NF) subsystems can be respectively analyzed, while the assumption of
an equal stress-sensitivity coefficient in the two subsystems is no longer required. It is found that non-Darcy flow mainly affects the
early stage bilinear and linear flow regime, leading to an increase in pressure drop and pressure derivative. The stress-sensitivity effect
is found to play a significant role in those flow regimes beyond the compound-linear flow regime. The existence of non-Darcy flow
makes the effect of stress sensitivity more remarkable, especially for the low-conductivity cases, while the stress sensitivity in fractures
has a negligible influence on the early time period, which is dominated by non-Darcy flow behavior. Other parameters such as storage
ratio and crossflow coefficient are also analyzed and discussed. It is found from field applications that the coupled model tends to obtain
the most-reasonable matching results, and for that model there is an excellent agreement between the measured and simulated
pressure response.

Introduction
Because new discoveries of conventional reservoirs have been declining steadily over the last few decades, rising energy demand has
imposed increasing interest in exploiting unconventional resources, such as tight oil and shale gas. Horizontal drilling with multistage-
fracturing technology has been shown to be the most effective way to increase well productivity for tight oil and shale gas formations.
In practice, the designed well does not typically meet expected well performance because of the composite contribution of non-Darcy
flow, multiphase flow, and stress-sensitivity effect (Vincent et al. 1999; Wang et al. 2015; Wang et al. 2017b; Zhang and Yang 2017;
Diwu et al. 2018). Numerous efforts have been made to prove that non-Darcy flow or stress-sensitivity effect plays an important role in
the well-performance prediction and HF design/optimization, respectively (Luo and Tang 2015; Gu et al. 2017; Li et al. 2017; Wang
et al. 2017a; Zhang and Yang 2017). So far, few attempts have been made to quantify the combined effects of non-Darcy flow and
stress sensitivity on the transient pressure behavior of a horizontal well with multistage fractures. Physically, the stress-sensitivity effect
mainly results from flow channel deformation in the matrix/NF and proppant crushing and embedment in proppant-filled HFs (Wang
et al. 2017b). Depletion-induced pore-pressure change results in effective-stress change, which will in turn change fracture and matrix
permeability and alter the flow behavior of the fluid within it (Clarkson et al. 2013; Moradi et al. 2017). The existence of the NF net-
work makes the aforementioned phenomena even more complicated. Therefore, it is of practical and fundamental importance to accu-
rately examine the combined effects of non-Darcy flow and stress sensitivity on the performance of a multifractured horizontal well
(MFHW) in a naturally fractured reservoir (NFR).
Because the single-porosity model is no longer applicable in an NFR, numerous efforts have instead been directed at using dual-
porosity models for describing flow behavior, assuming that the two subsystems (i.e., matrix subsystem and NF subsystem) function as
the main storage space and the main flow path, respectively (Warren and Root 1963; Kazemi 1969; de Swaan 1976; Jalali and Ershaghi
1987). The interaction between these two subsystems can be quantified by the exchange flow rate (Hassanzadeh et al. 2009); however,
transient pressure analysis for an NFR is commonly performed with the assumption that permeability for NFs remains constant, which
might not be physically applicable for stress-sensitive reservoirs. The dual-porosity model proposed by Warren and Root (1963) might
not be appropriate for the interpretation of transient pressure in the NFRs because it cannot be used to characterize the complexity of
the NFRs, multiple-size matrix blocks, and discrete NFs, and the fluid transport between the equivalent matrix subsystem and equiva-
lent fracture subsystem is not the actual flow condition in the matrix subsystem (Kuchuk and Biryukov 2012; Kuchuk et al. 2015).

* Corresponding author; email: tony.yang@uregina.ca


Copyright V
C 2019 Society of Petroleum Engineers

Original SPE manuscript received for review 16 July 2018. Revised manuscript received for review 2 November 2018. Paper (SPE 194501) peer approved 6 November 2018.

2019 SPE Journal 1

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 2 Total Pages: 20

A semianalytical model that is based on the boundary-element method was used to obtain transient pressure in NFRs by accounting for
the complex NF geometry and parameters. The discussion of complex fracture network in NFRs is outside the scope of this work, and
the Warren-Root model is assumed to satisfy the characteristics of NFRs in this work.
Numerous efforts have been made to examine the stress-sensitivity effect on transient pressure responses, while theoretical models
have been mainly developed for wells in single-porosity or dual-porosity reservoirs and fractured wells in single-porosity reservoirs
(Pedrosa 1986; Chen et al. 2008; Liu et al. 2015; Li et al. 2017; Wang et al. 2017a, b). Few attempts have been made to evaluate tran-
sient pressure responses of an MFHW in an NFR (Ren et al. 2017). Because the presence of stress-dependent permeability makes the
flow equations strongly nonlinear, it is difficult to solve the equations analytically. Pedrosa (1986) adopted the perturbation technique
for the first time to obtain the transient pressure solution for the flow equations with strong nonlinearity caused by stress-sensitive per-
meability. In practice, accurate results can be obtained through the perturbation technique with zero-order approximation; however, this
is only valid when assuming that the stress-sensitive coefficient (i.e., permeability modulus) is smaller than a threshold value (Wang
et al. 2015; Chen et al. 2016a). To obtain the analytical solution successfully, previous studies commonly assume that the permeability
modulus for the different subsystems is the same (Liu et al. 2015; Chen et al. 2016b, 2017). Obviously, this is not physically reasonable
because of the different characteristics for different subsystems. In fact, HFs have a broader range of permeability-modulus values,
depending on the type of proppants and their properties (Berumen and Tiab 1997; Wang et al. 2017b).
In addition to the stress-sensitivity effect, non-Darcy flow within HFs may impose a significant effect on the transient pressure
response. The most widely used model to describe non-Darcy flow in the porous media was originally proposed by Forchheimer
(1901), who added an inertial term into Darcy’s equation to account for the effects resulting from high-velocity flow. Subsequently,
numerous studies were made to describe the non-Darcy flow behavior on the basis of the Forchheimer equation (Guppy et al. 1982;
Hill et al. 2001; Luo and Tang 2015). By replacing the inertial term in the Forchheimer equation with two new parameters (i.e., mini-
mum permeability plateau and characteristic length), the Barree-Conway model is capable of handling non-Darcy flow in a wider
velocity range than the Forchheimer model (Barree and Conway 2004; Lai et al. 2012; Zhang and Yang 2014b, 2017). Therefore, it is
essential that the versatile Barree-Conway model be applied to describe non-Darcy flow behavior in hydraulically fractured wells in
an NFR.
In this paper, a coupled model has been developed to examine the combined effects of non-Darcy flow and stress sensitivity on the
transient pressure response of an MFHW in an NFR. More specifically, a slab source function was developed in the Laplace domain to
describe the fluid flow from the NF subsystem to the HF subsystem, while a new solution was derived to describe the fluid flow in the
HFs with consideration of the combined effects of non-Darcy flow and stress sensitivity. The stress-sensitivity coefficient for the HF
subsystem was different from that of the NF subsystem, which is also different from the traditional assumptions that the two coefficients
should be equal for the convenience of solving the equation matrix (Liu et al. 2015; Chen et al. 2016b). A perturbation method is used
to eliminate the nonlinearity caused by the stress sensitivity of the NFR, while for the HFs, a hybrid semianalytical and iteration
method is introduced to deal with the nonlinear problem caused by the combined effects of non-Darcy flow and stress sensitivity. To
verify the model, numerical validation was conducted under simplified conditions with the help of a commercial simulator. Effects of
various parameters on the transient pressure responses of such systems have been examined, analyzed, and discussed, while the newly
developed model has been extended to a field case.

Theoretical Formulations
In this study, a hydraulically fractured horizontal well is considered in a box-shaped NFR (Fig. 1), with the main assumptions being:
• The reservoir has a uniform initial pressure and constant thickness with no-flow outer boundaries.
• The equally spaced HFs fully penetrate a box-shaped NFR, and the multiple fractures are identical with a finite conductivity.
• Stress-sensitivity effects in the NF and HF subsystems are all considered, and the stress-sensitivity coefficients for the two fracture
subsystems are different.
• The Warren-Root model is assumed to characterize the dual-porosity media; the NFR is transformed into an equivalent homoge-
neous NF medium and an equivalent homogeneous matrix medium (Kuchuk et al. 2015), while pseudosteady crossflow occurs
between the matrix and NFs; and flow in the NFs can be described by using Darcy’s law.
• Non-Darcy flow is assumed to be in the HFs, and incompressible flow is assumed in the HF because the fracture volume is negligi-
ble compared with that of the NF subsystem (Cinco-Ley and Meng 1988).
• HFs are the only pathway for the fluid flow from the NFs to the wellbore, and the flow at the two toe points of the HF
is negligible.
Linear flow and radial flow could occur simultaneously in a transverse fracture intersected by a horizontal well (see Fig. 1c), which
is different from that in the vertical fracture. In transverse fractures, the convergence of the fluid flow into a horizontal wellbore results
in an additional pressure drop (Wang and Jia 2014). The fluid flow pattern in the transverse fracture can also be assumed as linear flow,
while the effect of flow convergence near the wellbore can be treated as a type of skin factor—namely, the convergence-flow skin
(Mukherjee and Economides 1991). In this work, the convergence-skin effect can be largely ignored by assuming a relatively small
formation thickness.
With the aforementioned assumptions, the matrix/fracture flow is schematically described in Fig. 2a. The fluid is released from the
matrix into the NFs and then flows into the HFs. The flow from NFs to HFs is 2D in the x–y plane, while only 1D flow occurs in the
HFs, because the radial flow from the HF to the wellbore is ignored.
In addition, the widely applied model to describe permeability vs. pressure is expressed in terms of the permeability modulus c
(Pedrosa 1986), which is defined as

1 dk
c¼ : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð1Þ
k dp

The definition of c assumes the exponential relationship between NF permeability and NF pressure, which can be written as follows
(Yang et al. 1999; Chen et al. 2008)

knf ðpnf Þ ¼ knfi exp½cnf ðpi  pnf Þ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð2Þ

where knf is the permeability for the NF subsystem, pnf is the pressure in the NF subsystem, pi is the initial pressure, knfi is the initial
permeability of NFs at initial pressure, and cnf is the permeability modulus for the NF subsystem.

2 2019 SPE Journal

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 3 Total Pages: 20

ye
Natural fractures
xe
z
x

hhf y
Horizontal
whf wellbore
h Hydraulic Hydraulic
fractures fractures
2xhf

(a) (b)

z x

(c)

Fig. 1—(a) 3D view and (b) top view of a hydraulically fractured horizontal well in a box-shaped naturally fractured reservoir, and
(c) convergence flow effect around the wellbore in an HF.

Natural fracture

Matrix Matrix

qex

qfDi,j
Wellbore
j

Hydraulic fracture qNDi,j qNDi,j +1

(a)
xhfD

qfDi,j qfDi,j +1

qNDi,j qNDi,j +1

M j +1 j j j +1 M x

xwDi,0 xwDi,j ΔxhfD


(b)

Fig. 2—Schematic of (a) fluid flow paths for an MFHW in an NFR and (b) the coupled matrix/fracture flow.

In HFs, the fracture permeability will remain constant after a specific value of effective stress on fractures even though the stress
continues to increase (Abass et al. 2009; Weaver et al. 2010). Therefore, a minimum fracture permeability khf min is introduced to char-
acterize the lower limit of stress-sensitive permeability, and the apparent HF permeability caused by the stress-sensitivity effect can be
described by a general model proposed by Zhang et al. (2014):

khf ðphf Þ ¼ khf min þ ðkhfi  khf min Þexp½chf ðpi  phf Þ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð3Þ

where khf is the permeability of the HF subsystem, phf is the pressure in the HF subsystem, khfi is the permeability of the HF subsystem
at initial pressure, chf is the permeability modulus of the HF subsystem, and khf min is the minimum HF permeability.

2019 SPE Journal 3

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 4 Total Pages: 20

The ratio of khf min and khfi can be determined on the basis of experimental measurements. Luo et al. (2018) indicated that khf min =khfi
can be equal to 0.64, 0.15, or 0.10 for different types and concentrations of proppants. The value of khf min =khfi refers to the lower limit
of the permeability for the proppant-filled HFs. The case with a larger value of permeability modulus chf and larger value of khf min =khfi
means that the proppant-filled fractures are sensitive to pressure change and can easily reach the lower limit with high remaining perme-
ability. Therefore, the stress-sensitivity effect might have a negligible influence on the transient pressure response, though the perme-
ability modulus chf is large.
The Forchheimer equation is used most often to describe non-Darcy flow behavior, which can be described as follows
khfi
khf ¼ ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð4Þ
1 þ khfi bqv=l
where b is the inertial flow parameter, q is fluid density, l is fluid viscosity, and v is the fluid superficial velocity.
The Barree-Conway model takes a more general form for non-Darcy flow by replacing the constant b in the Forchheimer equation
with two new parameters (i.e., minimum permeability plateau and characteristic length) to account for the plateau area that occurred at
a high flow rate (Barree and Conway 2004, 2009),
 
1  kmr
khf ¼ khfi kmr þ ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð5Þ
1 þ qv=ls
where kmr is the minimum permeability relative to Darcy permeability and s is the characteristic length.
The Barree-Conway model not only can overcome the limitation of the Forchheimer non-Darcy equation at high flow rates, but it
also provides a single equation to describe the entire range of flow rates (Lai et al. 2012). The Barree-Conway model can be simplified
to the Forchheimer equation by assuming the relative minimum permeability kmr ¼ 0 and characteristic length s ¼ 1=ðbkhfi Þ. In the fol-
lowing validation section, such a simplified Barree-Conway model was used to describe the non-Darcy flow because only the
Forchheimer equation can be chosen in the commercial numerical simulator used in our study.

Governing Equations. Matrix Subsystem. The permeability of the matrix is usually far less than that of the surrounding NFs
(km  knf ), so the flow in the matrix can be neglected, and the matrix is considered as evenly distributed sources that release fluids into
NF subsystems (Li et al. 2017). Therefore, the governing equation for the matrix subsystem can be expressed as follows:
@ðq/m Þ
þ qex ¼ 0; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð6Þ
@t
where /m is porosity of the matrix subsystem, t is real time, and crossflow (i.e., qex ) between the matrix subsystem and the NF subsys-
tem can be written as
akm q
qex ¼ ðpm  pnf Þ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð7Þ
l
where qex is the mass crossflow between the matrix and NF subsystem, pm is pressure in the matrix subsystem, km is the permeability of
the matrix subsystem, and a is the shape factor.
It is assumed that there is no flow in the matrix block, and the boundary conditions can be expressed as the following:
@pm @pm
¼ ¼ 0; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð8aÞ
@x @y
with the following initial condition

pm ðx; y; t ¼ 0Þ ¼ pi ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð8bÞ

Natural-Fracture Subsystem. The governing partial-differential equation for the NF subsystem by taking stress sensitivity into
account can be described as follows
   
@ cnf ðpi pnf Þ @pnf @ cnf ðpi pnf Þ @pnf km /nf lctnf @pnf
e þ e þ aðpm  pnf Þ ¼ ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð9Þ
@x @x @y @y knfi knfi @t

where ctnf is the total compressibility for the NF subsystem and /nf is the porosity of the NF subsystem.
In addition, the initial and boundary conditions can be described as follows:

pnf ðx; y; t ¼ 0Þ ¼ pi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ð10aÞ


 
@pnf  @pnf 
¼ 0 0  y  y ; ¼ 0; 0  x  xe ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð10bÞ
@x x¼0;xe @y y¼0;ye
e

where xe is the reservoir length along the x-direction, and ye is the reservoir length along the y-direction.
Because HF surfaces serve as the inner boundaries, flow from the NF subsystem to the HF subsystem can be accounted for by the
source term, qf ðx; tÞ, which represents the flow rate from the NFs to the unit length of HFs.
Hydraulic-Fracture Subsystem. It is assumed that unsteady-state flow in an HF is negligible because the fracture volume is far less
than the reservoir volume (Cinco-Ley and Samaniego-V. 1981). Therefore, the governing equation with the combined effects of non-
Darcy flow and stress sensitivity can be written as follows:
      
khfi @ khf min khf min chf ðpi phf Þ kmr @phf qf
þ 1 e kmr þ 1  þ ¼ 0; . . . . . . . . . . . . . . . . . . . . . . ð11Þ
l @x khfi khfi 1 þ qv=ls @x whf hhf

where whf is HF width and hhf is HF height.

4 2019 SPE Journal

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 5 Total Pages: 20

The boundary and initial conditions can be written as follows:


      
khfi hhf whf khf min khf min chf ðpi phf Þ kmr @phf 
þ 1 e kmr þ 1  ¼ qNi;1 i ¼ 1; 2; …; N . . . . . . . . . . . . ð12aÞ
l khfi khfi 1 þ qv=ls @x x¼xw

@phf 
¼ 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð12bÞ
@x x¼xw þxhf

phf ðx; t ¼ 0Þ ¼ pi ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ð12cÞ

where qNi;1 is the flow rate from the first node of the ith HF to the wellbore and N is the number of the HF.

Nondimensionalization. For the convenience of analysis, storage ratio x, crossflow coefficient k, and other dimensionless variables
are defined as follows (Li et al. 2017; Wang et al. 2017b; Zhang and Yang 2017):
khf whf
CfD ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13aÞ
knfi xhf
q Qsc B
FND ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13bÞ
ls whf hhf
2pknfi hðpi  phf Þ
phfD ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13cÞ
Qsc Bl
2pknfi hðpi  pm Þ
pmD ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13dÞ
Qsc Bl
2pknfi hðpi  pnf Þ
pnfD ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13eÞ
Qsc Bl
2qf xhf
qfD ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13fÞ
Qsc B
qN
qND ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13gÞ
Qsc B
knfi t
tD ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13hÞ
ð/m ctm þ /nf ctnf Þlx2hf
x y
xD ¼ ; yD ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13iÞ
xhf xhf
km 2
k¼a x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13jÞ
knfi hf
/nf ctnf
x¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13kÞ
ð/m cm þ /nf ctnf Þ
Qsc Bl
cnfD ¼ c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13lÞ
2pknfi h nf
Qsc Bl
chfD ¼ c ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð13mÞ
2pknfi h hf

where Qsc is the surface flow rate, B is the formation volume factor (FVF), CfD is the conductivity for HFs, FND is the non-Darcy
number, and qND is the dimensionless flow rate inside the HFs.
For a dual-porosity model, crossflow coefficient (i.e., k) and storage ratio (i.e., x) are the two main characteristic parameters, which
are defined as the ratio of elastic energy of NFs to the total (fractures and matrix) elastic energy and the ability of crossflow from matrix
to the NFs, respectively (Li et al. 2017). The non-Darcy effect can be determined by two dimensionless parameters (i.e., FND and kmr).
Non-Darcy number FND is defined on the basis of the total flow rate from the fractured well (Zhang and Yang 2017), which can be
assumed as a constant for the prescribed production-rate condition. With the definition of qND , the dynamic non-Darcy flow behavior
within an HF can be characterized by the product of FND and qND . Relative minimum permeability kmr is introduced to account for the
plateau area occurring at a high flow rate, which is caused by flow heterogeneity or streamlining (Barree and Conway 2004, 2009). To
tackle the problem of stress sensitivity, the dimensionless permeability modulus for an NF subsystem and an HF subsystem (i.e., cnfD
and chfD ) are defined, which have different values to account for stress sensitivity for the two aforementioned subsystems.
Matrix Subsystem. With the aforementioned definitions, Eqs. 6 through 8b can be rewritten as follows:
@pmD
kðpmD  pnfD Þ þ ð1  xÞ ¼ 0; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð14Þ
@tD
with the following initial and boundary conditions

pmD ðxD ; yD ; tD ¼ 0Þ ¼ 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð15aÞ

@pmD @pmD
¼ ¼ 0: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð15bÞ
@xD @yD

2019 SPE Journal 5

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 6 Total Pages: 20

Natural-Fracture Subsystem. Similarly, nondimensionalized forms of Eqs. 9 through 10b can be described as follows:
"    #
2
cnfD pnfD @ pnfD @ 2 pnfD @pnfD 2 @pnfD 2 @pnfD
e 2
þ 2
 cnfD  cnfD ¼x  kðpmD  pnfD Þ; . . . . . . . . . . . . . . . . . . . . . . . . . ð16Þ
@xD @yD @xD @yD @tD

with boundary conditions


 
@pnfD  @pnfD 
¼ 0; 0  yD  yeD ; ¼ 0; 0  xD  xeD ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð17aÞ
@xD xD ¼0;xeD @yD yD ¼0;yeD

and the initial condition

pnfD ðxD ; yD ; tD ¼ 0Þ ¼ 0; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð17bÞ

Hydraulic-Fracture Subsystem. With the dimensionless definition, Eqs. 11 through 12c can be rewritten as follows:
     
@ CfDmin CfDmin chfD phfD 1  kmr @phfD pqfD
þ 1 e kmr þ  ¼ 0; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð18Þ
@xD CfDi CfDi 1 þ FND qND @xD CfDi

with boundary and initial conditions


 
@phfD  pqNDi;1
dSS dND ¼ ; N ¼ 1; 2; …N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð19aÞ
@xD xD ¼xwD CfDi

@phfD 
¼ 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð19bÞ
@xD xD ¼xwD þxfD

phfD ðxD ; yD ; tD ¼ 0Þ ¼ 0; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð19cÞ

where
 
CfDmin CfDmin chfD phfD
dSS ¼ þ 1 e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð19dÞ
CfDi CfDi
1  kmr
dND ¼ kmr þ : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð19eÞ
1 þ FND qND

Semianalytical Solutions. Matrix/Natural-Fracture Subsystems. It is difficult to obtain the analytical solution of the mathematical
model for the matrix/NF systems because of the strong nonlinearity caused by the stress sensitivity of the NFs. Thus, a perturbation
method is used to weaken the nonlinearity (Pedrosa 1986). The relationship between the Pedrosa deformation function gnfD and the
pressure variable pnfD can be written as follows:
1
pnfD ¼  lnð1  cnfD gnfD Þ: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð20Þ
cnfD

The perturbation technique is applied to rewrite gnfD in the form of the power series

gnfD ¼ gnfD0 þ cnfD gnfD1 þ c2nfD g2nfD2 þ ::: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð21aÞ


1
¼ 1 þ cnfD gnfD þ c2nfD g2nfD þ ::: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð21bÞ
1  cnfD gnfD

Given that the dimensionless permeability modulus is usually small (cnfD  1), the zero-order approximate solution can meet the
required engineering precision (Wang et al. 2015; Li et al. 2017). Therefore, Eqs. 14 and 16 can be respectively converted into the fol-
lowing forms with the Pedrosa’s substitution associated with the zero-order approximation:

@pmD
kðpmD  gnfD0 Þ þ ð1  xÞ ¼ 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð22Þ
@tD
@ 2 gnfD0 @ 2 gnfD0 @gnfD0 @pmD
þ ¼x þ ð1  xÞ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð23aÞ
@x2D @y2D @tD @tD
 
@gnfD0  @gnfD0 
 ¼ 0; 0  y  y ;  ¼ 0; 0  xD  xeD ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð23bÞ
@xD xD ¼0;xeD @yD yD ¼0;yeD
D eD

and the initial condition

gfD0 ðxD ; yD ; tD ¼ 0Þ ¼ 0: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð23cÞ

To solve the diffusivity equations (Eqs. 22 through 23c), Laplace transformation was applied. In the Laplace domain, the solutions
of the aforementioned diffusivity equations can be obtained from their counterparts for a homogeneous reservoir by replacing the
Laplace-transform parameter sD with sD  f ðsD Þ, where f ðsD Þ is the transfer function defined as (Warren and Root 1963)

6 2019 SPE Journal

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 7 Total Pages: 20

xð1  xÞsD þ k
f ðsD Þ ¼ ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð24Þ
ð1  xÞsD þ k
where sD is the Laplace-transform variable.
For a homogeneous reservoir, a slab source function in the Laplace domain was applied. Because Newman’s principle cannot be
used directly in the Laplace domain (Yang et al. 2015), a method proposed by Chen et al. (1991) was adopted to generate the solution
in the Laplace domain and is presented in Appendix A. The solutions for an NFR can be written as follows:
" #
pxfD X1
npxwD npxD
g nfD0 ðxD ; yD ; sD ; uD Þ ¼ q ðsD Þ S p ðyD ; ywD ; yfD ; uD Þ þ 2 Fn ðxfD =xeD Þcos cos S pðnÞ ðyD ; ywD ; yfD ; anD ; uD Þ ;
xeD yfD f n¼1
xeD xeD
                                       ð25Þ

where

uD ¼ sD  f ðsD Þ: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð26Þ

The expression of other terms in Eq. 25 is presented in Appendix A. Compared with the traditional plane source function proposed
by Ozkan and Raghavan (1991a), the solution in Eq. 25 is derived directly in the Laplace domain rather than the real-time domain. As
explained by Ozkan and Raghavan (1991b), the solution is complicated and difficult to calculate with the hyperbolic functions and
infinite sums. The ratio of hyperbolic functions in Eqs. A-10a and A-10b might cause computational divergence when their arguments
approach zero or infinity. Similarly, the computational difficulty can be avoided with the method proposed by Ozkan and
Raghavan (1991b).
Hydraulic-Fracture Subsystem. The diffusivity equation for the HF subsystem is also strongly nonlinear because of the existence
of dSS and dND , which are caused by the effect of non-Darcy flow and stress sensitivity, respectively. A semianalytical method is first
used to discretize each HF into several segments (see Fig. 2b). Then, the average value approximation method is used to cope with the
spatial variation of dSS and dND at each fracture segment. It is worthwhile to note that the value of dimensionless permeability modulus
for the NF subsystem is not identical to that of the HF subsystem (i.e., cnfD 6¼ chfD ), and the magnitude of chfD can be on the order of 100
(Berumen and Tiab 1997; Wang et al. 2017b). To obtain an analytical solution successfully in the Laplace domain, dSS is rewritten as
the product of two terms:

dSS ¼ dSSa  dSSb ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð27Þ

where

dSSa ¼ expðcnfD phfD Þ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð28aÞ


 
CfDmin CfDmin
dSSb ¼ expðcnfD phfD Þ þ 1  exp½ðchfD  cnfD ÞphfD : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð28bÞ
CfD CfD

Then, the average value of dSSb and dND at each fracture segment is used to approximate dSSb and dND variation along that location,
respectively. The function of dSSb and dND vs. time is simplified as

dSSbi; j ðtD Þ ¼ dSSb ðxD ; tD Þ; xDi; j  xD < xDi; jþ1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð29aÞ

dNDi; j ðtD Þ ¼ dND ðxD ; tD Þ; xDi; j  xD < xDi; jþ1 : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð29bÞ

Substituting Eqs. 29a and 29b into Eq. 18, and substituting phfD with the Pedrosa deformation function ghfD , we obtain
1
phfD ¼  lnð1  cnfD ghfD Þ: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð30Þ
cnfD

Then, the governing equation for segment j in fracture i can be rewritten as follows:

@ 2 ghfD pqfDi; j
 ¼ 0; xDi; j  xD  xDi; jþ1 ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð31Þ
@x2D CfDi

with the boundary and initial conditions


 
@ghfD  pqNDi; j
 ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð32aÞ

@xD xD ¼xDi; j CfDi
 
@ghfD  pqNDi; jþ1
 ¼  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð32bÞ
@xD xD ¼xDi; jþ1 CfDi

ghfD ðxD ; tD ¼ 0Þ ¼ 0; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð32cÞ

where
qfDi; j
qfDi; j ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð33aÞ
dSSbi; j dNDi; j

2019 SPE Journal 7

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 8 Total Pages: 20

qNDi; j
qNDi; j ¼ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð33bÞ
dSSbi; j dNDi; j
qNDi; jþ1
qNDi; jþ1 ¼ : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð33cÞ
dSSbi; j dNDi; j

Subsequently, using Laplace transformation, the solution for the HF subsystem can be obtained by integrating Eq. 34 with the
boundary conditions:
!
p X
j1
Dx2hfD X
j1
Dx2hfD 
  
g hfDi; j ¼ g wD  2DxhfD q NDi;k þ DxhfD q NDi; j  q  q NDi; j ; . . . . . . . . . . . . . . . . . . . . . . . . ð34Þ
CfDi k¼1
2 k¼1 NDi;k 8

where g hfDi; j is the value of g hfD at the middle point of segment j in fracture i, g wD is the value of g hfD at the first node of the fracture,
and DxhfD is the dimensionless length of the HF segment.
Coupling Matrix and Fracture Subsystems. It is assumed that HFs are symmetrical with regard to the horizontal wellbore, and
each fracture is discretized into 2M segments (see Fig. 2b), and just one-half of the fracture is considered, totaling the unknown varia-
bles as 2N  M þ 1. Flux from the NF to the HF is denoted by q fDi; j (N  M), flux inside the HF node by q NDi; j (N  M), and the well
bottomhole pressure by g wD .
The pressure-continuity condition for the NF and HF subsystems can be obtained as follows:

phfDi; j ¼ pnfDi; j : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð35Þ

Thus, for the solutions in the Laplace domain, we can obtain g hfDi; j ¼ g nfD0i; j , and N  M equations can be obtained for the
N fractures.
Because it is assumed that incompressible flow occurs in the HF subsystem, the flow rate into the fracture segment equals the flow
rate out of the fracture segment according to the mass-conservation law,
1
q NDi; j  q NDi; jþ1  DxhfD q fDi; j ¼ 0: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð36Þ
2
As such, N  M equations can be obtained for the N fractures.
For the prescribed production-rate condition, the constraint condition can be written as follows:
X
N
1
q NDi;1 ¼ ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð37Þ
i¼1
2sD

where q NDi;1 is the flux flow from the first segment of the ith fracture to the wellbore.
On the basis of the constraint equations, an equation system is obtained,
~ ¼ B;
AX ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð38Þ

where matrix A is a coefficient matrix of dimension (2NM þ 1)(2NM þ 1), X ~ is the vector of unknowns, and ~B is the known vector.
An iterative method is applied here to solve the nonlinear-equation matrix because the variables dSSbi; j and dNDi; j are dependent on
the pressure pfDi; j and flux qNDi; j , respectively. The detailed flow chart is illustrated in Fig. 3.
1. Assign an initial guess for the nonlinear parameter ~ d SSb and ~
d ND .
2. On the basis of the constraint equations (Eqs. 35 through 37), the Eq. 38 system is obtained: AX ~ ¼~ B.
3. Calculate the unknown vector X. ~
4. Compute the pressure and flux for each fracture segment pfDi; j and qNDi; j .
5. Update the nonlinear parameters ~ d SSb and ~ d ND on the basis of Eqs. 8b and 19e and compare them with the assumed values.
~
6. Terminate the iteration if jd SSbnew  ~ d SSb j < e and j~d NDnew  ~d ND j < e; otherwise, use the calculated ~
d SSbnew and ~
d NDnew as the
next estimate, and return to Step 2 until the convergence is achieved.
7. Select a new time and repeat Steps 1 through 6 until maximum time has been reached.
When a stable solution in the Laplace domain has been obtained with iterations, the Stehfest inverse algorithm (Stehfest 1970) can
be used to convert the final solution as that in the real-time domain, where the Stehfest number equals 10. Then, with Eq. 39, we can
obtain the bottomhole-pressure response for the fractured horizontal well incorporating the combined stress-sensitivity effect and non-
Darcy flow effect:
1
pwD ¼  lnð1  cnfD gwD Þ: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ð39Þ
cnfD

The pressure derivative is numerically computed on the basis of the differentiation algorithm presented by Bourdet et al. (1989).
The differentiation algorithm with respect to the logarithm of time uses one point before and one point after the point of interest, calcu-
lates the corresponding derivatives, and places their weighted mean at the point considered (Bourdet et al. 1989).
Fig. 4 shows the efficiency of the iteration method as previously mentioned. The computational accuracy is set as e ¼ 105 , and the
process takes a central-processing-unit time of 75 seconds and 320 seconds, with 50 log-time steps to simulate the transient pressure
behavior for a horizontal well intercepted by one HF and three HFs, respectively. As shown in Fig. 4, the maximum iteration number
is 10, showing that computational accuracy can be quickly achieved after 10 iterations.

Model Validation
The advantage of the model developed in this work is to simultaneously take the stress-sensitive permeability and the non-Darcy flow
within the HFs into account. To the best of our knowledge, there are no appropriate semianalytical solutions in the literature for the
combined effects. Hence, two special cases were carried out to validate the accuracy of the semianalytical model by comparing the

8 2019 SPE Journal

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 9 Total Pages: 20

results with numerical simulation. Because only the Forchheimer model can be chosen to describe the non-Darcy flow behavior in the
commercial numerical simulator, the simplified Barree-Conway model was adopted for the comparisons by assuming kmr ¼ 0 and the
bq khfi Qsc B
non-Darcy number can be calculated as FND ¼ .
l whf hhf

Initialization

Assume parameters
δSSbi,j = 1, δNDi,j = 1

Solution of each segment in the Solution of each fracture segment


matrix-natural fracture subsystems in the hydraulic-fracture subsystem

AX = B

Update the parameters


γ p –(γ –γ )p
δSSbi,jnew = CfDmin/CfDi e nfD hfDi,j + (1 – CfDmin/CfDi ) e hfD nfD hfDi,j
1 – kmr
δNDi,jnew = kmr +
1 + FND qNDi,j

δSSb = δSSbnew No δSSbnew – δSSb < ε


δND = δNDnew δNDnew – δND < ε

Yes
tD = tD new

No
tD < tD max

Yes

Stop

Fig. 3—Flow chart for solving the nonlinear problems.

12
11 N = 3, M = 10, γ nfD = 0.05, γ hfD = 1,
10 FND = 10, kmr = 0.01, ε = 10–5

9
Iteration Number

8
7
6

5
4
3

2
1
0
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD

Fig. 4—Iteration number at different times during the solution for three HF cases.

2019 SPE Journal 9

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 10 Total Pages: 20

Stress-Sensitivity Effect and Darcy Flow. To verify the newly developed models, the case with considering the stress-sensitivity
effect was compared with a commercial numerical simulator. To make the pressure response more obvious, the pay-zone thickness is
set to be a small value (3.0 m) to allow for ignoring the convergence-skin effect (Mukherjee and Economides 1991). The parameters
used for model validation are listed in Table 1. As can be seen in Fig. 5a, comparison of the solution in this work with the results from
the numerical simulation shows a good agreement.

Parameters Value Unit

Reservoir thickness, h 3 m

Reservoir initial pressure, pi


8
1.00×10 Pa

Porosity, φm 0.1 Fraction

Natural fracture permeability, knfi


–16 2
Reservoir 5.00×10 m

Storage ratio, ω 0.1 Fraction

Interporosity coefficient, λ
–3
1.00×10 Fraction

Permeability modulus, γnf


–9 –1
5.35×10 Pa

Well length, LH 1000 m

Horizontal well Wellbore radius, rw 0.1 m

Production rate, Qsc


–5 3
8.0×10 m /s

Fracture number, N 3 integer

Fracture half-length, xhf 100 m

Hydraulic fracture Fracture conductivity, CfDi 50 Dimensionless

Fracture width, whf 0.001 m

Fracture permeability modulus, γhf


–7 –1
1.07×10 Pa

Viscosity, μ 0.001 Pa⋅s

Formation volume factor, B


3 3
Oil 1.1 m /std m

Density, ρ
3
800 kg/m

Table 1—Main parameters used for model validation with stress-sensitivity effect.

101 101

100 100
pwD, dpwD /dlntD

pwD, dpwD /dlntD

10–1 10–1

10–2 Numerical simulation 10–2 Numerical simulation


This work This work

10–3 10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD tD
(a) (b)

Fig. 5—Comparison of pressure and pressure derivative obtained from this work and those from a commercial numerical simulator
with the (a) stress-sensitivity effect (cnfD 5 0:05 and chfD 5 1:0) and Darcy flow (FND 5 0 and kmr 5 1:0); (b) stress-sensitivity effect
(cnfD 5 0:05 and chfD 5 1:0) and non-Darcy flow (FND 5 7:22 and kmr 5 0).

10 2019 SPE Journal

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 11 Total Pages: 20

Stress-Sensitivity Effect and Non-Darcy Flow. The newly developed models are then validated with considering both non-Darcy
flow and stress-sensitivity effect. Because non-Darcy flow can be modeled only for the gas phase in the numerical simulator, we com-
pared the transient pressure response between this work and numerical simulation for a single gas phase. The main parameters used for
model validation are listed in Table 2. As shown in Fig. 5b, there is an excellent agreement between this work and those from numerical
simulation, confirming the reliability of the newly developed models to describe the combined non-Darcy flow and stress-
sensitivity effects.

Parameters Value Unit

Reservoir temperature, T 361 K

Reservoir thickness, h 3.05 m

Reservoir initial pressure, pi


7
2.07×10 Pa

Porosity, φm 0.1 Fraction


Reservoir
Natural fracture permeability, knfi
–17 2
9.87×10 m

Storage ratio, ω 0.1 Fraction

Interporosity coefficient, λ
–3
1.00×10 Fraction

Permeability modulus, γnf


–8 –1
2.72×10 Pa

Well length, LH 1219 m

Horizontal well Wellbore radius, rw 0.1 m

Production rate, Qsc


3
0.0328 m /s

Fracture number, N 3 integer

Fracture half-length, xhf 60.96 m

Hydraulic fracture Fracture conductivity, CfDi 374.94 Dimensionless

Fracture width, whf 0.01 m

Fracture permeability modulus, γhf


–7 –1
5.45×10 Pa

Viscosity, μi
–5
1.89×10 Pa⋅s

Gas Specific gravity, γg 0.6 Fraction

Inertial factor, β
5 –1
3.22×10 m

Table 2—Main parameters used for model validation with non-Darcy flow and stress-sensitivity effect.

Although both approaches can be used to solve such a coupled problem, the numerical method usually tends not only to increase the
computational costs for long time periods but also to adopt a small timestep to avoid the undesirable numerical oscillations in the solu-
tion. In addition to performing the required sensitivity analysis within a consistent and unified framework, the method proposed in this
work can be used to generate an accurate solution at any specific time, and it only requires 5–10 iterations at each timestep.

Results and Discussion


In this section, type curves of a horizontal well with multistage fractures are obtained, and the flow regimes are identified. The effects
of some main parameters, including the non-Darcy number, relative minimum permeability, NF permeability modulus, HF permeability
modulus, the value of CfDmin =CfDi , storage ratio, and crossflow coefficient, on the characteristics of the type curves are examined
and analyzed.

Flow-Regime Identification. Fig. 6 depicts the type curves for a horizontal well intercepted by multiple fractures considering non-
Darcy flow and stress-sensitivity effect, and basic data are used to generate the type curves: N ¼ 3, LD ¼ 10, CfDi ¼ 10, x ¼ 0:1,
k ¼ 0:001, FND ¼ 10, kmr ¼ 0:01, cnfD ¼ 0:05, chfD ¼ 1:0, and CfDmin =CfDi ¼ 0:020. Clearly, there are seven possible flow regimes for
the fluid flow in such a well, which can be described as follows:
1. Bilinear flow: This period is characterized by a slope of 0.25 on the pressure-derivative curve (Cinco-Ley et al. 1978) without
considering non-Darcy flow behavior. The slope of the curve will not be affected by the existence of non-Darcy flow; however,
the pressure and its derivative will increase because of the non-Darcy flow effect (Guppy et al. 1982).
2. Linear flow: It can be identified by a slope of 0.50 on the pressure-derivative curve. During this period, fluid flows linearly from
the NFs to each HF (Cinco-Ley and Meng 1988).
3. Early radial flow: This period is present whenever fracture spacing and fracture half-length are suitable (Raghavan et al. 1997).
4. Compound-linear flow: In this period, there will be interactions among the HFs. The flow pattern is predominately normal to the
vertical plane with the wellbore (van Kruysdijk and Dullaert 1989).
5. Transition flow: This is the transition stage between compound-linear flow and diffusion flow (Cinco-Ley and Meng 1988).

2019 SPE Journal 11

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 12 Total Pages: 20

6. Diffusion flow: This stage is marked by a “dip” in the pressure-derivative curves. In this period, fluid starts to flow from the
matrix to the NFs (Wang et al. 2017a).
7. Compound-radial flow: This period usually appears at a later time and can be identified by a fixed value of the pressure deriva-
tive. When considering the stress-sensitivity effect, this pressure derivative is not a constant (Chen et al. 2016b).

102
N = 3, LD = 10, CfDi = 10, ω = 0.1, λ = 0.001,
FND = 10, kmr = 0.05, γ nfD = 0.05, γ hfD = 1.0,
101 CfDmin/CfDi = 0.020

pwD, dpwD /dlntD


100

(7)
10–1
(5)
(4) (6)
(3)
(2)
Non-Darcy flow with stress sensitivity
10–2
Non-Darcy flow without stress sensitivity
Darcy flow without stress sensitivity
(1)
10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD

Fig. 6—Type curves for a horizontal well intercepted by three pairs of symmetrical fractures in an NFR.

As indicated by Zhang and Yang (2017), non-Darcy flow effect is found to impose a significant effect on the early-stage bilinear/
linear flow regime, resulting in an additional pressure drop that is similar to lowering the fracture conductivity. As noted by Wang et al.
(2017b), stress sensitivity of fractures and matrix mainly affects the intermediate- and late-time period, during which the pressure drop
in HFs and matrix is considerable. As can be seen from Fig. 5, the combined effects of non-Darcy flow and stress sensitivity play a sig-
nificant role in all the flow regimes. It is worthwhile noting that erroneous interpretations will be obtained from the classical transient
pressure analysis without considering the combined effects of non-Darcy flow and stress sensitivity. The HF conductivity would be
underestimated because of the non-Darcy flow behavior. The traditional analysis method can result in more than 50% error in the esti-
mation of HF conductivity when the non-Darcy effect exists (Zhang and Yang 2014b), and it also would have mistakenly treated stress-
sensitivity effect as the boundary-dominated flow regime (Thompson et al. 2010).

Sensitivity Analysis. Unlike the traditional models based on the Forchheimer equation, the non-Darcy flow behavior can be character-
ized by two dimensionless parameters (i.e., FND and kmr ) with the Barree-Conway model in this study (Zhang and Yang 2014a). The
stress-sensitivity effect can be determined with the dimensionless permeability modulus (i.e., cnfD and chfD ), and it is worth noting that
stress-sensitivity coefficients in NFs cnfD and in HFs chfD are no longer assumed as equal to achieve a successful solution. Traditionally,
the transient pressure response for the dual-porosity model can be represented by two parameters (i.e., x and k). Accordingly, effects of
these parameters on pressure response are discussed here because other parameters, such as fracture conductivity CfD , wellbore storage,
and fracture spacing LD , have been extensively analyzed (Li et al. 2017; Ren et al. 2017; Wang et al. 2017b).
Effect of Non-Darcy Number. Figs. 7a and 7b illustrate the type curves for FND ¼ 0 (Darcy flow), 1, 10, and 30 at CfDi ¼ 10 and
100, respectively. Other parameters used to generate the type curves are kmr ¼ 0:05, cnfD ¼ 0:05, chfD ¼ 1:0, CfDmin =CfDi ¼ 0:02,
x ¼ 0:1, and k ¼ 0:001. As shown in Figs. 6a and 6b, the pressure drop and pressure-derivative curves increase as the non-Darcy
number FND increases because of the strong non-Darcy effect. If the stress-sensitivity effect is not taken into account, the non-Darcy
flow is found to impose a significant influence in the early time period, resulting in an additional pressure drop, which is similar to
reducing the fracture conductivity (Zhang and Yang 2017). As shown in Fig. 7, however, the non-Darcy number is found to affect the
full-time period in the presence of the stress-sensitivity effect, especially for the low-HF-conductivity case. This is ascribed to the fact
that the apparent fracture conductivity decreases because of the non-Darcy effect, and that the smaller HF conductivity can make the
stress-sensitivity effect more obvious, which shows that the non-Darcy effect can affect the entire flow regime (Yang et al. 2015). In the
presence of stress sensitivity, the smaller fracture conductivity caused by a larger non-Darcy number will result in a larger pressure
drop, which will make the stress-sensitivity effect more remarkable at the late-stage flow regimes, especially for fractures with
low conductivity.
Relative Minimum Permeability. Figs. 8a and 8b present the drawdown type curves for kmr ¼ 0:01, 0.05, 0.10, and 1.00 (Darcy
flow) at CfDi ¼ 10 and 100, respectively. Other parameters used to generate the type curves are FND ¼ 10, cnfD ¼ 0:05, chfD ¼ 1:0,
CfDmin =CfDi ¼ 0:02, x ¼ 0:1, and k ¼ 0:001. As illustrated in Figs. 7a and 7b, the pressure drop and pressure derivative increase when
kmr decreases. This is ascribed to the fact that as kmr is increased to 1.00, the Barree-Conway model can be simplified as Darcy’s model.
The relative minimum permeability is less sensitive than the non-Darcy number (Zhang and Yang 2017). At the late-stage flow regime,
a smaller value of kmr will also result in a larger pressure drop because of the larger decrease of fracture conductivity caused by non-
Darcy flow at the early time period. However, the influence on the late-time period of relative minimum permeability can be negligible
compared with that of the non-Darcy number.
Permeability Modulus of Natural Fractures. Figs. 9a and 9b show the pressure and pressure-derivative curves for cnfD ¼ 0, 0.025,
0.050, and 0.100 at CfDi ¼ 10 and 100, respectively. Other parameters used to generate the type curves are FND ¼ 10, kmr ¼ 0:05,
chfD ¼ 1:0, CfDmin =CfDi ¼ 0:02, x ¼ 0:1, and k ¼ 0:001. The pressure drop becomes larger as the permeability modulus cnfD is
increased. Early-stage flow regimes are approximately independent of the NF permeability modulus. As analyzed previously, the

12 2019 SPE Journal

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 13 Total Pages: 20

bilinear and linear flows are affected by the fracture conductivity, irrespective of the reservoir stress-sensitivity effect, while the
compound-linear flow is determined by the combined influence of the NF permeability modulus, HF permeability modulus, and HF
geometry (Wang et al. 2017b). As such, the stress-sensitivity mainly takes effect on those beyond the compound-linear flow regime,
and such an influence becomes more remarkable for the HFs with low conductivity.

102 102

101 101
pwD, dpwD /dlntD

pwD, dpwD /dlntD


100 100

10–1 10–1

FND = 30 FND = 30
10–2 FND = 10 10–2 FND = 10
FND = 1 FND = 1
FND = 0 (Darcy flow) FND = 0 (Darcy flow)

10–3 10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD tD
(a) (b)

Fig. 7—Effect of non-Darcy number FND on transient pressure behavior at (a) CfDi 5 10 and (b) CfDi 5 100.

102 102

101 101
pwD, dpwD /dlntD

pwD, dpwD /dlntD

100 100

10–1 10–1

kmr = 0.01 kmr = 0.01


kmr = 0.05 kmr = 0.05
10–2 kmr = 0.10 10–2 kmr = 0.10
kmr = 1.00 (Darcy flow) kmr = 1.00 (Darcy flow)

10–3 10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD tD
(a) (b)

Fig. 8—Effect of relative minimum permeability kmr on the transient pressure behavior at (a) CfDi 5 10 and (b) CfDi 5 100.

102 102

101 101
pwD, dpwD /dlntD

pwD, dpwD /dlntD

100 100

10–1 10–1

γnfD = 0.100, γhfD = 1.0 γnfD = 0.100, γhfD = 1.0


10–2 γnfD = 0.050, γhfD = 1.0 10–2 γnfD = 0.050, γhfD = 1.0
γnfD = 0.025, γhfD = 1.0 γnfD = 0.025, γhfD = 1.0
γnfD = γhfD = 0 (without stress sensitivity) γnfD = γhfD = 0 (without stress sensitivity)

10–3 10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD tD
(a) (b)

Fig. 9—Effect of NF permeability modulus cnfD on the transient pressure behavior at (a) CfDi 5 10 and (b) CfDi 5 100.

2019 SPE Journal 13

ID: jaganm Time: 21:01 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 14 Total Pages: 20

Permeability Modulus of Hydraulic Fractures. Figs. 10a and 10b display the pressure response and pressure-derivative curves for
chfD ¼ 0, 0.2, 0.5, and 1.0 at CfDi ¼ 10 and 100, respectively. Other parameters used to generate the type curves are FND ¼ 10,
cnfD ¼ 0:05, cnfD ¼ 0:05, CfDmin =CfDi ¼ 0:02, x ¼ 0:1, and k ¼ 0:001. The larger the HF permeability modulus, the larger the pressure
drop and pressure derivative. Unlike the NF stress sensitivity, which mainly affects the intermediate- and late time period, the HF stress
sensitivity affects the entire flow regime. This is attributed to the fact that the pressure depletion along the HF results in fracture closure,
and then the reduced fracture conductivity causes an extra drop in pressure. As the HF permeability modulus is increased, the influence
becomes more evident. However, the HF stress sensitivity has a negligible effect on the early-time flow regimes for fractures with high
conductivity (Wang et al. 2017b).

102 102

101 101
pwD, dpwD /dlntD

pwD, dpwD /dlntD


100 100

10–1 10–1

γhfD = 2.0, γnfD = 0.050 γhfD = 2.0, γnfD = 0.050


10–2 γhfD = 1.0, γnfD = 0.050 10–2 γhfD = 1.0, γnfD = 0.050
γhfD = 0.5, γnfD = 0.050 γhfD = 0.5, γnfD = 0.050
γhfD = γnfD = 0 (without stress sensitivity) γhfD = γnfD = 0 (without stress sensitivity)

10–3 10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD tD
(a) (b)

Fig. 10—Effect of HF permeability modulus chfD on the transient pressure behavior at (a) CfDi 5 10 and (b) CfDi 5 100.

Minimum Fracture Conductivity. Figs. 11a and 11b demonstrate the effect of minimum HF conductivity CfDmin =CfDi on the pres-
sure and pressure-derivative curves at CfDi ¼ 10 and 100, respectively. Other parameters used to generate the type curves are
FND ¼ 10, kmr ¼ 0:05, cnfD ¼ 0:05, chfD ¼ 1:0, x ¼ 0:1, and k ¼ 0:001. The lower the value of CfDmin =CfDi , the more serious the
stress-sensitivity effect in the HFs. Fractures lose a small part of their initial conductivity with a high value of CfDmin =CfDi . As can be
seen in Fig. 10, the smaller the value of CfDmin =CfDi , the larger the pressure drop and pressure derivative. This parameter mainly affects
the late-time flow regime, as CfDmin refers to the limit condition of stress-sensitive permeability. In fact, the presence of CfDmin prevents
the pressure curve and pressure derivative from increasing upward. For a smaller CfDmin , the pressure and pressure derivative can con-
tinuously increase quickly because of the stress-sensitivity effect (Zhang et al. 2014; Wang et al. 2017b).

102 102
CfDmin /CfDi = 0.005
CfDmin /CfDi = 0.020
CfDmin /CfDi = 0.100
101 101
CfDmin /CfDi = 1.000 (without stress sensitivity
pwD, dpwD /dlntD

pwD, dpwD /dlntD

100 100

10–1 10–1

CfDmin /CfDi = 0.005


10–2 CfDmin /CfDi = 0.020 10–2
CfDmin /CfDi = 0.100
CfDmin /CfDi = 1.000 (without stress sensitivity
10–3 10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD tD
(a) (b)

Fig. 11—Effect of minimum fracture conductivity CfDmin /CfDi on the transient pressure behavior at (a) CfDi 5 10 and (b) CfDi 5 100.

Storage Ratio. Figs. 12a and 12b demonstrate the effect of storage ratio on the pressure and pressure-derivative curves at CfDi ¼
10 and 100, respectively. Other parameters used to generate the type curves are FND ¼ 10, kmr ¼ 0:05, cnfD ¼ 0:05, chfD ¼ 1:0,
CfDmin =CfDi ¼ 0:02, and k ¼ 0:001. As can be seen in Fig. 10, storage ratio x mainly determines the width and depth of the concave
portion of the pressure-derivative curve. The smaller the x, the larger the pressure and pressure derivative for the early and intermediate
time periods, and the wider and deeper the concave portion becomes at the diffusion flow stage (Li et al. 2017). However, the
compound-radial flow regime is not affected by the storage ratio.
Crossflow Coefficient. Figs. 13a and 13b show the effect of crossflow coefficient k on the pressure and pressure-derivative curves
at CfDi ¼ 10 and 100, respectively. Other parameters used to generate the type curves are FND ¼ 10, kmr ¼ 0:05, cnfD ¼ 0:05,
chfD ¼ 1:0, CfDmin =CfDi ¼ 0:02, and x ¼ 0:1. It is shown that the lower the crossflow coefficient k, the later the diffusion occurs.

14 2019 SPE Journal

ID: jaganm Time: 21:02 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 15 Total Pages: 20

A smaller k results in a higher position of the pressure and pressure-derivative curves and a later appearance of the concave portion of
the derivative curve. A smaller value of k means the difference between the matrix and NFs is greater, and the crossflow occurs with
more difficulty compared with that of a larger k (Li et al. 2017).

102 102

101 101
pwD, dpwD /dlntD

pwD, dpwD /dlntD


100 100

10–1 10–1

10–2 ω = 0.05 10–2 ω = 0.05


ω = 0.10 ω = 0.10
ω = 0.30 ω = 0.30

10–3 10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD tD
(a) (b)

Fig. 12—Effect of elastic storage ratio x on the transient pressure behavior at (a) CfDi 5 10 and (b) CfDi 5 100.

102 102

101 101
pwD, dpwD /dlntD

pwD, dpwD /dlntD

100 100

10–1 10–1

λ = 0.0001 λ = 0.0001
10–2 λ = 0.001 10–2 λ = 0.001
λ = 0.01 λ = 0.01

10–3 10–3
10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104 105 106
tD tD
(a) (b)

Fig. 13—Effect of crossflow coefficient k on the transient pressure behavior at (a) CfDi 5 10 and (b) CfDi 5 100.

Field Application. Raghavan et al. (1997) presented the field case as a transient pressure test. The horizontal well was drilled in a shal-
low shelf-deposit reservoir and stimulated with four stages of fractures. The reservoir and fluid properties are listed in Table 3. The
pressure profile is well-matched by using the newly formulated model with single phase and multiple fractures (Fig. 14). It is shown
from the derivative curve that there is a short wellbore-storage effect, and a linear flow regime with the slope of 0.50 develops. Eventu-
ally, the compound-linear flow is found at the late stage. It is difficult to obtain unique match results by the type-curve matching
method. When non-Darcy flow and stress sensitivity are incorporated in the matching process, the matching results are CfDi ¼ 55,
FND ¼ 5, kmr ¼ 0:1, cnfD ¼ 0:15, and cnfD ¼ 3:0, and an excellent agreement exists between the measured and simulated pressure of the
well test for the whole flow period.

Parameters Value Unit

Reservoir thickness, h 10.67 m

Porosity, φm 0.108 Fraction

Total compressibility, ct
–9 –1
1.56×10 Pa

Viscosity, μ
–3
1.675×10 Pa⋅s

Formation volume factor, B


3 3
1.097 m /std m

Table 3—Reservoir and fluid properties for the tested well (Raghavan et al. 1997).

2019 SPE Journal 15

ID: jaganm Time: 21:02 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 16 Total Pages: 20

102

101
10,000

100

Δpw , dpw /dlndt (psi)


1,000

pwD, dpwD /dlntD


10–1
100

10–2
10

10–3
0
0.001 0.01 0.1 1 10 100 1,000
10–4 Δt (hours)

10–5

10–6
10–7 10–6 10–5 10–4 10–3 10–2 10–1 100 101 102 103 104
tD

Fig. 14—Type-curve matching of the transient pressure profile.

Conclusions
In this study, a generalized semianalytical model has been developed and validated to examine the combined effects of non-Darcy flow
and stress sensitivity on the transient pressure response for an MFHW in a naturally fractured formation. Non-Darcy flow behavior in
the HFs imposes effects on the early-stage period, and the stress sensitivity in fractures and reservoirs mainly affects the transient pres-
sure response for the intermediate- and late-stage flow regimes. The existence of non-Darcy flow behavior in the HFs can make the
stress-sensitivity effect more significant. The early flow regimes, which are mainly affected by the non-Darcy flow, are hardly influ-
enced by the stress-sensitivity level. The non-Darcy flow behavior can be described by two parameters (i.e., non-Darcy number and rel-
ative minimum permeability), while the non-Darcy number is more sensitive than the minimum relative permeability. In the fractures
with low conductivity, the effect of non-Darcy flow behavior and stress sensitivity in the NFs and HFs on the transient pressure
response can be more obvious than in high-conductivity fractures. The characterizations of the flow regimes for an MFHW can be
greatly changed because of the non-Darcy flow and stress-sensitivity effect.

Nomenclature
B ¼ FVF, bbl/STB
ctm ¼ total compressibility for matrix, psi–1
ctnf ¼ total compressibility for NFs, psi–1
CfD ¼ dimensionless HF conductivity, fraction
h ¼ formation thickness, ft
hhf ¼ HF height, ft
khf ¼ permeability for HF subsystem, md
km ¼ matrix permeability, md
kmr ¼ minimum relative permeability, fraction
knf ¼ permeability for NF subsystem, md
L ¼ fracture spacing, ft
LH ¼ horizontal well length, ft
N ¼ HF number
phf ¼ pressure in HF subsystem, psi
pm ¼ pressure in matrix subsystem, psi
pnf ¼ pressure in NF subsystem, psi
pwf ¼ well bottomhole pressure, psi
qex ¼ crossflow between matrix subsystem and NF subsystem, lbm/(D-ft3)
qf ¼ flux from NF system to HF system, STB/(D-ft)
qN ¼ flow rate at the node along HF, STB/D
Qsc ¼ surface flow rate, STB/D
rw ¼ wellbore radius, ft
s ¼ Laplace-transformation variable
t ¼ real time, hours
v ¼ velocity, ft/sec
whf ¼ HF width, ft
xe ¼ reservoir length along x-direction, ft
xhf ¼ HF half-length, ft
xw ¼ well location along x-direction, ft
ye ¼ reservoir length along y-direction, ft
yw ¼ fracture location along y-direction, ft
a ¼ shape factor of the matrix subsystem, ft–2
b ¼ inertial factor, ft–1
chf ¼ permeability modulus for HFs, psi–1
cnf ¼ permeability modulus for NFs, psi–1

16 2019 SPE Journal

ID: jaganm Time: 21:02 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 17 Total Pages: 20

e ¼ computational accuracy
l ¼ fluid viscosity, cp
q ¼ fluid density, lbm/ft3
s ¼ characteristic length, ft–1
g ¼ Pedrosa deformation function, psi
dND ¼ parameter defined in Eq. 19e
dSS ¼ parameter defined in Eq. 19d
dSSa ¼ parameter defined in Eq. 28a
dSSb ¼ parameter defined in Eq. 28b
/m ¼ matrix porosity, fraction
/nf ¼ NF porosity, fraction

Subscripts
D ¼ dimensionless
hf ¼ hydraulic fracture
i ¼ initial condition, or the ith node
j ¼ the jth node
m ¼ matrix
nf ¼ natural fracture
0 ¼ zero-order approximation

Acknowledgments
The authors acknowledge a Discovery Development Grant, a Discovery Grant, and a Collaborative and Research Development Grant
awarded to D. Yang from the Natural Sciences and Engineering Research Council of Canada and the financial support from DM Oil-
field High-Tech Ltd. The authors appreciate the anonymous reviewers for their constructive comments and valuable suggestions.

References
Abass, H., Sierra, L., and Tahini, A. 2009. Optimizing Proppant Conductivity and Number of Hydraulic Fractures in Tight Gas Sand Wells. Presented at
the SPE Saudi Arabia Section Technical Symposium, Al-Khobar, Saudi Arabia, 9–11 May. SPE-126159-MS. https://doi.org/10.2118/126159-MS.
Barree, R. D. and Conway, M. W. 2004. Beyond Beta Factors: A Complete Model for Darcy, Forchheimer and Trans-Forchheimer Flow in Porous
Media. Paper presented at the SPE Annual Technical Conference and Exhibition, Houston, Texas, 26–29 September. SPE-89325-MS. https://doi.org/
10.2118/89325-MS.
Barree, R. D. and Conway, M. W. 2009. Multiphase Non-Darcy Flow in Proppant Packs. SPE Prod & Oper 24 (2): 257–268. SPE-109561-PA. https://
doi.org/10.2118/109561-PA.
Berumen, S. and Tiab, D. 1997. Interpretation of Stress Damage on Fracture Conductivity. J. Pet. Sci. Eng. 17 (1–2): 71–85. https://doi.org/10.1016/
S0920-4105(96)00057-5.
Bourdet, D., Ayoub, J. A., and Pirard, Y. M. 1989. Use of Pressure Derivative in Well Test Interpretation. SPE Form Eval 4 (2): 293–302. SPE-12777-PA.
https://doi.org/10.2118/12777-PA.
Chen, H. Y., Poston, S. W., and Raghavan, R. 1991. An Application of the Product Solution Principle for Instantaneous Source and Green’s Functions.
SPE Form Eval 6 (2): 161–167. SPE-20801-PA. https://doi.org/10.2118/20801-PA.
Chen, S., Li, H., Zhang, Q. et al. 2008. A New Technique for Production Prediction in Stress-Sensitive Reservoirs. J Can Pet Technol 47 (3): 49–54.
PETSOC-08-03-49. https://doi.org/10.2118/08-03-49.
Chen, Z., Liao, X., Zhao, X. et al. 2016a. Development of a Trilinear Flow Model for Carbon Sequestration in Depleted Shale. SPE J. 21 (4):
1386–1399. SPE-176153-PA. https://doi.org/10.2118/176153-PA.
Chen, Z., Liao, X., Zhao, X. et al. 2016b. A Semi-Analytical Mathematical Model for Transient Pressure Behavior of Multiple Fractured Vertical Well
in Coal Reservoirs Incorporating With Diffusion, Adsorption and Stress-Sensitivity. J. Nat. Gas Sci. Eng. 29 (February): 570–582. https://doi.org/
10.1016/j.jngse.2015.08.043.
Chen, Z., Liao, X., Zhao, X. et al. 2017. A Comprehensive Productivity Equation for Multiple Fractured Vertical Wells With Non-Linear Effects Under
Steady-State Flow. J. Pet. Sci. Eng. 149: 9–24. https://doi.org/10.1016/j.petro.2016.09.050.
Cinco-Ley, H. and Meng, H. 1988. Pressure Transient Analysis of Wells With Finite Conductivity Vertical Fractures in Double Porosity Reservoirs. Pre-
sented at the SPE Annual Technical Conference and Exhibition, Houston, Texas, 2–5 October. SPE-18172-MS. https://doi.org/10.2118/18172-MS.
Cinco-Ley, H. and Samaniego-V., F. 1981. Transient Pressure Analysis for Fractured Wells. J Pet Technol 33 (9): 1749–1766. SPE-7490-PA. https://
doi.org/10.2118/7490-PA.
Cinco-Ley, H., Samaniego-V., F., and Dominguez, A. N. 1978. Transient Pressure Behavior for a Well With a Finite-Conductivity Vertical Fracture.
SPE J. 18 (4): 253–264. SPE-6014-PA. https://doi.org/10.2118/6014-PA.
Clarkson, C. R., Qanbari, F., Nobakht, M. et al. 2013. Incorporating Geomechanical Changes Into Rate-Transient Analysis: Example From the Haynes-
ville Shale. SPE Res Eval & Eng 16 (3): 303–316. SPE-162526-PA. https://doi.org/10.2118/162526-PA.
de Swaan, O. A. 1976. Analytic Solutions for Determining Naturally Fractured Reservoir Properties by Well Testing. SPE J. 16 (3): 117–122. SPE-
5346-PA. https://doi.org/10.2118/5346-PA.
Diwu, P., Liu, T., You, Z. et al. 2018. Effect of Low Velocity Non-Darcy Flow on Pressure Response in Shale and Tight Oil Reservoirs. Fuel 216:
398–406. https://doi.org/10.1016/j.fuel.2017.11.041.
Forchheimer, P. H. 1901. Wasserbewegung Durch Boden. Zeitz Ver Duetch Ing. 45: 1782–1788.
Gu, D., Ding, D., Gao, Z. et al. 2017. Pressure Transient Analysis of Multiple Fractured Horizontal Wells in Naturally Fractured Unconventional Reser-
voirs Based on Fractal Theory and Fractional Calculus. Petroleum 3 (3): 326–339. https://doi.org/10.1016/j.petlm.2016.12.005.
Guppy, K. H., Cinco-Ley, H., Ramey Jr., H. J. et al. 1982. Non-Darcy Flow in Wells With Finite-Conductivity Vertical Fractures. SPE J. 22 (5):
681–698. SPE-8281-PA. https://doi.org/10.2118/8281-PA.
Hassanzadeh, H., Pooladi-Darvish, M., and Atabay, S. 2009. Shape Factor in the Drawdown Solution for Well Testing of Dual-Porosity Systems. Adv.
Water Resour. 32 (11): 1652–1663. https://doi.org/10.1016/j.advwatres.2009.08.006.
Hill, R. J., Koch, D. L., and Ladd, A. C. J. 2001. Moderate Reynolds Number Flows in Ordered and Random Arrays of Spheres. J. Fluid Mech.
448: 243–278. https://doi.org/10.1017/s0022112001005936.

2019 SPE Journal 17

ID: jaganm Time: 21:02 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 18 Total Pages: 20

Jalali, Y. and Ershaghi, I. 1987. Pressure Transient Analysis of Heterogeneous Naturally Fractured Reservoirs. Paper presented at the SPE California
Regional Meeting, Ventura, California, 8–10 April. SPE-16341-MS. https://doi.org/10.2118/16341-MS.
Kazemi, H. 1969. Pressure Transient Analysis of Naturally Fractured Reservoirs With Uniform Fracture Distribution. SPE J. 9 (4): 451–462. SPE-2156-A.
https://doi.org/10.2118/2156-A.
Kuchuk, F. and Biryukov, D. 2012. Transient Pressure Test Interpretation From Continuously and Discretely Fractured Reservoirs. Paper presented at
the SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 8–10 October. SPE-158096-MS. https://doi.org/10.2118/158096-MS.
Kuchuk, F., Biryukov, D., Fitzpatrick, T. et al. 2015. Pressure Transient Behavior of Horizontal Wells Intersecting Multiple Hydraulic and Natural Frac-
tures in Conventional and Unconventional Unfractured and Naturally Fractured Reservoirs. Paper presented at the SPE Annual Technical Conference
and Exhibition, Houston, Texas, 28–30 September. SPE-175037-MS. https://doi.org/10.2118/175037-MS.
Lai, B., Miskimins, J. L., and Wu, Y. 2012. Non-Darcy Porous Media Flow According to the Barree and Conway Model: Laboratory and Numerical
Modelling Studies. SPE J. 17 (1): 70–79. SPE-122611-PA. https://doi.org/10.2118/122611-PA.
Li, X., Cao, L., Luo, C. et al. 2017. Characteristics of Transient Production Rate Performance of Horizontal Well in Fractured Tight Gas Reservoirs With
Stress-Sensitivity Effect. J. Pet. Sci. Eng. 158: 92–106. https://doi.org/10.1016/j.petrol.2017.08.041.
Liu, M., Xiao, C., Wang, Y. et al. 2015. Sensitivity Analysis of Geometry for Multi-Stage Fractured Horizontal Wells With Consideration of Finite Con-
ductivity Fractures in Shale Gas Reservoirs. J. Nat. Gas Sci. Eng. 22 (January): 182–195. https://doi.org/10.1016/j.jngse.2014.11.027.
Luo, W. and Tang, C. 2015. A Semianalytical Solution of a Vertical Fractured Well With Varying Conductivity Under Non-Darcy-Flow Condition. SPE J.
20 (5): 1028–1040. SPE-178423-PA. https://doi.org/10.2118/178423-PA.
Luo, W., Tang, C., Feng, Y. et al. 2018. Mechanism of Fluid Flow Along a Dynamic Conductivity Fracture With Pressure-Dependent Permeability
Under Constant Wellbore Pressure. J. Pet. Sci. Eng. 166: 465–475. https://doi.org/10.1016/j.petrol.2018.03.059.
Moradi, M., Shamloo, A., and Dezfuli, A. D. 2017. A Sequential Implicit Discrete Fracture Model for Three-Dimensional Coupled Flow-Geomechanics
Problems in Naturally Fractured Porous Media. J. Pet. Sci. Eng. 150: 312–322. https://doi.org/10.1016/j.petrol.2016.12.027.
Mukherjee, H. and Economides, M. J. 1991. A Parametric Comparison of Horizontal and Vertical Well Performance. SPE Form Eval 6 (2): 209–216.
SPE-18303-PA. https://doi.org/10.2118/18303-PA.
Ozkan, E. and Raghavan, R. 1991a. New Solutions for Well-Test-Analysis Problems: Part 1—Analytical Considerations. SPE Form Eval 6 (3):
359–368. SPE-18615-PA. https://doi.org/10.2118/18615-PA.
Ozkan, E. and Raghavan, R. 1991b. New Solutions for Well-Test-Analysis Problems: Part 2—Computational Considerations. SPE Form Eval 6 (3):
369–378. SPE-18616-PA. https://doi.org/10.2118/18616-PA.
Pedrosa, O. A. 1986. Pressure Transient Response in Stress-Sensitive Formations. Paper presented at the SPE California Regional Meeting, Oakland,
California, 2–4 April. SPE-15115-MS. https://doi.org/10.2118/15115-MS.
Raghavan, R. S., Chen, C.-C., and Agarwal, B. 1997. An Analysis of Horizontal Wells Intercepted by Multiple Fractures. SPE J. 2 (3): 235–245. SPE-
27652-PA. https://doi.org/10.2118/27652-PA.
Ren, Z., Wu, X., Han, G. et al. 2017. Transient Pressure Behavior of Multi-Stage Fractured Horizontal Wells in Stress-Sensitive Tight Oil Reservoirs.
J. Pet. Sci. Eng. 157: 1197–1208. https://doi.org/10.1016/j.petrol.2017.07.073.
Stehfest, H. 1970. Algorithm 368: Numerical Inversion of Laplace Transforms. Commun. ACM 13 (1): 47–49. https://doi.org/10.1145/361953.361969.
Thompson, J. M., Nobakht, M., and Anderson, D. M. 2010. Modeling Well Performance Data From Overpressured Shale Gas Reservoirs. Presented at
the SPE Canadian Unconventional Resources and International Petroleum Conference, Calgary, Alberta, 19–21 October. SPE-137755-MS. https://
doi.org/10.2118/137755-MS.
van Kruysdijk, C. P. J. W. and Dullaert, G. M. 1989. A Boundary Element Solution to the Transient Pressure Response of Multiply Fractured Horizontal
Wells. Presented at the ECMOR I-1st European Conference on the Mathematics of Oil Recovery, Cambridge, England, 1 July. https://doi.org/
10.3997/2214-4609.201411306.
Vincent, M. C., Pearson, C. M., and Kullman, J. 1999. Non-Darcy and Multiphase Flow in Propped Fractures: Case Studies Illustrate the Dramatic
Effect on Well Productivity. Presented at the SPE Western Regional Meeting, Anchorage, Alaska, 26–27 May. SPE-54630-MS. https://doi.org/
10.2118/54630-MS.
Wang, H., Guo, J., and Zhang, L. 2017a. A Semi-Analytical Model for Multilateral Horizontal Wells in Low-Permeability Naturally Fractured Reser-
voirs. J. Pet. Sci. Eng. 149: 564–578. https://doi.org/10.1016/j.petrol.2016.11.002.
Wang, J. and Jia, A. 2014. A General Productivity Model for Optimization of Multiple Fractures With Heterogeneous Properties. J. Nat. Gas Sci. Eng.
21: 608–624. https://doi.org/10.1016/j.jngse.2014.09.024.
Wang, J., Jia, A., Wei, Y. et al. 2017b. Approximate Semi-Analytical Modeling of Transient Behavior of Horizontal Well Intercepted by Multiple
Pressure-Dependent Conductivity Fractures in Pressure-Sensitive Reservoir. J. Pet. Sci. Eng. 153: 157–177. https://doi.org/10.1016/j.petrol.
2017.03.032.
Wang, S., Ma, M., Ding, W. et al. 2015. Approximate Analytical-Pressure Studies on Dual-Porosity Reservoirs With Stress-Sensitive Permeability. SPE
Res Eval & Eng 18 (4): 523–533. SPE-174299-PA. https://doi.org/10.2118/174299-PA.
Warren, J. E. and Root, P. J. 1963. The Behavior of Naturally Fractured Reservoirs. SPE J. 3 (3): 245–255. SPE-426-PA. https://doi.org/10.2118/426-PA.
Weaver, J. D., Rickman, R. D., and Luo, H. 2010. Fracture-Conductivity Loss Caused by Geochemical Interactions Between Man-Made Proppants and
Formations. SPE J. 15 (1): 116–124. SPE-118174-PA. https://doi.org/2118/118174-PA.
Yang, D., Zhang, F., Styles, J. A. et al. 2015. Performance Evaluation of a Horizontal Well With Multiple Fractures by Use of a Slab-Source Function.
SPE J. 20 (3): 652–662. SPE-173184-PA. https://doi.org/10.2118/173194-PA.
Yang, D., Zhang, Q., Fan, L. et al. 1999. Inflow Performance of Horizontal Wells in Naturally Fractured Reservoirs. J. Univ. Pet. (Nat. Sci. Ed.) 23 (6):
44–49.
Zhang, F. and Yang, D. 2014a. Determination of Minimum Permeability Plateau and Characteristic Length in Porous Media With Non-Darcy Flow
Behaviour. J. Pet. Sci. Eng. 119 (7): 8–16. https://doi.org/10.1016/j.petrol.2014.04.018.
Zhang, F. and Yang, D. 2014b. Determination of Fracture Conductivity in Tight Formations With Non-Darcy Flow Behaviour. SPE J. 19 (1): 34–44.
SPE-162548-PA. https://doi.org/10.2118/162548-PA.
Zhang, F. and Yang, D. 2017. Effects of Non-Darcy Flow and Penetrating Ratio on Performance of Horizontal Wells With Multiple Fractures in a Tight
Formation. J. Energy Resour. Technol. 140 (3): 032903-1–032903-11. https://doi.org/10.1115/1.4037903.
Zhang, Z., He, S., Liu, G. et al. 2014. Pressure Buildup Behavior of Vertically Fractured Wells With Stress-Sensitive Conductivity. J. Pet. Sci. Eng.
122: 48–55. https://doi.org/10.1016/j.petrol.2014.05.006.

Appendix A: Derivation of the Slab Source Solution in the Laplace Domain


A slab source function can be obtained by considering the intersection of a slab source in a linear reservoir in the x- and y-direction.
The instantaneous source function can be written as follows:

18 2019 SPE Journal

ID: jaganm Time: 21:02 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 19 Total Pages: 20

Sðx; y; xw ; yw ; xf ; yf ; tÞ ¼ VIIðx; xw ; xf ; tÞVIIðy; yw ; yf ; tÞ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-1Þ

where VIIðx; xw ; xf ; tÞ is the slab source (prescribed flux) in the x-direction and VIIðy; yw ; yf ; tÞ is the slab source (prescribed flux) in
the y-direction.
In the Laplace domain, the product source solution of Eq. A-1 can be expressed as follows:
  X1  
hðs0 Þ hðsn Þ
Sðx; y; xw ; yw ; xf ; yf ; sÞ ¼ 0 S p ðy; yw ; yf ; sÞ þ S pðnÞ ðy; yw ; yf ; an ; sÞ; . . . . . . . . . . . . . . . . . . . . . . . . ðA-2Þ
g ðs0 Þ ð1Þ n¼1
g0 ðsn Þ ð1Þ

where
pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi
2sinhðyf s=vÞ cosh½ðye  yÞ s=vcoshðyw s=vÞ
S p ðy; yw ; yf ; sÞ ¼ pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-3aÞ
s=v v s=vsinhðye s=vÞ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2sinhðyf s=v þ an Þ cosh½ðye  yÞ s=v þ an coshðyw s=v þ an Þ
S pðnÞ ðy; yw ; yf ; an ; sÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; . . . . . . . . . . . . . . . . . . . . . . ðA-3bÞ
s þ an v sinhðye s=v þ an Þ

and

an ¼ ðnp=xe Þ2 : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-4Þ

Then, Eq. A-2 can be rewritten as follows:


" #
2xf X 1
npxw npx
Sðx; y; xw ; yw ; xf ; yf ; sÞ ¼ S p ðy; yw ; yf ; sÞ þ 2 Fn ðxf =xe Þcos cos S pðnÞ ðy; yw ; yf ; an ; sÞ ; . . . . . . . . . . . . . . . ðA-5Þ
xe n¼1
xe xe

where
sin½ðnpÞðxf =xe Þ
Fn ðxf =xe Þ ¼ : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-6Þ
ðnpÞðxf =xe Þ

Because a slab source is used, the source-shape-dependent rate in terms of withdrawal rate is
q
q^ ¼ : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-7Þ
4xf yf h

Using Eqs. A-5 and A-7, the pressure source function in the Laplace domain can be written as follows:
1
Dp hf ðx; y; sÞ ¼ qðsÞSðx; y; xw ; yw ; xf ; yf ; sÞ
4/ct xf yf h
" #
qðsÞ X1
npxw npx
¼ S p ðy; yw ; yf ; sÞ þ 2 Fn ðxf =xe Þcos cos S pðnÞ ðy; yw ; yf ; an ; sÞ :                 ðA-8Þ
2/ct xe yf h n¼1
xe xe

The dimensionless variables and parameters defined in Eqs. 13a through 13m and Eqs. A-2 through A-6 can be rewritten as follows:
" #
pxfD X1
npxwD npxD
p hfD ðxD ; yD ; sD Þ ¼ q fD S p ðyD ; ywD ; yfD ; sD Þ þ 2 Fn ðxfD =xeD Þcos cos S pðnÞ ðyD ; ywD ; yfD ; anD ; sD Þ ; . . . . . ðA-9Þ
xeD yfD n¼1
xeD xeD

where
pffiffiffiffiffi pffiffiffiffiffi pffiffiffiffiffi
2sinhðyfD sD Þcosh½ðyeD  yD Þ sD coshðywD sD Þ
S p ðyD ; ywD ; yfD ; sD Þ ¼ pffiffiffiffiffi ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-10aÞ
sD  sinhðyeD sD Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2sinhðyfD sD þ anD Þcosh½ðyeD  yD Þ sD þ anD coshðywD sD þ anD Þ
S pðnÞ ðyD ; ywD ; yfD ; anD ; sD Þ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; . . . . . . . . . . . . . . ðA-10bÞ
ðsD þ anD Þ  sinhðyeD sD þ anD Þ
s
sD ¼ L2r ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-10cÞ
v
anD ¼ ðnp=xeD Þ2 ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-10dÞ
sin½ðnpÞðxfD =xeD Þ
Fn ðxfD =xeD Þ ¼ : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-10eÞ
ðnpÞðxfD =xeD Þ

Liwu Jiang is currently a PhD degree graduate student in the Petroleum Systems Engineering Program in the Faculty of Engineer-
ing and Applied Science at the University of Regina. His major research interests include reservoir modeling and simulation, tran-
sient pressure analysis, fluid flow in porous media, and exploitation of unconventional resources. Jiang holds a BSc degree and
an MASc degree in petroleum engineering from the China University of Petroleum.
Tongjing Liu is an associate professor in the Enhanced Oil Recovery (EOR) Institute at the China University of Petroleum, Beijing.
His major research areas include decision-making for mature oilfield development, reservoir engineering, conformance control,
transient pressure analysis, tracer survey, software development for EOR processes, and exploitation of unconventional resour-
ces. Liu holds BSc and PhD degrees in petroleum engineering from the China University of Petroleum.

2019 SPE Journal 19

ID: jaganm Time: 21:02 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150


J194501 DOI: 10.2118/194501-PA Date: 12-March-19 Stage: Page: 20 Total Pages: 20

Daoyong Tony Yang is a professor in the Petroleum Systems Engineering Program in the Faculty of Engineering and Applied Sci-
ence at the University of Regina. Previously, he worked as a petroleum engineer for 3 years in the PetroChina TuHa Oilfield Com-
pany with the China National Petroleum Corporation (CNPC) and as a reservoir engineer for 4 years in the Software
Development and Information Centre of Petroleum Engineering, CNPC. Yang’s major research areas include reservoir descrip-
tion and dynamics, phase behavior, mass and heat transfer, assisted history matching, formation evaluation, production optimi-
zation, CO2 EOR and storage, pressure/rate transient analysis, reservoir nanoagents, jet dynamics, artificial-lift methods,
transport phenomena, interfacial interactions in EOR processes, heavy-oil recovery, unconventional-resources exploitation, and
oilfield wastewater treatment. He has authored or coauthored more than 200 refereed-journal articles and conference papers
and holds three patents. Yang is the recipient of the 2011 SPE Canadian Region Formation Evaluation Award, 2013 SPE Cana-
dian Region Reservoir Description and Dynamics Award, and 2018 SPE Canadian Region Distinguished Achievement Award for
Petroleum Engineering Faculty. He is an associate editor for the Journal of Energy Resources Technology (2015–2021). Yang is a
member of SPE, the American Chemical Society, the Society of Exploration Geophysicists, the American Society of Mechanical
Engineers, and the Society of Petrophysicists and Well Log Analysts, and is a registered professional engineer with the Association
of Professional Engineers and Geoscientists of Saskatchewan, Canada. He holds BSc and PhD degrees in petroleum engineering
from the China University of Petroleum (East China) and a PhD degree in petroleum systems engineering from the University
of Regina.

20 2019 SPE Journal

ID: jaganm Time: 21:02 I Path: //chenas03/Cenpro/ApplicationFiles/Journals/SA/SPE/J###/Vol00000/180150/Comp/APPFile/SA-J###180150

Você também pode gostar