Journal of Natural Gas Science and Engineering 59 (2018) 224–236

Contents lists available at ScienceDirect

Journal of Natural Gas Science and Engineering

journal homepage: www.elsevier.com/locate/jngse

Analytical dual-porosity gas model for reserve evaluation of naturally T

fractured gas reservoirs using a density-based approach
Zhenzihao Zhanga,∗, Luis F. Ayala H.b
a
Department of Geoscience, University of Calgary, 2500 University Drive NW, Calgary, Alberta, T2N 1N4, Canada
b
John and Willie Leone Family Department of Energy and Mineral Engineering, The Pennsylvania State University, University Park, PA, 16802, USA

A R T I C LE I N FO A B S T R A C T

Keywords: The development of naturally fractured gas reservoirs often requires the deployment of rigorous techniques for
Reserve evaluation production data analysis incorporating dual-porosity gas behavior. It has been a prominent problem to linearize
Rate-transient analysis and analytically solve the governing equations for dual-porosity gas systems. This study applies a pseudo-
Naturally fractured reservoir pressure-based interporosity flow equation to derive a density-based rate-transient analysis method to accurately
Dual-porosity system
predict the gas production rate and estimate the amount of original gas in place (Gi ) for the systems. The
Gas
Transfer function
methodology also predicts the gas production rate by transforming the response of its liquid counterpart via a
decoupling of the pressure-dependent effects using dimensionless depletion-driven parameters.
For the first time, the density-based flowing material balance method is derived for the dual-porosity gas
reservoir. More than that, an innovative fracture productivity equation that was missing for the dual-porosity
system is derived as well. This study provided detailed derivations for the model and relationship used in past
density-based dual-porosity rate-transient analysis. The dual-porosity productivity equation and the relationship
between average matrix pseudopressure and average fracture pseudopressure are rigorously derived. The re-
scaling relationship between the dual-porosity liquid solution and the dual-porosity gas solution is also de-
monstrated in detail. An appropriate interporosity flow equation for gas is used. Based on that, the results show
that the density-based approach is able to successfully capture the dual-porosity behavior of gas for constant
bottomhole pressure condition.

1. Introduction
Naturally fractured reservoirs are heterogeneous in nature. Many of
them comprise discrete volumes of matrix rock separated by fractures.
The fractures disrupt the matrix blocks and form continuous networks.
Fractures are highly permeable and occupy a small percentage of total
reservoir volume. The matrices, on the other hand, are of low perme-
ability and have a large storage capacity. Because of high fracture
permeability, most of the fluid flow within a reservoir occurs in the
fracture network, while the majority of the fluid is stored in the re-
servoir's matrix.
Studies of naturally fractured reservoirs have been conducted for
decades. This research has proven increasingly important because of
the ballooning production from unconventional reservoirs. Barenblatt
et al. (1960) proposed a dual-porosity model for liquid, using an in-
terporosity flow equation and a diffusivity equation for the fracture
system. The fracture system is treated as a continuous porous
medium, while the matrix system is treated as a source of
hydrocarbon that provides fluids to the fracture system. The dual-
porosity model has been described using partial-differential equa-
tions. Warren and Root (1963) used Barenblatt et al. (1960) method
for analyzing well test data of oil reservoirs. They assumed a sys-
tematic array of identical, rectangular parallelepiped representing
matrix blocks throughout an anisotropic reservoir. Warren and Root
(1963) analytically solved their governing equations for constant-rate
production, examined pressure-buildup performance, and obtained
asymptotic solutions. Techniques for analyzing drawdown and build-
up data were then proposed by them. They pointed out that storativity
ratio and interporosity flow coefficient are controlling parameters in
dual-porosity behavior. Of the two controlling parameters, the stor-
ativity ratio, ω , is a measure of fluid capacitance of the fracture
system, and the interporosity flow coefficient, ξ , represents degree of
heterogeneity in the dual-porosity system. The equations for de-
scribing the dual-porosity reservoir for homogeneous and isotropic
reservoirs following Warren and Root's model in dimensionless form
could be written as follows:

∗
Corresponding author.
E-mail address: zhang.zhenzihao@gmail.com (Z. Zhang).

https://doi.org/10.1016/j.jngse.2018.08.022
Received 12 May 2018; Received in revised form 14 August 2018; Accepted 21 August 2018
Available online 28 August 2018
1875-5100/ © 2018 Elsevier B.V. All rights reserved.
Z. Zhang, L.F. Ayala H.
∂2pfD 1 ∂pfD ∂p ∂pfD
+ = (1 − ω) mD + ω
∂rD2 rD ∂rD ∂tD ∂tD
∂pmD
(1 − ω) = ξ (pfD − pmD )
∂tD
kf h ⎛⎜pi − pf ⎞⎟
ϕf cf k
where ω= ϕf cf + ϕm cm
; ξ = α km rw2 ; pDf =
f
kf h (pi − pm )
fracture pressure ; pDm = qBo μ
, dimensionless matrix pressure;
kf t
rD = rr , dimensionless radius ; tD =
w ⎛ ⎞ 2
⎜ϕf cf + ϕm cm ⎟ μrw
⎝ ⎠
is fracture compressibility plus liquid compressibility; cm is matrix
compressibility plus liquid compressibility; kf is fracture permeability ;
km is matrix permeability; α is shape factor; ϕm is matrix porosity; and
ϕf is fracture porosity. Note that ξ is used instead of λ to avoid con-
fusion with λ in λ and β rescaling approach. Equation (2) is the
pseudo-steady state interporosity flow equation in dimensionless
form, as used in the Warren and Root model proposed for liquid.
Zimmerman et al. (1993) showed derivation of dimensional form of
Equation (2) assuming quasi-steady state flow. Crawford et al. (1976)
analyzed pressure buildup data from known naturally fractured re-
servoirs and found that the Warren and Root model can successfully
describe behaviors of these reservoirs. Da Prat et al. (1981) derived a
constant bottom-hole-pressure solution in a closed, circular dual-
porosity system using the Warren and Root model. Moench (1984)
proposed a transient interporosity flow model incorporating skin ef-
fect at the interface between matrix and fracture. The skin may result
from underground water circulating in groundwater and geothermal
reservoirs. Moench found that the pseudo-steady state interporosity
flow applies even in early stages when the skin effect was strong en-
ough. Wang et al. (2018) extend the dual-porosity to model cases with
fractal property distribution. Yuan et al. (2017) developed a workflow
for analyzing production data of naturally fractured shale formations.
On the gas side, efforts have been made to linearize these governing
equations using concepts of pseudo-pressure and pseudo-time.
However, for gas dual-porosity systems, fluids in the two overlapping
continua may be found at different pressures at any given location,
leading to markedly different gas properties. Resulting flow equations
thus become difficult to linearize and solve analytically. As a result,
traditional production decline analysis using pseudo-functions has yet
to be proven fully successful.
Spivey and Semmelbeck (1995) used pseudo-time and pseudo-
pressure to forecast long-term gas production in shale gas and dewa-
tered coal seams. Pseudo-time accounts for desorption effects by using
modified compressibility. This may predict the long-term gas produc-
tion for a wide range of parameters. However, it is not applicable when
2
there is a small ξreD or drawdown. Moreover, the applicability of this
approach to naturally fractured reservoirs has not been rigorously
proven.
Gerami et al. (2007) used pseudo-functions to study dual-porosity
systems. Pseudo-time has been defined so as to incorporate the water
saturation and compressibilities (Gerami et al., 2007). They presented
an innovative interporosity flow equation for gas without derivation.
The equation they presented is similar to the equation for liquid but
with a changing storativity ratio, ω , given as a function of pressure,
which implies both nonlinearity and an unsolvable nature. To tackle
this problem, Gerami et al. (2007) first solved the equations assuming a
constant ω . The authors assumed that the same analytical result may be
used for gas if ω is updated at each time step. The error in using this
method increases with the CMG-IMEX results as production continues.
Ye and Ayala (2012, 2013) and Ayala and Ye (2012, 2013) proposed
a λ and β density-based approach to analyzing unsteady state flow
under constant bottom-hole pressure. Zhang and Ayala (Ayala and
Zhang, 2013; Zhang and Ayala, 2014a, 2014b) rigorously derived
λ and β and extended them to the variable pressure drawdown/variable
Journal of Natural Gas Science and Engineering 59 (2018) 224–236
rate systems in later work. This density-based approach mostly cir-
cumvents the concepts of pseudo-functions and enables straightforward
(1)
Gi prediction and gas well performance forecasting. Zhang and Ayala
(2016) applied the density-based approach to naturally fractured gas
reservoirs and observed that the density-based method provides accu-
(2)
rate estimations.
In this study, we presented a theoretical basis for using the density-
⎝
qBo μ
⎠
, dimensionless based method in naturally fractured reservoirs and show that behavior
of constant-bottomhole-pressure dual-porosity systems is able to be
captured by this extended approach at the late stages. This work pro-
, dimensionless time; cf vides a theoretical background for Zhang and Ayala (2017), which
applies the developed dual-porosity gas model for more complex vari-
able drawdown/variable-rate cases and extends the theory to the liquid
case. The relationship between matrix gas pseudo-pressure and fracture
gas pseudo-pressure used in that work is rigorously developed in this
study. This work also presents the rigorous development of the dual-
porosity productivity equation used in the previous work. The rescaling
relationship between the dual-porosity liquid solution and the dual-
porosity gas solution is derived as well, explaining the rescaling re-
lationship discovered in Zhang and Ayala (2016).
1.1. Interporosity flow Equation
The interporosity flow equation employed in Barenblatt et al.
(1960) and Warren and Root (1963) has been proposed for a pseudo-
steady state liquid flow from the matrix blocks to the fracture system.
Starting from physical principles, Zimmerman et al. (1993) derived this
interporosity flow equation for liquid using spherical matrix blocks. The
development procedure has assumed the “quasi-steady state” approx-
imation, which treats fracture pressure on the outer boundary, pf , as a
constant throughout the derivation.
Owing to pressure-dependent nature of compressibility and visc-
osity for gas, gas flow from the matrix gridlock behaves differently than
the liquid flow. Thus, it is inadequate to use the interporosity flow
equation for gas. Instead, The use of an appropriate gas interporosity
flow equation is necessary to provide a better description of the gas
flow between the matrix blocks and the fracture system. Following
Zimmerman's et al. (1993) approach and incorporating the compressi-
bility-viscosity effects over time, we are able to symbolically derive an
interporosity flow equation for gas from physical principles as follows:
dρm αkm
ϕm = (m (pf ) − m (pm ) ⎞⎟
dt 2θ ⎠ (3)
where ρm is matrix fluid density, θ = RT / MW , and m (pf ) and m (pm ) are
fracture pseudo-pressure and matrix pseudo-pressure, respectively.
Appendix A shows the development of Equation (3) in detail. In the
development process, an incompressible matrix is assumed for simpli-
city. A general form of interporosity flow equation considering forma-
tion compressibility and connate water is also presented in Appendix A.
We consider no desorption or slippage effects and focus instead on the
dual-porosity behavior of gas flow. Equation (3) is in the similar form to
the interporosity flow equation written by Gerami et al. (2007) without
derivation1. Equation (3) takes into account the compressibility-visc-
osity effects for gas flow out of the matrix blocks. By substituting the
definition of pseudo-pressure, we can rewrite Equation (3) as:
pf pm
dρm
ϕm
dt
⎛
= αkm ⎜ ∫ μ 1cgf dρf − ∫ μ 1 ⎞
dρm ⎟
⎝ 0 gf 0 gm cgm ⎠ (4)
In the case of liquid, Equation (4) collapses to
1
Per personal communication with Dr. Pooladi-Darvish where he indicated
that and they wrote it using an analogy with the liquid formulation.
225
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

dρm αkm constants after integration with respect to the closed boundary condi-
ϕm = (ρ − ρm )
dt μc f (5) tion gives:
ρf ρf ρf
Substituting pf − pm =
1
cl
ln ( ) and ln ( ) ≈
ρm ρm ρm
− 1 into Equation
m (pf ) − m (pwf ) = ⎛⎜
psc qgsc T ⎞
⎟ ln
⎢
⎜
2
⎡ ⎛r ⎞− r ⎤ ⎟

2re ⎥
(5) gives: ⎝ πhkf Tsc ⎠ ⎣ ⎝ rw ⎠ ⎦ (13)
1 ∂ρm αkm Following Dake (1978), we calculate volume-averaged m (pf ) from
ϕm = (pf − pm )
ρm ∂t μ (6) Equation (13) and extend the resulting equation to a reservoir with a
random reservoir shape following Dake (1978). Writing the result in
Equation (6) demonstrates that the gas interporosity flow equation
customary units gives:
collapses to the liquid interporosity flow equation in the Warren and
Root model. Owing to different levels of pressure in the fracture system Tsc kf h
and the matrix system in the presence of gas, the viscosity-compressi- qgsc = (m (pf ) − m (pwf ) ⎞⎟
50294psc TbD, PSS ⎠ (14)
bility terms are different for the gas system, unlike in the liquid system.
The interporosity flow equation written for liquid in the Warren and where bD, PSS is the pseudo-steady state component and m (pf ) is the
Root model is unable to capture the viscosity-compressibility effects volume-averaged fracture pseudo-pressure.
and hence unable to accurately capture the behavior of gas. Equation (14) is derived assuming no skin. Also it is only applicable
to a vertical well. Skin effect can be added as needed by adapting the
2. Deliverability Equation for dual-porosity system equation. Equation (14) is valid for boundary-dominated flow, whether
the production scenario is constant-rate production or constant bot-
The deliverability test plays an important role in determining pro- tomhole-pressure production. We test Equation (14) against 118 dif-
duction capacity of a given reservoir. The deliverability equation for ferent dual-porosity reservoir models whose reservoir and fluid prop-
conventional single-porosity reservoirs has been well established. With erties are different and randomly generated. The comparison of
proper manipulation of the interporosity flow equation, we are able to numerical production results and production from Equation (14) de-
derive the interporosity flow equation for a dual-porosity system, which monstrate exact match in boundary-dominated flow stage for all the
is then able to be utilized for the deliverability test. cases. The wide applicability of deliverability equation beyond pseudo-
Given a circular reservoir, the diffusivity equation in the fracture steady state condition is the same as its single-porosity counterpart. For
system can be written in Equation (7). conventional reservoirs, the applicability of deliverability equation for
constant bottomhole pressure case and variable rate/bottomhole pres-
1 ∂ ∂ (ϕf ρf ) ∂ (ϕm ρm )
− (rρ ur ) = + sure case has been identified and proven in previous studies (Winestock
r ∂r f ∂t ∂t (7) and Colpitts, 1965; Palacio and Blasingame, 1993).
Assuming Darcy flow in the fracture system, we substitute Darcy's Equation (14) correlates the gas production rate with the average
1 fracture pseudo-pressure. However, it is hard to quantify the average
law and d (m (pf )) = 2θ μ c dρf into Equation (7) and multiply both
gf gf
fracture pseudo-pressure from production data or reservoir parameters
sides with reservoir volume, Vres . The equation can then be written in
collected from monitors. For this reason, a technique transferring the
Equation (8).
average fracture pseudo-pressure to predictable average reservoir
MWkf 1 ∂ ⎛ ∂m (pf ) ⎞ ∂ (ϕf ρf + ϕm ρm ) pseudo-pressure is developed in Appendix B. Applying the technique
Vres ⎜r ⎟ = Vres
2RT r ∂r ⎝ ∂r ⎠ ∂t (8) yields correlation among average fracture pseudo-pressure, average
matrix pseudo-pressure, and bottom-hole pseudo-pressure.
Timing both sides of Equation (8) with rdr and integrating the re- Substituting Equation B-7 into Equation B-6 and rearranging the
sulting equation gives Equation (9). terms gives:
re re
MWkf ∂m (pf ) ⎞ ∂ (ϕf ρf + ϕm ρm ) 1 K
Vres
2RT
∫ d ⎛r⎜
∂r
⎟ = Vres ∫ ∂t
rdr m (pf ) =
1+K
m (pm ) +
1+K
m (pwf )
(15)
r ⎝ ⎠ r (9)
0.0062841kf h
For boundary-dominated flow (BDF), following Equation 3.11 in where K = αkm Vres bD, PSS
.
Camacho (1987), we propose an approximation as Equation (10) for Then we can write Equation (14) in terms of m (pm ) instead of m (pf )
relating fluid amount change at every point to average fluid mass by substituting Equation (15) into it:
change across the reservoir:
Tsc kf h
re re re qgsc = (m (pm ) − m (pwf ))
∂ (ϕf ρf + ϕm ρm ) ∂ (ϕf ρf + ϕm ρm ) 50294psc TbD, PSS (1 + K ) (16)
∫ ∂t
rdr ≈ ∫ ∂t
rdr = − ∫ V1res qgsc ρsc rdr
r r r In the long-term, the average matrix pressure corresponding to
(10) m (pm ) is able to be approximated using average pressure p from the
where ρf and ρm represent average fracture fluid density and average material balance equation that treats the reservoir as a tank for con-
matrix fluid density, respectively. Substituting Equation (10) into ventional reservoirs. This approximation is based on the physical fact
Equation (9) gives: that the majority of production in a naturally fractured reservoir ori-
ginates in the matrix as fluids are depleted in the fracture system.
re re
MWkf ∂m (pf ) ⎞ Replacing m (pm ) with m (p) gives:
Vres
2RT
∫ d ⎛r⎜
∂r
⎟ = −qgsc ρsc ∫ rdr
r ⎝ ⎠ r (11) Tsc kf h
qgsc = (m (p) − m (pwf ))
Substituting Vres = πre2 h , ρsc = psc MW / RTsc into Equation (11) gives: 50294psc TbD, PSS (1 + K ) (17)
re
∂m (pf ) ⎞ psc T ⎛ ⎞ where m (p) is pseudo-pressure corresponding to average pressure from
∫ d ⎛⎜r ⎟ = − qgsc ⎜re2 − r 2⎟
∂r πr 2
hk f Tsc
the material balance equation using the tank model.
r ⎝ ⎠ e ⎝ ⎠ (12)
Equation (17) provides a clear relationship between production and
Equation (12) yields the gas rate equation for a dual-porosity re- average reservoir pressure. It should be noted that Equations (15)–(17)
servoir using techniques provided in Dake (1978). Integrating both is also valid for dual-porosity systems with formation compressibility
sides of Equation (12) with respect to radius again and eliminating and connate water as long as rock compressibility and water

226
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

compressibility are much smaller than gas compressibility at reservoir Substituting Equation (25) into Equation (24) and multiplying both
average pressure. sides by dt /(ρm − ρwf ) gives:
d (ρm − ρwf ) N
3. λ and β rescaling for dual-porosity systems =− λm dt
(ρm − ρwf ) 1+K (26)
Ye and Ayala (2012, 2013), and Ayala and Ye (2012, 2013) pro- where (−ρwf ) is added in the differential term because of constant
posed a density-based approach to analyzing the unsteady state flow of bottom-hole pressure. Integrating Equation (26) and substituting
single-porosity gas reservoirs. Using depletion-driven dimensionless d (βm t )
λm = dt
gives:
variables λ and β , they successfully decoupled pressure-dependent ef-
ρm t
fects from depletion. Ye and Ayala (2012) were able to show that the
dimensionless gas rate solution under constant flowing bottom-hole ∫ dln (ρm − ρwf ) = 1−+NK ∫ dβm t
pi 0 (27)
pressure can be rescaled from its liquid counterpart using the depletion-
driven dimensionless variables λ and β . Zhang and Ayala (2014a) Finishing the integration in Equation (27) gives:
subsequently derived λ and β . The rescaling relationship and λ and β as
(ρm − ρwf ) −N
defined by Zhang and Ayala (2014a) are as follows: ln = β t
(ρi − ρwf ) 1+K m (28)
qDgas (tD ) = λ ·qDliq − s (β tD ) (18)
Taking the exponentials of both sides and transforming the resulting
μgi cgi equations gives:
λ =
2θ ⎛⎜ρ − ρ wf ⎟⎞ −N
⎝ ⎠ (ρm − ρwf ) = (ρi − ρwf ) exp ⎛ βm t ⎞
⎝1 + K ⎠ (29)
m (p) − m ⎛⎜pwf ⎞⎟
⎝ ⎠ (19)
Substituting Equation (23) and Equation (29) into Equation (16)
t
∫0 λdt gives:
β =
t (20) 2θTsc kf h λm −N
qsc = − (ρ − ρwf ) exp ⎛ βm t ⎞
where is predicted analytical gas flow rate response at a constant
qDgas 50294psc TbD, PSS μgi cgi 1 + K i ⎝ +K
1 ⎠ (30)
bottom-hole pressure in a single-porosity system with a closed Introduce definitions of qDgas and tD as follows:
boundary, and qDliq − s is the analytical response of the liquid single-por-
osity system with a closed boundary at constant bottom-hole pressure. 158021ρsc μgi cgi qgsc
qDgas =
For a single-porosity system, λ and β have been proven to successfully 2πkf h (ρi − ρwf ) (31)
capture the effects of pressure-dependent properties on the system re-
sponse. With Equation (16), λ and β rescaling is found to be an intrinsic 0.0063283kf t
tD =
characteristic of dual-porosity systems. μgi cgi (ϕm + ϕf ) rw2 (32)
Substituting Equation B-3 into Equation (16) and rearranging the
terms assuming incompressible matrix gives: Substituting Equation (25), Equation (31), and Equation (32) into
Equation (30) in sequence gives:
d (ρm ) ρsc Tsc kh
=− (m (pm ) − m (pwf )) 1 λ 1 2πrw2 1 ϕm + ϕf ⎞
dt 50294ϕm Vres psc TbD, PSS (1 + K ) (21) qDgas = · m exp ⎛⎜− βm tD ⎟
bD, PSS 1 + K b
⎝ D, PSS A 1 + K ϕm ⎠ (33)
Define λm as follows:
Equation (33) is derived similarly to those of Zhang and Ayala
μgi cgi
λm = (2014a). The exponential solution for a single-porosity liquid reservoir
2θ ⎛⎜ρm − ρ wf ⎟⎞ with fracture permeability, kf , and total porosity, ϕm , in dimensionless
⎝ ⎠
form is modified from Fetkovich (1980):
m (pm ) − m ⎜⎛pwf ⎟⎞
⎝ ⎠ (22)
1 1 2πrw2 ϕm + ϕf ⎞
qDliq = exp ⎛⎜− tD ⎟
For long-term behavior, the fracture pressure's fraction in the bD, PSS ⎝ bD, PSS A ϕm ⎠ (34)
average reservoir pressure is minimal considering that the matrix stores 0.0063283kf t
the majority of remaining gas. λm is equivalent to λ and evaluated at where definition of tD is modified to be.
μc ⎛⎜ϕm + ϕf ⎞⎟ rw
2
average pressure, p , as obtained from the material balance equation for ⎝ ⎠
conventional reservoirs. A comparison between Equation (33) and Equation (34) produces
Deriving representation of m (pm ) − m (pwf ) in terms of λm from the λ and β rescaling approach from a single-porosity liquid solution to
Equation (22) gives: dual-porosity gas behavior in the long-term, is shown as:

2θ (ρm − ρwf ) λm qDgas = λd qDliq − s (βd tD ) (35)

m (pm ) − m (pwf ) =
μgi cgi (23) λm μgi cgi βm ∫0t λm dt
where. λd = 1+K
; λm = ; βd = 1+K
; βm = t
.
2θ ⎛⎜ρm − ρwf ⎟⎞
Substituting Equation (23) into Equation (21) and rearranging the ⎝ ⎠
terms gives: m (pm ) − m ⎛⎜pwf ⎞⎟
⎝ ⎠
d (ρm ) θρsc Tsc kh Moreover, with λm and βm , rescaling from a dual-porosity liquid
=− λm (ρm − ρwf )
dt 25147ϕm Vres psc TbD, PSS μgi cgi (1 + K ) (24) solution is found to yield a match between behavior of the liquid so-
lution and its gas counterpart. To explain the match, we derive liquid
Take the notation: rate equation for a well producing in a closed-boundary dual-porosity
θρsc Tsc kf h reservoir. By following derivation process of its gas counterpart, as
N= explained in Equations (7)–(14), we are able to obtain the liquid rate
25147ϕm Vres psc TbD, PSS μgi cgi (25)
equation:

227
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

kf πh Table 2
qsc = (ρf − ρwf )
79014μcl ρsc bD, PSS (36) Reservoir size and Gi for three scenarios.
Scenario 1: Units Values
where qsc is the liquid flow rate, ρf is average fracture liquid density,
and cl is liquid compressibility. Equation (5) can be used to derive the Reservoir outer radius, re Ft 1750
average interporosity flow equation for liquid by integrating Equation Dimensionless outer radius, reD 7000
(5) with respect to volume and dividing the reservoir volume on both Drainage area Acres 221
Gi Bscf 119.9
sides. The resultant equation is written as follows:
Scenario 2:
∂ (ϕm ρm ) αk Reservoir outer radius, re Ft 3500
= 0.0063288 m (ρf − ρm ) Dimensionless outer radius, reD 14000
∂t μcl (37) Drainage area Acres 884
Gi Bscf 479.6
We subsequently develop an exponential liquid rate equation for a Scenario 3:
dual-porosity system in a manner similar to its gas counterpart's deri- Reservoir outer radius, re ft 7000
vation, giving: Dimensionless outer radius, reD 28000
Drainage area acres 3534
1 1 1 2πrw2 1 ϕm + ϕf ⎞ Gi Bscf 1918.4
qDliq = exp ⎜⎛− tD ⎟
1 + K bD, PSS ⎝ bD , PSS A 1 + K ϕm ⎠ (38)
158021ρsc μcqsc 0.0063283kf t presents three scenarios with different reservoir sizes. The rate-time
where qDliq − d = and tD = .
2πkf h ⎜⎛ρi − ρ wf ⎟⎞ μgi cgi (ϕm + ϕf ) rw2 production data are transformed into dimensionless form by means of
⎝ ⎠ their definitions and are compared against rescaled dimensionless gas
Comparing Equation (33) and Equation (38) gives:
production rates from analytical liquid flow rates.
qDgas = λm qDliq − d (βm tD ) (39) The line curves in Fig. 1 are single-porosity liquid solution to con-
stant bottomhole pressure production in a closed-boundary radial
where qDgas is the predicted analytical gas flow rate response at constant system, qD liq −s
. The dotted curves are rescaled liquid solution,
bottom-hole pressure in a dual-porosity system with a closed boundary, liq −s
λ dqD (βd tD) , namely qDgas . The rescaled liquid analytical solution, from
and qDliq − s is the analytical response of the liquid dual-porosity system the plot, is far apart from the liquid analytical solution. The shape of the
with a closed boundary at constant bottom-hole pressure. The re- curve is also stretched from the shape of the single-porosity liquid so-
presentation of qDliq − s in Laplace space is provided in the Raghavan lution. Fig. 2 shows the comparison between the rescaled liquid solu-
(1993). Da Prat et al. (1981) presented qDliq − d in Laplace space. By nu- tion and dimensionless production data from the in-house dual-porosity
merically inverting qDliq − s and qDliq − d to real space by means of the simulator. The lines represent production data from the in-house si-
Stehfest algorithm (1970), we are able to obtain exact values of qDliq − s mulator. A good match between the two production rates in the second
and qDliq − d for rescaling purposes. decline stage in Fig. 2 demonstrates the correctness of the rescaling
theory represented by Equation (35). Figs. 1 and 2 demonstrate that the
4. Case study on λ and β Scaling relationship liquid analytical solution for a single-porosity reservoir is able to be
rescaled to accurately predict the long-term behavior of the corre-
Two proposed λ and β rescaling approaches are validated against sponding dual-porosity gas reservoir. The rescaling approach we de-
three numerically generated reservoir scenarios that exhibit strong veloped is valid for the second decline stage of dual-porosity reservoir
dual-porosity behaviors. Because of the lack of simulators with the production. The second decline stage denotes the shift from the pro-
derived interporosity flow equation for gas in our study, we developed duction mainly contributed by fracture fluids to the production con-
an in-house simulator. An isotropic circular reservoir is described by its tributed to by matrix fluids as depletion in the fracture system develops.
radial system. The well is at the center and fully penetrated with no The derivation of the rescaling approach originates from assumptions
skin. The simulator implementation follows Abou-Kassem et al. (2006). about boundary-dominated flow and ρm = ρ that apply in the second
The simulator adopts implicit discretization and the simple-iteration decline stage. Systems with smaller fracture storage capacities, larger
method (SIM). The method proposed by Lee et al. (1966) is used for the reservoir sizes, or stronger interporosity flows satisfy the approximation
viscosity calculation. The compressibility calculation follows the better than systems with the reverse properties. The more the system
method proposed in Abou-Kassem et al. (1990), and Dranchuk and meet the assumptions, the better the match will be.
Abou-Kassem's (1975) method is used for the compressibility factor Note that the previous rescaling is from the single-porosity liquid
calculation. Gas with rg = 0.55 is assumed in the case for which Table 1 solution for the gas rate. Now we apply the rescaling approach from the
provides a summary of the reservoir and fluid properties. Table 2 dual-porosity liquid rate solution using Equation (39). Lines and scat-
tered points in Fig. 3 represent the dimensionless dual-porosity liquid
Table 1
rate and rescaled dimensionless gas rate, respectively. Fig. 3 represents
Reservoir and fluid properties. the analytical solution for dimensionless production rates of the liquid
dual-porosity reservoir at constant bottomhole pressure and rescaled
Properties Units Values
liquid dual-porosity solution, λm qDliq − d (βm tD ) in Equation (39). The re-
Matrix permeability, km md 0.005 scaled liquid analytical solution is stretched and shifted from liquid
Fracture permeability, kf md 50 analytical solution. Fig. 4 demonstrating the validity of Equation (39)
Matrix porosity, ϕm 0.15 by presenting the match between the rescaled liquid dual-porosity so-
Fracture porosity, ϕf 0.01 lution and qDgas , which is dimensionless production data from in-house
Pay zone thickness, h ft 300 reservoir simulation. Figs. 3 and 4 demonstrate that the liquid analy-
Gas specific gravity, SG (air = 1) 0.55
tical solution for a dual-porosity reservoir may be used to accurately
Wellbore radius, rw ft 0.25
Initial pressure, pi psia 5000 predict the long-term behavior of a corresponding dual-porosity gas
Initial temperature, T °F 200 reservoir using the λ and β rescaling approach from the dual-porosity
Specified wellbore flowing pressure, p wf psia 100 liquid solution. The deviation in the transition stage between the first
Shape factor, α 1/ft2 9.98959 × 10−5 decline stage and the second decline stage in Fig. 4 is because of vio-
Interporosity flow coefficient, ξ 6.2435 × 10−10 lation of the assumption, ρm = ρ , at the transition stage. It is also

228
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

Fig. 1. Rescaled single-porosity liquid analytical solution (gas analytical solution) vs. single-porosity liquid analytical solution.

observed that λ and β rescaling is able to capture gas production be- ρwf
rρ = 1 −
havior at the very beginning. ρi (42)

5. Reserve prediction for naturally fractured reservoirs Substituting Equation (41) and Equation (42) into Equation (40)
gives:
Following Zhang and Ayala (2013), we develop a rate-transient 2ρi kf h 1 ⎛ Gp
qgsc = rρ − ⎞ λm
analysis in a dual-porosity gas system. This method is applicable for ⎜

50294ρsc bD, PSS μgi cgi 1 + K ⎝ OGIP ⎠

⎟

(43)
variable pressure drawdown/variable rate production because of the
applicability of gas deliverability equation for such scenarios. Denote:
Substituting Equation (23) into Equation (16) and adding ρi / ρi into the
e
2ρi kf h
RHS gives: qgdi =
50294ρsc bD, PSS μgi cgi (1 + K ) (44)
ρ wf

qgsc =
ρi Tsc kf h 1
2θ ( ρm
ρi
− ρi )λ m
With Equation (44), Equation (43) develops into:
50294psc T bD, PSS 1 + K μgi cgi (40) rρ 1 Gp 1
λm = λm + e
Consider two correlations: qgsc OGIP qgsc qgdi (45)
ρm Gp
=1− Equation (45) is applicable to dual-porosity system with formation
ρi OGIP (41) compressibility and connate water as long as formation compressibility

Fig. 2. Rescaled single-porosity liquid analytical solution (gas analytical solution) vs. dimensionless production data of dual-porosity reservoir simulation (gas
numerical solution).

229
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

Fig. 3. Rescaled dual-porosity liquid analytical solution (gas analytical solution) vs. dual-porosity liquid analytical solution.

and connate water compressibility are much smaller than gas com-
pressibility.
To use Equation (45), an analysis should be conducted to identify
the radial flow regime in the production history. A diagnostic plot
should also be implemented first to examine whether boundary-domi-
nated flow stage has been achieved. If implemented for transient data,
this specific analysis method would underestimate the reserve due to a
Gp rρ
concave shape of transient data in λm q vs. λm q plot. A pessimistic
gsc gsc
evaluation would be achieved if transient stage data is used for evalu-
ating reserve. Once the boundary-dominated-flow period is identified,
the data collected during the boundary-dominated flow-stage can then
be used for analysis. It could be used for Gi prediction following steps
described below:
rρ Gp
1. First of all, we plot vs. data in a Cartesian plot. rρ is calcu-
qgsc qgsc
lated using Equation (42). λm is calculated using Equation (22). In
the second decline stage, λm is close to λ so we can evaluate it at
average pressure, p . p is obtained from the material balance
equation for conventional systems. Gp is calculated using the
production history from simulation results. We identified the points
at the second decline stage. This could be done by plotting flow rate
against time in a semi-log plot and taking the points at the second
rρ Gp
decline stage. Then, in q vs. q plot, we obtain the best-fit straight
gsc gsc
line through the points at the second decline stage, and calculate the
gradient 1 of the fitted straight line. Afterwards, we use the gradient
Gi
to calculate Gi .
2. For the second step, using the Gi derived from the last step, we
Gp rρ
calculate λm at each time step again and plot λm q vs. λm q . Then
gsc gsc
we obtain the best-fit straight line through the points at the second
decline stage. We also obtain the gradient 1 to calculate Gi .
Gi
3. Thirdly, we calculate the difference between the two Gi values in the
last two iterations. If the difference is larger than the preset cri-
terion, we repeat steps two and three; otherwise, we use the Gi as
calculated.
Equation (45) may be utilized for predicting production by solving
its discretized form and discretized material balance equation together
after obtaining the Gi . The procedure for predicting flow rate is

Fig. 4. Rescaled dual-porosity liquid analytical solution (gas analytical solution) vs. dimensionless production data of dual-porosity reservoir simulation (gas nu-
merical solution).

230
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

λm Gp λm rρ
λm Gp λm rρ Fig. 7. vs. q straight-line analysis for Scenario 3.
Fig. 5. qgsc
vs. q straight-line analysis for Scenario 1. qgsc gsc
gsc

p i
presented in steps 1 to 8: 5. Calculate ()Z
using material balance equation, then calculate p i .
Gp p i p Gi
1. Knowing horizontal-axis coordinate λm q
gsc
in last time step, ⎛ ⎞ = i ⎛⎜1 − p ⎞⎟
i−1 i−1 ⎝Z⎠ Zi
⎝
Gi
⎠ (46)
G G G
⎛λm p ⎞ p
, increase ⎛λm q ⎞
p
by a small value, Δ ⎛λm q ⎞ giving
⎝ qgsc ⎠ ⎝ gsc ⎠ ⎝ gsc ⎠
i
i ρ r
G
⎛λm p ⎞ . 6. Calculate (λm)i using Equation (22). Calculate qigsc knowing ⎛λ m q ⎞ ,
⎝ qgsc ⎠ ⎝ gsc ⎠

ρ r
i (λ m)i and rρ .
2. Calculate corresponding ⎛λ m q ⎞ which is the vertical coordinate of 7. Calculate time using Equation
⎝ gsc ⎠
i
G
p
the point whose horizontal coordinate is ⎛λm q ⎞ on the straight (Gp)i − (Gp)i − 1
⎝ gsc ⎠ (t)i = (t )i − 1 + 2·
(qgsc )i + (qgsc )i − 1 (47)
line.
i i i
3. Calculate ( ) by dividing ⎛⎝λ
rρ
Gp
⎞ by ⎛λm ⎞ .
⎝ qgsc ⎠
mq
rρ
gsc ⎠
Gp
where i denotes the current time step and i-1 denotes the previous time
step.
4. Obtaining rρ from Equation (42), we can calculate (Gp)i knowing rρ
i
and ( ).rρ
Gp
8. Repeat steps 1 to 8.

λm Gp λm rρ
Fig. 6. vs. q straight-line analysis for Scenario 2.
qgsc gsc Fig. 8. Production data of the Austin Chalk field case.

231
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

Table 3
Reservoir and fluid properties for the field example of Austin Chalk.
Properties Units Values

Initial pressure, pi psia 3800

Initial GOR, R si scf/STB 1050
Formation volume factor, Bo RB/STB 1.46
Oil API °API 40
Viscosity, μ o cp 0.26
Temperature, T °F 225
Drainage area acres 40
Drainage radius, re ft 745
Wellbore radius, rw ft 0.25

p G
We suggest a sensitivity analysis on the effect of Δ ⎛λm q ⎞ on pro-
⎝ gsc ⎠
Gp
duction profile. This process could set the optimal Δ ⎛λm q ⎞ for pro-
⎝ gsc ⎠
duction prediction. The petroleum engineers who are seeking an ef-
fective method to predict their reservoir and forecast their production
can use this semi-analytical model to constrain the empirical models. Fig. 9.
Np
vs. q
rρ
straight-line analysis for the field case.
qosc osc
This rate-transient-analysis technique can be extended to cases
where well is put at production at variable bottomhole pressure/rate.
Np rρ
The modification and validation have been discussed in Zhang and Plotting vs. gives Fig. 9. Boundary-dominated flow period is
qosc qosc
Ayala (2017). identified from the straight-line section in Fig. 9. The start of boundary-
dominated flow could also be identified from diagnostic plot.
6. Case study on reserve evaluation The slope of the fitted straight-line is 5.59 E− 8 1/STB . Thus, the
1
reserve is evaluated as 5.59 E− 8 = 1.79E 7 STB . Fitting the transient-stage
Production data from the three scenarios are utilized to validate the data would yield a higher slope and lower estimated reserve. This
proposed rate-transient analysis for predicting Gi in dual-porosity re- evaluated reserve falls well within the range of reserve evaluated from
servoirs. The resulting plots after four iterations of Gi prediction are geological settings, from 7.78E6 STB to 3.34E7 STB .
shown in Fig. 5, Fig. 6, and Fig. 7, corresponding to the three afore-
mentioned scenarios.
7. Concluding Remarks
Negligible differences between derived Gi and the actual Gi are
observed. For Scenario 1, slopes for the fitted straight lines are
This study demonstrates that reserve and long-term production
8.2317 × 10−9 Mscf −1, yielding a Gi estimation of 121.5 Bscf with a
performance in dual-porosity gas reservoirs can be successfully cap-
relative error of 1.32% with respect to the actual Gi . The Gi estimations
tured using a density-based approach as long as boundary-dominated
for Scenarios 2 and 3 are 485 Bscf and 1903.6 Bscf with relative errors
flow prevails. The gas productivity equation was rigorously derived for
of 1.113% and −0.75%, respectively. Straight lines in Figs. 5–7 have
Mscf −1 Mscf −1 dual-porosity systems. Assumptions, development of diffusivity equa-
intercepts of 1.4101 × 10−6 ( )
D
, 4.96 × 10−7 ( )
D
and tions, and proof of the λ and β rescaling relationship between dual-
Mscf −1
2.7862 × 10−7 ( )
, respectively.
D
porosity gas solution and dual-porosity liquid solution were also pre-
sented.
Production data from a well completed in naturally fractured Austin
Chalk (Chen et al., 1986) was used for validation of the rate-transient An in-house simulator for dual-porosity gas systems using the de-
analysis technique developed in this study. The formation thickness rived interporosity flow equation for gas was used to validate the
ranges between 100 feet and 400 feet. The matrix porosity in the for- proposed λ and β rescaling approach for dual-porosity reservoirs.
mation ranges from 3% to 7%. Well 1 in the field cases of the original Successful matches between the dimensionless gas flow rate and the
study is selected. The well's production profile is given in Fig. 8. rescaled dimensionless flow rate with λ and β were obtained, showing
The reservoir and fluid properties are listed in Table 3. that the λ and β rescaling approach successfully captures well perfor-
The bottomhole pressure is assumed 100 psia, a low value, since the mance under constant bottom-hole pressure production.
well is kept being pumped off. In our field case study, the GOR is as- Moreover, a rate-transient analysis method was proposed using a
sumed 10 scf/STB to satisfy our single-phase assumption in the theo- density-based approach to predict Gi and gas production rates.
retical development. The oil compressibility is assumed 1.5 E− 4 1/psi . Naturally fractured oil reservoirs' reserve could also be evaluated by
For a liquid reservoir, gas straight-line analysis equation is rewritten using the same method under the liquid assumption. The successful
in Equation (48) honoring that viscosity and compressibility of liquid implementation of the method proves the validity of the density-based
stays constant. approach as well as its strong effectiveness in reserve evaluation and
production prediction.
rρ 1 Np 1
= + e
qosc Ni qosc qodi (48) Acknowledgments
where
The authors would like to thank members of the Unconventional
0.0022548πkf hρi
e
qodi = Natural Resources Consortium (UNRC) at The Pennsylvania State
μcl ρsc bD, PSS (1 + K ) (49) University for their support throughout this research.

232
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

Nomenclature

Roman

am radius of spherical matrix, L , ft

A area, L2 , ft 2
bD, PSS
1
2
ln ( 4 A
e γ CA rw
′2 ), pseudo-steady component, dimensionless
Bo oil formation volume factor, RB/STB
c compressibility, L − t 2/ M , 1/psi
cf fracture compressibility plus liquid compressibility, L − t 2/ M , 1/psi
CA Dietz's reservoir shape factor, dimensionless
Gi original gas in place, L3 , Mscf
Gp cumulative gas production, L3 , Mscf/D
h thickness, L , ft
k permeability, L2 , md
K rescaling component for dual-porosity system, dimensionless
2
m (p) pseudo-pressure, M / L − t 3 , psia
cp 2
m (p) average pseudo-pressure in a reservoir, M / L − t 3 , psia
cp
MW molecular weight of gas, M / N , lbm/lbmol
Ni original oil in place, L3 , STB
p pressure, M / L − t 2 , psia
qD dimensionless flow rate, dimensionless
qgie initial decline rate for density-based model under full potential drawdown, L3 / t , Mscf/D
e
qodi initial oil decline rate for density-based model under full potential drawdown, L3 / t , STB/D
r radius, L, ft
re external Radius, L, ft
rw wellbore radius, L, ft
rρ wellbore-to-initial density ratio, dimensionless
3
R molar gas constant, M − L2 / N − t 2 − T , 10.73 psia – ft
lbmol – °R
s skin factor or Laplace variable, dimensionless
SG specific gravity, dimensionless
t time, t, days
T temperature, T, °R
lb
u ρm r , M/L2 , cf –ft
ur fluid flow rate in porous media, L/ t , ft/s
Vres reservoir volume, L3 , Mcf
Z compressibility factor, dimensionless

Greek

α shape factor, 1/L2 , 1/ ft 2

β time-averaged λ , dimensionless
β time-averaged λ , dimensionless
βd time-averaged λd , dimensionless
βm time-averaged λm , dimensionless
∗
βm∗ time-averaged λm , dimensionless
γ Euler's constant, 0.5772156649
3
θ RT
, L2 / t 2 , psia – ft
MW lbm
λ viscosity-compressibility dimensionless ratio, dimensionless
λ space-averaged viscosity-compressibility ratio for single-porosity system, dimensionless
λd space-averaged viscosity-compressibility ratio for dual-porosity system, dimensionless
λm average viscosity-compressibility ratio between average matrix pressure and bottom-hole pressure, dimensionless
λm∗ viscosity-compressibility ratio for matrix fluid, dimensionless
μ viscosity, M/L-t, cp
ξ Interporosity flow coefficient, dimensionless
ρ density of fluids, M/L3 , lb/cf
ϕ porosity, dimensionless
ω storativity ratio, dimensionless

Subscript

avg average value in the matrix

D dimensionless
f fracture
g gas

233
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

i initial
l liquid
m matrix
o matrix
sc standard condition
wf wellbore condition

Superscript

gas gas
liq-d liquid in dual-porosity system
liq-s liquid in single-porosity system
− average

Appendix A. Interporosity Gas Flow Equation Development

Starting from the diffusivity equation of gas flow in matrix blocks gives:
∂ (ϕm ρm )
− ∇ ·(ρm um) =
∂t (A-1)
Substituting Darcy's law into Equation A-1 gives:

⎛ k ⎞ ∂ (ϕm ρm )
∇ ·⎜ρm m ∇pm ⎟ =
μgm ∂t (A-2)
⎝ ⎠
Assuming incompressible matrix rock, multiplying both sides by θ and adding term μgm cgm/ μgm cgm at the RHS gives:

⎛ 1 ⎞ 1 ∂ρm
∇ ·⎜km θ ∇ρ = ϕm μgm cgm θ
μgm cgm m ⎟ μgm cgm ∂t (A-3)
⎝ ⎠
1
Substituting dm(pm) = 2θ μ c dρm into Equation A-3 gives:
gm gm
∂m (pm )
∇ ·(km ∇m (pm )) = ϕm μgm cgm
∂t (A-4)
ϕm μgm cgm
Dividing both sides of Equation A-4 by km and dividing both sides by gives:
km
∂m (pm ) km
= ∇2 m (pm )
∂t ϕm μgm cgm (A-5)
Take spherical matrix shape with radius am . The fracture pressure acts as an outer boundary condition of the matrix block. Substituting
μgi cgi
λm∗ = μ c into Equation A-5 gives:
gm gm
∂m (pm ) km
= ∇2 m (pm )
λm∗ ∂t ϕm μgi cgi (A-6)
∗
∫ λm dt
Substituting βm∗ = t
into Equation A-6 gives:
∂m (pm ) km
= ∇2 m (pm )
∂ (βm∗ t ) ϕm μgi cgi (A-7)
∗
where βm∗ t
is in the same form of normalized pseudo-time, is λm∗
corresponding to average pressure in the matrix block. For gas reservoirs, average
λm
reservoir pressure was utilized to evaluate pseudo-time and worked very well during boundary-dominated flow. Expanding Equation A-7 in spherical
coordinates and taking u (r , t ) = m (pm ) r gives:
∂u k m ∂ 2u
∗
=
∂ (βm t ) μgi cgi ϕm ∂r 2 (A-8)
We take spherical matrix shape with radius am . The fracture is around the matrix. And pressure at matrix surface is the same as fracture pressure.
Also we have boundary condition:
u (0, βm∗ t ) = 0 (A-9)

u (am , βm∗ t ) = am m (pm ) (A-10)

u (r , 0) = rm (pm ) (A-11)
Solving Equation A-8 to A-11 for m (pm ) distribution and calculating the average pseudo-pressure gives (Crank, 1975):
∞ 2 2 ∗
mavg (pm ) − m (pi ) 6 1 ⎛ π km n βm t ⎞
=1− ∑ exp ⎜−
m (pf ) − m (pi ) π2 n2
μgi cgi ϕm am2 ⎟ (A-12)
n=1 ⎝ ⎠
Long-term approximation truncates to the first term of the infinite series, giving:

234
Z. Zhang, L.F. Ayala H. Journal of Natural Gas Science and Engineering 59 (2018) 224–236

2 ∗
mavg (pm ) − m (pi ) 6 ⎛ π km βm t ⎞
=1− exp −
⎜ μ c ϕ a2 ⎟
m (pf ) − m (pi ) π2 gi gi m m ⎠ (A-13)
⎝
Lim and Aziz (1995) validated long-term approximation in their derivation for liquid system. We observe this approximation is accurate for
∗t
π 2km βm
2 > 0.5. For a wide variety of cases, the approximation should be viable throughout a long production life. Moving terms in Equation A-13
μgi cgi ϕm am
gives:
2 ∗ m (pf ) − mavg (pm )
6 ⎛ π km βm t ⎞
exp ⎜− =
π 2
μgi cgi ϕm am2 ⎟ m (pf ) − m (pi ) (A-14)
⎝ ⎠
Taking the derivative of Equation A-14 with respect to βm∗ t gives:
∗
1 d (mavg (pm )) 6 ⎛ π 2k
⎞ m βm tm π 2k
= 2 exp ⎜−
m (pf ) − m (pi ) d (βm∗ t ) π μgi cgi ϕm am2 ⎟ μgi cgi ϕm am2 (A-15)
⎝ ⎠
d(mavg (pm )) d(mavg (pm )) dt ∗
Writing * t)
in Equation A-15 as dt * t) and substituting d (βm∗ t ) = λm dt into the resulting equation gives:
d(βm d (βm
∗
1 d (mavg (pm )) 6 ⎛ π 2k m βm t
⎞ π 2km ∗
= 2 exp ⎜− λm
m (pf ) − m (pi ) dt π μgi cgi ϕm am2 ⎟ μgi cgi ϕm am2 (A-16)
⎝ ⎠
Substituting Equation A-14 into Equation A-16 gives:
d (mavg (pm )) π 2km
∗ = (m (pf ) − mavg (pm ))
λm dt μgi cgi ϕm am2 (A-17)
π2
2
is a constant known as shape factor, α , that changes with the geometry of the matrix. Moreover, replacing average pseudo-pressure in matrix
am
∗
volume, mavg (pm ) , with point-specific matrix pseudo-pressure and replacing λm with λm∗ since matrix is treated as points when solving the equations
gives:
d (m (pm )) αkm
ϕm = (m (pf ) − m (pm ))
λm∗ dt μgi cgi (A-18)
1
Application of definition of λm∗ and d(m (pm )) = 2θ μ c dρm into Equation A-18 gives:
gm gm

dρm αkm
ϕm = (m (pf ) − m (pm ) ⎞⎟
dt 2θ ⎠ (A-19)
For reservoirs including formation compressibility and connate water, the interporosity flow equation, Equation A-20, can be derived following
the same procedure as presented from Equations A-1 to A-19.
∂ (ϕm (1 − Swconm ) ρm ) αkm
= (m (pf ) − m (pm ))
∂t 2θ (A-20)
where S wconm is matrix connate water saturation.

Appendix B. Correlating m (pf ) and m (pm )

Writing Equation (3) in customary units gives:

dρm αkm
ϕm = (m (pf ) − m (pm ))
dt 316θ (B-1)
Integrating with respect to volume and dividing reservoir volume on both sides of Equation B-1 gives:
dρm αkm
ϕm = (m (pf ) − m (pm ))
dt 316θ (B-2)
where ρm , m (pf ) and m (pm ) are volume-averaged density, fracture pseudo-pressure and matrix pseudo-pressure respectively. Introducing a valid
approximation for the long-term:
Vres d (ϕm ρm )
qgsc = −
ρsc dt (B-3)
Equation B-3 is a valid assumption for long-term considering that the matrix holds the vast majority of hydrocarbon while the fracture holds a
negligible portion. A majority of the production in naturally fractured reservoir originates from the matrix in the long term as fluids are depleted in
the fracture system and incurs larger interporosity flow. Substituting Equation B-3 into Equation (14) with incompressible matrix and fracture
assumption gives:
Vres d (ρm ) Tsc kf h
ϕ =− (m (pf ) − m (pwf ))
ρsc m dt 50294psc TbD, PSS (B-4)
The difference between volumetric averaged density and average density from material balance is negligible for a wide range of cases. Thus,
substituting Equation B-2 into Equation B-4 gives:

235
Z. Zhang, L.F. Ayala H.
Vres αkm Tsc kf h
(m (pf ) − m (pm )) = − (m (pf ) − m (pwf ))
ρsc 316θ 50294psc TbD, PSS
Moving terms in Equation B-5 gives:
0.0062841kf h
(m (pf ) − m (pm )) = − (m (pf ) − m (pwf ))
αkm Vres bD, PSS
Take the notation:
0.0062841kf h
K= αkm Vres bD, PSS
Substituting Equation B-7 into Equation B-6 and moving terms gives:
1 K
m (pf ) = m (pm ) + m (pwf )
1+K 1+K
For dual-porosity systems with formation compressibility and connate water, Equation B-8 can also be derived following the same procedure.
References
Abou-Kassem, J.H., Mattar, L., Dranchuk, P.M., 1990. Computer calculations of com-
pressibility of natural gas. J. Can. Petrol. Technol. 29 (5) SPE-PETSOC-90-05-10.
https://doi.org/10.2118/90-05-10.
Abou-Kassem, J.H., Farouq Ali, S.M., Islam, M.R., 2006. Petroleum Reservoir Simulations:
a Basic Approach. Elsevier.
Ayala, H.,L.F., Ye, P., 2012. Analysis of unsteady responses of natural gas reservoirs via a
Universal natural gas Type-curve formulation. In: Paper SPE-159956 Presented at the
SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, 8-10
October 2012, . https://doi.org/10.2118/159956-ms.
Ayala, H.,L.F., Ye, P., 2013. Unified decline Type-curve analysis for natural gas wells in
boundary-dominated flow. SPE J. 18 (1), 97–113.
Ayala, H.,L.F., Zhang, M., 2013. Rescaled exponential and density-based decline models:
Extension to variable rate/pressure-drawdown conditions. J. Can. Petrol. Technol. 52
(6), 433–440. SPE-168223-PA. https://doi.org/10.2118/168223-PA.
Barenblatt, G.I., Zheltov, Iu P., Kochina, I.N., 1960. Basic concepts in the theory of
Seepage of homogeneous liquids in Fissured rocks (Strata). J. Appl. Math. Mech. 24,
852–864.
Camacho, V.,R.G., 1987. Well Performance under Solution Gas-drive. Doctoral dis-
sertation, Ph. D. Dissertation. U. Tulsa, Tulsa, OK.
Chen, H.Y., Poston, S.W., Wu, C.H., 1986. Characterization of the Austin Chalk producing
Trend. In: Paper SPE 15533 Presented at the SPE Annual Technical Conference and
Exhibition Held in New Orleans, Louisiana, 5-8 October.
Crank, J., 1975. The Mathematics of Diffusion. Clarendon Oxford, pp. 90–91.
Crawford, G., Hagedorn, A., Pierce, A.E., 1976. Analysis of pressure buildup tests in a
naturally fractured reservoir. J. Petrol. Technol. 28 (11), 1295–1300.
Da Prat, G., Cinco-Ley, H., Ramey, H., 1981. Decline curve analysis using Type curves for
two-porosity systems. Old SPE J. 21 (3), 354–362.
Dake, L.P., 1978. Fundamentals of Reservoir Engineering. Elsevier, pp. 145–150.
Dranchuk, P.M., Abou-Kassem, H., 1975. Calculation of Z Factors for natural Gases using
equations of state. J. Can. Petrol. Technol. 14 (3) SPE-PETSOC-75-03-03. https://doi.
org/10.2118/75-03-03.
Fetkovich, M.J., 1980. Decline curve analysis using Type curves. J. Petrol. Technol. 32,
1065–1077.
Gerami, S., Pooladi-Darvish, M., Hong, H., 2007. Decline curve analysis for naturally
fractured gas reservoirs: a study on the applicability of“ pseudo-time” and“ material
balance pseudo-time”. In: International Petroleum Technology Conference.
Lee, A.L., Gonzalez, M.H., Eakin, B.E., 1966. The viscosity of natural Gases. https://doi.
org/10.2118/1340-PA.
Lim, K., Aziz, K., 1995. Matrix-fracture transfer shape factors for dual-porosity simulators.
236
Journal of Natural Gas Science and Engineering 59 (2018) 224–236
(B-5)
(B-6)
(B-7)
(B-8)
J. Petrol. Sci. Eng. 13 (3), 169–178.
Moench, A.F., 1984. Double‐porosity models for a Fissured groundwater reservoir with
fracture skin. Water Resour. Res. 20 (7), 831–846.
Palacio, J.C., Blasingame, T.A., 1993. Decline Curve Analysis Using Type
Curves—analysis of Gas Well Production Data. pp. 12–14 Paper SPE-25909.
Raghavan, R., 1993. Well Test Analysis. Prentice Hall, Englewood Cliffs, NJ.
Spivey, J., Semmelbeck, M., 1995. Forecasting long-term gas production of dewatered
coal seams and fractured gas shales. In: Low Permeability Reservoirs Symposium.
Stehfest, H., 1970. Algorithm 368: numerical Inversion of Laplace Transforms. Commun.
ACM 13, 47–49.
Wang, W., Yuan, B., Su, Y., Sheng, G., Yao, W., Gao, H., Wang, K., 2018. A composite
dual-porosity fractal model for channel-fractured horizontal well. Eng. Appl. Comput.
Fluid Mech.s. Eng. Appl. Computational Fluid Mech. 12 (1), 104–116.
Warren, J., Root, P.J., 1963. The behavior of naturally fractured reservoirs. Old SPE J. 3
(3), 245–255.
Winestock, A., Colpitts, G.P., 1965. Advances in estimating gas well deliverability. J. Can.
Petrol. Technol. 4 (03), 111–119.
Ye, P., Ayala, H.,L.F., 2012. A density-diffusivity approach for the unsteady state analysis
of natural gas reservoirs. J. Nat. Gas Sci. Eng. 7, 22–34.
Ye, P., Ayala, H.,L.F., 2013. Straightline analysis of flow rate vs. Cumulative-Production
data for the Explicit Determination of gas reserves. J. Can. Petrol. Technol. 52 (4),
296–305. SPE-165583-PA. https://doi.org/10.2118/165583-pa.
Yuan, B., Zheng, D., Moghanloo, R.G., Wang, K., 2017. A novel integrated workflow for
evaluation, optimization, and production predication in shale plays. Int. J. Coal Geol.
180, 18–28.
Zhang, M., Ayala, H.,L.F., 2014a. Gas-rate forecasting in boundary-dominated flow:
constant-bottomhole-pressure decline analysis by Use of rescaled exponential models.
SPE J. 19 (3), 410–417. SPE-168217-PA. https://doi.org/10.2118/168217-PA.
Zhang, M., Ayala, H.,L.F., 2014b. Gas-production-data analysis of variable-pressure-
drawdown/variable-rate systems: a density-based approach. SPE Reservoir Eval. Eng.
17 (4), 520–529. SPE-172503-PA. https://doi.org/10.2118/172503-PA.
Zhang, Z., Ayala, H.,L.F., 2016. Decline curve analysis using a pseudo-pressure-based
interporosity flow equation for naturally fractured gas reservoirs. J. Petrol. Explor.
Prod. Technol. 7 (2), 555–567. https://doi.org/10.1007/s13202-016-0277-z.
Zhang, Z., Ayala, H.,L.F., 2017. Production data analysis for dual-porosity gas/liquid
systems: a density-based approach. J. Nat. Gas Sci. Eng. 46, 143–159. https://doi.
org/10.1016/j.jngse.2017.07.007.
Zimmerman, R.W., Chen, G., Hadgu, T., Bodvarsson, G.S., 1993. A numerical dual‐por-
osity model with Semianalytical Treatment of fracture/matrix flow. Water Resour.
Res. 29 (7), 2127–2137.