SPE 135564

Accurate Simulation of Shale-Gas Reservoirs

Juan Andrade, SPE, Faruk Civan, SPE, Deepak Devegowda, SPE, and Richard Sigal, SPE, University of Oklahoma

Copyright 2010, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Florence, Italy, 19–22 September 2010.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by
the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members.
Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an
abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
Current simulators make assumptions that may be inappropriate for modeling shale gas reservoirs. These assumptions are: 1) that the
system exhibits instantaneous capillary equilibrium, 2) that transport can be completely defined by viscous flow (Darcy’s law), and 3)
that relative permeability is not flow rate dependant. To investigate these assumptions, a one-dimensional reservoir simulator was
developed. This was then modified to implement a solution that does not force instantaneous capillary equilibrium and the effects of
this phenomenon were studied.
In this paper, the requirements for a realistic shale gas simulator are first reviewed and results from a model study of a shale gas
reservoir using a commercial simulator are briefly discussed. A simple example illustrates why instantaneous capillary equilibrium
may be inappropriate and therefore justifies the need to develop simulators that do not make the above assumptions. The mathematical
formulation used to implement a non-equilibrium capillarity model is developed and discussed.
Lastly, a set of simulation studies are discussed where saturation profiles for two-phase flow displacement are compared for capillary
equilibrium and non-equilibrium conditions. We also investigate the effect of wettability by considering a 100% water-wet and a
100% oil-wet formation. The results indicate a dramatic impact on the saturation profiles produced by relaxing the capillary
equilibrium requirement.

Introduction
A valid reservoir simulator is a key enabling technology for the characterization, development, and management of a producing
reservoir. It is a standard tool to plan, model, and predict the results of gas production operations. Furthermore it provides the platform
to integrate geologic, geochemical, geophysical, and petrophysical parameters, and examine their effects on production. The ability to
use simulation to investigate the sensitivity of production to reservoir parameters is particularly important for reservoirs such as shale
gas reservoirs where traditional core measurements are difficult or impractical to perform. Simulation provides the means to both
model gas production and produced water.
As shale gas reservoir rocks are studied in more detail their pore geometry is recognized as being extremely complex. The shales
contain hydrophobic pores in the organics, pores of indeterminate wettability in the non-organic part of the matrix, natural fractures
most likely dominantly water wet, and stimulation produced fractures that possibly have fractional wettability (Sondergeld et al. 2010)
The organics also store methane as adsorbed gas and absorbed gas. This storage mechanism depends on pressure, and as gas is
released from the organics both permeability and porosity could increase (Ambrose et al. 2010; Kang et al. 2010).
The simulator needs to provide for this. There is as yet no clear understanding on how these pore systems are connected and it is very
possible that different reservoirs have different connectivity, so a simulator needs to provide for tools to connect them in multiple
ways and test the effects of changing the connectivity. Correctly simulating production depends on getting the pore geometry and
pressure effects correct.
In shale gas reservoirs such as the Barnett, wells that have been massively hydraulically fractured typically produce only a
fraction of the stimulation fluid. Understanding the deposition of the stimulation fluid is a vital tool for designing completions that
minimize the production of stimulation fluids. A common practice for increasing production and extending the commercial life of a
well is re-stimulation. Re-stimulation of partially depleted shale gas wells carries the hazard of producing a water block. Correct
modeling of the deposition of stimulation fluid is necessary to investigate and avoid this.
Standard commercial simulators make approximations that are inappropriate for porous media with very small pore sizes such as
shale gas reservoir rocks. The most serious approximations are: 1) the assumption of instantaneous capillary equilibrium which is used
in standard simulators to eliminate one of the fluid pressure variables, 2) that transport can be completely defined by viscous flow
(Darcy’s law), and 3) that relative permeability is not flow rate dependent. These assumptions result in simulators that: do not
correctly predict the amount of produced water; do not properly handle the changes in gas transport rates with time, so probably
cannot correctly model gas production; and do not correctly predict the deposition of stimulation water especially during re-
2 SPE 135564

stimulation. The objective of this study is to produce 1-d simulator and use it to evaluate the effects of introducing physical
assumptions that should more correctly represent storage and transport in shale gas reservoirs. .
As part of a comprehensive study of the Barnett shale gas reservoir, Qin, Sigal, and Zhang (Qin 2007) performed an extensive
simulation study using a commercial simulator. The study included examining the fate of the stimulation fluid. For a wide range of
assumed values for permeability, capillary pressure, relative permeability, and hydraulic fracture geometry the fracture fluid was
imbibed into the formation, and much less water was produced back in the study then is typically observed in producing wells. Also if
the gas production rate was taken as too high the water pressure became negative.
The modeling of negative water pressure is almost certainly the result of forcing instantaneous capillary equilibrium. The inability to
simulate the correct water production is most likely the result of incorrect physics, and/or assuming a homogeneous shale reservoir.
Including zones with low gas saturation, very low permeability, very low porosity, or heterogeneous pore geometry can potentially
lead to increased water production. Without the correct physics these cases cannot be separated.
In this work, we incorporate the effect of non-instantaneous capillary equilibrium for oil-water and gas-water systems and study
the impact of the new physics on the movement of the saturation fronts within a 1-D coreflood. The numerical simulator is part of an
in-house code developed as a part of this project. Extensive validation studies were performed with both analytic solutions and
commercially available simulators to assess the accuracy of our numerical implementation. The paper is organized as follows: First,
we present a discussion of capillary equilibrium theory and its application to shale-gas reservoirs. Next, we highlight some of the
challenges in shale-gas reservoirs related to the presence of systems of mixed-wettability interacting with each other and contributing
to the overall flow and storage mechanisms. Following this, we present the mathematical formulation of our numerical reservoir
simulator and follow this with several case studies that incorporate the effects of mixed-wettability and non-instantaneous capillary
equilibrium.
The incorporation of non-instantaneous equilibrium resulted in significant modification to the saturation profiles and their time
dependence that were calculated for the coreflood. This illustrates the need to relax this assumption to correctly model shale gas
reservoirs.

Capillary Equilibrium Theory

A simple physical model illustrates that for tight gas sand and shale gas reservoirs the assumption of instantaneous capillary
equilibrium is a poor approximation. Consider a cone-shaped water wetting capillary tube with an infinite water reservoir at pressure
Pwr at the small end and an infinite gas reservoir at pressure at Pgr at the large end (Fig. 1). During water drainage, the rate of gas
inhibition and the rate at which the gas-water interface moves to the left are controlled by the rate at which water can leave the small
orifice that connects the tube to the infinite water reservoir. It is clear that when the gas phase pressure in the reservoir is increased, the
interface will move to the left and reach a new equilibrium position defined by the capillary equilibrium condition (Eq. 1) and the
capillary pressure Pc.

During the approach to equilibrium, local gas and water pressures will deviate from their corresponding values in the reservoir,
however assuming instantaneous capillary equilibrium may be sufficient to predict the new equilibrium position of the interface. The
situation becomes different if the pressure in the gas reservoir is rapidly lowered. The pressure of the gas in the capillary tube will
rapidly decrease; however, because water has a limited compressibility and because flow rate out of the water reservoir can be made
arbitrarily small, it may not be able to move rapidly enough to replace the displaced gas in order to satisfy Eq. 1. It is entirely possible
that for sufficiently large values of Pc, the local water phase pressure can only satisfy Error! Reference source not found.Eq. 1 by
becoming negative. Consequently, under such conditions, the use of instantaneous capillary equilibrium must be revisited.

Fig. 1. Instantaneous Equilibrium Problem

Tyagi et al. (2007) show that the relative permeability and capillary pressure curves cannot be assumed functions of phase
saturations alone and these curves are functions of both space and time. When fluid conditions in porous media are varied, the fluid
SPE 135564 3

adjusts its distribution throughout the porous media accordingly until a new steady-state is attained. Barenblatt et al. (2003) emphasize
that transition and redistribution from one steady state to another steady state does not occur instantaneously and may in fact require a
substantial amount of time. The redistribution time representing the characteristic time scale of this process varies with the saturation
of the wetting fluid. Its value is zero at zero saturation and approaches infinity as saturation approaches unity. Barenblatt et al. (2003)
considered power-law saturation dependence at low saturations values. They assumed that, for non-equilibrium processes, there is an
effective saturation Seff value different than the non-equilibrium saturation value S, at which the non-equilibrium values of relative
permeability and capillary pressure are equal to their conventional equilibrium values. They estimated the non-equilibrium saturation
value using a kinetic equation and demonstrated the successful application of this approach. However, they had to introduce a number
of simplifying assumptions in order to be able to generate analytical solutions, including constant rock and fluid properties, constant
characteristic redistribution time, and unit proportionality coefficient value in the definition of non-equilibrium kinetics.
The relaxation time concept reasonably describes the transition through the meta-stable states. However, its quantification has been
implemented in various ways. Barenblatt et al. (1995) assumed that the relaxation time is a function of the reciprocal-capillary
pressure derivative. However, Barenblatt et al. (2003) used a constant value for the relaxation time. Silin and Patzek (2004) assumed
that the relaxation time is a function of instantaneous saturation, becoming zero at the irreducible water saturation and infinity at the
residual oil saturation. Multivariate power-law type correlations were also recommended and demonstrated for determination of the
relaxations time (Downar-Zapolski et al. 1996; Badur and Bananskiewicz 1998; Civan 2006; and Michel and Civan 2008).
Alternatively, Gupta and Civan (1994) and Rasmussen and Civan (1998) presented approaches based on the kinetics description of
the fluid redistribution processes at three stages, each described by separate rate equations with different orders of process magnitudes.
Gallego et al. (2006) demonstrated the validity of this approach by means of experimental data. The advantage of this approach is that
it can help determine the characteristic time and other phenomenological parameters by analyzing experimental data obtained by core
testing. In general, although well developed theories to account for these effects are presented in previously published literature, they
have not been tested in the context of numerical reservoir simulation. Additionally, because previous studies focused on the laboratory
scale assessment of oil-water flow in corefloods, any formulation for shale-gas reservoirs must account for the considerable
compressibility of the gas phase.

Wettability Theory
Over the last few years, as an increasing number of shale reservoir samples have been imaged using SEMs with nano-meter resolution,
it has become evident that a significant, portion of the pore geometry is contained as voids in the organic material (Sondergeld et al
2010). These pores should have walls that are hydrophobic and gas-wetting, since these voids in the organic material were created by
the generation of oil and/or gas so never contained brine. Additionally, the organic material plays a second significant storage role
through adsorption. Consequently, the location, distribution and amount of the organics within gas shales are very likely to impact our
estimates of gas reserves. Gas shales also contain brine associated with clays that are water wet. NMR studies (Sondergeld et al. 2010)
show that the water relaxation spectra typically are bi-modal. The slower mode could be interpreted as water in pores with walls with
heterogeneous wettability. Although there is no direct evidence for this, it is very possible that in some cases gas is also present in
non-organic pores of indeterminate wettability. This complex matrix porosity is overlain by a natural fracture system, probably water
wet, and the complex fracture system created by the simulation. As yet, our understanding of the connectivities between these various
pore and wettability systems is somewhat limited. The reported observation that many shale gas reservoir rocks imbibe oil-based mud,
and also that some imbibe water-based mud speaks to the complex wettability and pore geometry structure of the system. Unpublished
NMR measurements show strong surface relaxation for both oil and water implying a heterogeneous wettability system.
Consequently, any valid reservoir model must be able to accommodate the four porosity elements potentially present and provide
for the ability to have arbitrary connectivity between them. Connectivity between systems of different wettability needs to be
established to account for the complex characteristics of the shale gas rocks. The numerical simulator developed as a part of this study
currently does not incorporate flow modeling in mixed wettability systems. There are ongoing efforts to account for these storage and
flow mechanisms. Our work in this paper is focused on the impact of non-instantaneous capillary equilibrium in single-wettability
systems.

Mathematical Formulation
This section discussed the general one-dimensional mathematical framework behind the numerical formulation of our in-house
simulator. Numerical solution to the set of differential equations shown below was handled by finite difference modeling. Central
difference in space and trapezoidal rule in time were coupled to expand the saturation general equation after combining the
formulations presented below. Uniform partition both in space and in time were assumed, therefore solutions are obtained for grid-
blocks equally spaced and at constant intervals of time. Spatial derivatives are approximated by second order central difference while
the time derivative uses a backward difference operator. After discretizing the equations they were programmed in computational tool
to generate solutions such as saturation profiles.
The mass balance equation of phases 1 (water) and 2 (oil for incompressible and gas for compressible flow) are given by

0 1,2 1 .2

Flow equations of each phase are defined by the equations below based on Darcy’s law
4 SPE 135564

.3

Capillary pressure under equilibrium conditions can be defined as

.4
Initial condition indicates that the saturation at every-point in the system is known
, 0 , 0 .5
The general boundary condition can be expressed by Civan (1996). Note that terms enclosed by (+) are calculated immediately
outside the permeable zone while terms enclosed by (-) refers to properties assessed immediately inside the permeable zone

. . . . . .6

Relative permeability and capillary pressure are given as function of Seff as

, .7

, .8

, .9
Consider that, for non-equilibrium processes, there is an effective saturation of Seff value different than the non-equilibrium
saturation value S, at which the non-equilibrium values of relative permeability and capillary pressure are equal to their conventional
equilibrium values. The relationship between these values is given by the following expression (Barenblatt et al. 2003).

. 10

Note that τ value of zero represents capillary equilibrium conditions. The magnitude of τ indicates how fast fluids reach
equilibrium.

Numerical Simulations
The numerical simulation studies utilizing the in-house simulator are organized as follows: We first validate the simulator code with
analytic solutions derived from the frontal advance theory of Welge (1952) and Buckley and Leverett (1942). Next, we incorporate the
effects of non-instantaneous capillary equilibrium and evaluate the impact of formation wettability on the recovery and saturation
front movement in a 1-dimensional coreflood.

Case 1: Validation and Comparison with Commercial Reservoir Simulators and Analytic Solutions.
The 1-dimensional model used for validation represents a core sample with constant properties (Table 1). The relative permeability
curves used for this study are shown in Fig. 2. The model is initially 100% saturated with wetting fluid (water phase), a non-wetting
fluid (oil) is injected at a constant rate through the inlet face of the sample to simulate a drainage process. The outlet pressure is held
constant at 1 atm. We ignore the effects of gravity-induced segregation and capillary pressure and Fig. 4 plots the numerical results
with the analytical solution obtained by applying the fractional flow theory (Buckley and Leverett 1942; Welge 1952). Excellent
agreement between the numerical and analytic solutions at different times and locations are obtained, thereby confirming the validity
of the in-house code.
Core and fluid Properties
Single Rate Drainage
μw 1 cp
μo 1.5 cp
L 10 cm
D 3.5 cm
φ 0.3 fraction
k 50 mD

qinj 0.1 cc/min

Pout 1 Atm

Table 1. Rock and fluid properties (Kulkarni et al. 1998).

We also compare the performance of the in-house simulator with the results obtained from a commercially available simulator. For
this scenario, we include a more realistic non-zero capillary pressure curve for a water-wet core shown in Fig. 3. We plot the resulting
SPE 135564 5

saturation profiles for 3 different times during the coreflood (Before breakthrough, after breakthrough and after infinite injection
sufficiently long period of injection) in Fig. 5 and the results indicate good agreement with the commercial reservoir simulator. The
cumulative water production is presented in Fig. 6.

0.8 Krw
kro

0.6
Kr

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1

Fig. 2. Relative Permeability curves (Kulkarni et al., 1998).

0.8

100% Oil-Wet
0.6
100% Water-Wet

0.4
Pc [atm]

0.2

-0.2

-0.4
0 0.2 0.4 0.6 0.8 1

Fig. 3. Capillary pressure for water-wet and oil-wet rocks (Kulkarni et al. 1998; McDougall et al. 1995).
6 SPE 135564

1.0

0.8

Solid lines: Analytical solution

0.6
Markers: Numerical Solution
Sw

0.4

0.2

20 min 70 min 100 min 250 min

0.0
0 0.2 0.4 0.6 0.8 1
x, Length [nomralized]

Fig. 4. Case 1: Drainage with no-capillary pressure. Solutions for frontal advanced theory and numerical solution.

0.9

0.8

0.7
Sw

0.6

0.5

0.4

20 min 50 min 70 min 100 min 120 min 140 min 200 min 1000 min
0.3
0 0.2 0.4 0.6 0.8 1
x, Length [normalized]

Fig. 5. Case 1: Drainage with capillary pressure effect. Solutions for commercial simulator and numerical solution.
SPE 135564 7

18

16

14

12
Com. Sim.
Qw [cc]

10
Num. Sol.
8

0
0 200 400 600 800 1000
Time [min]

Fig. 6. Case 1: Cumulative water production.

Case 2: Assessing the impact of relaxation time during a drainage process in a water-wet system.
This test was designed to illustrate the dramatic effect of capillary non-equilibrium on the saturation profiles during a drainage
process. The rock and fluid properties utilized for this study are identical to the previous case study. Notice that when the relaxation
time (τ) has a value of zero, the system achieves instantaneous capillary equilibrium. We use the formulation of Barenblatt et al.
(2003) shown in Eq. 10 through the use of an effective saturation Seff that allows us to utilize existing relative permeability and
capillary pressure functions as shown in Eqs. 7-9. The results are plotted in Fig. 7 and 8.

0.9

0.8

0.7
Sw

20 min
100 min
70 min
0.6
50 min

140 min
0.5

1000 min

0.4

τ = 0 (black) τ = 10 (red) τ = 30 (green) τ = 60 (blue)

0.3
0 0.2 0.4 0.6 0.8 1
x, Length [normalized]

Fig. 7. Case 2: Effect of relaxation time on saturation profiles (Drainage).

8 SPE 135564

18

15

12

9 τ=0
Qw [cc]

τ = 10
6 τ = 30
τ = 60
3

0
0 200 400 600 800 1000
Time [min]

Fig. 8. Case 2: Cumulative water production.

Case 3: Assessing the impact of relaxation time during an imbibition process in a water-wet system.
In order to simulate the effects of non-instantaneous capillary equilibrium during an imbibition process, we begin with a core sample
initially saturated with 70% oil. The petrophysical and fluid properties are identical to Cases 1 and 2. Water (wetting phase) is injected
at a constant rate through the inlet face of a core sample while the outlet face is held at a constant pressure. The resulting saturation
profiles at various times are presented in Fig. 9, while the effect on cumulative water production is shown in Fig. 10.

0.9

1000 min
0.8

140 min

0.7
20 min
Sw

0.6

70 min 100 min

0.5
50 min

0.4

τ = 0 (black) τ = 10 (red) τ = 30 (green)

0.3
0 0.2 0.4 0.6 0.8 1
x, Length [normalized]

Fig. 9. Case 3: Effect of relaxation time on saturation profiles (Imbibition).

SPE 135564 9

18

15

12

9
Qo [cc]

τ=0

6 τ = 10
τ = 30
3

0
0 200 400 600 800 1000
Time [min]

Fig. 10. Case 3: Cumulative oil production.

Case 4: Assessing the impact of relaxation time during a drainage process in a strongly oil-wet system.
We repeat the set-up from Case 3 and define the core sample to be strongly oil-wet. Capillary pressure input data is described by the
red dotted curve in Fig. 3. The drainage displacement under capillary non-equilibrium conditions is reproduced in Fig. 11.

0.9

0.8

0.7
Sw

20 min
100 min
70 min
0.6
50 min

140 min
0.5

1000 min

0.4

τ = 0 (black) τ = 5 (green) τ = 10 (red)

0.3
0 0.2 0.4 0.6 0.8 1
x, Length [normalized]

Fig. 11. Case 4: Effect of relaxation time on saturation profiles (Drainage, Oil-wet Rock).

Analysis of the Results

The first case study was primarily utilized to demonstrate the validity of the in-house simulator. The subsequent examples are of
primary importance and illustrate non-instantaneous capillary equilibrium effects during drainage and imbibitions in oil- and water-
wet rock. The results from Case 2, shown in Figs. 7 and 8, illustrate the dramatic effect of including the concept of relaxation time for
10 SPE 135564

fluid transport. The results indicate that the saturation front tends to move faster than the corresponding fronts obtained assuming
capillary equilibrium, thereby suggesting that the displacing fluid might have an earlier breakthrough. An analysis of the cumulative
water production (Fig. 8) also suggests that these faster breakthrough times might result in lesser recovery of the water originally in
situ.
Capillary non-equilibrium effects are seen to play a significant role even in imbibitions processes and the corresponding effects on
saturation front movement and cumulative production of the displaced are identical to those seen in the previous case study (Case 2).
Consequently, it is worthwhile to note that the relaxation time has a considerable impact on the cumulative production of the displaced
fluids (Figs. 8 and 10).
The results from Case 4 for drainage in a strongly oil-wet system (Fig. 11) also indicate similar effects. Consequently, one of the
findings of this study is to evaluate the impact of relaxation time on systems of mixed wettability such as those in shale gas rocks with
water-wet nanopores and strongly oil-wet organic material. Additionally, Eq. 10 suggests that for an imbibition process (for instance,
case 3), the effective saturation of the wetting phase has a value that is always larger than or identical to the corresponding saturation
calculated using capillary equilibrium. Therefore, under capillary non-equilibrium conditions, relative permeability of the wetting
phase is always larger whereas relative permeability of the non-wetting phase and capillary pressures are lower. This simple fact
immediately implies that the displacement is going to be affected by the formulation.

Conclusions
• The paper presents an approach to account for non-instantaneous capillary pressure equilibrium effects. A 1-dimensional reservoir
simulator was developed that incorporates these effects and utilized to study the impact of capillary pressure relaxation times on
fluid transport within a 1-dimensional coreflood.

• A preliminary study of the rock wettability is accomplished for two cases, strongly water-wet and strongly oil-wet rocks.
Saturation profiles within a 1-dimension coreflood are seen to be a function of the relaxation time in both water-wet and oil-wet
rocks although the respective profiles are not identical. Consequently, the wettability of the formation can significantly impact
fluid transport within the reservoir.

• Because shales tend to comprise of localized systems of mixed wettability, future work will focus on the development of a
methodology to incorporate these effects.

Nomenclature
τ = relaxation time
Seff = effective water saturation
k = permeability
L = sample length
D = sample diameter
μo = oil viscosity
μw = water viscosity
φ = porosity
Swi = Initial water saturation
v = volumetric influx
α = angle of inclination
ρ = density
Sw = water saturation
Pc = capillary pressure
t = time
P = pressure
kr = relative permeability
qinj = injection rate

References
1. Ambrose, R.J., Hartman, R.C., Diaz-Campos, M., Akkutlu, I.Y. and Sondergeld, C.H. 2010. New Pore-scale Considerations for Shale
Gas in Place Calculations. Paper SPE 131772-MS presented at the SPE Unconventional Gas Conference, Pittsburg, Pennsylvania,
USA, 23-25 February. doi 10.2118/131772-MS
2. Barenblatt, G.I., Patzek, T.W., and Silin, D.B. 2003. The Mathematical Model of Nonequlibrium Effects in Water-Oil Displacment.
SPE Journal , 8 (4): 409-416. SPE-87329-PA. doi: 10.2118/87329-PA.
3. Buckley, S.E., and Leverett, M.C. 1942. Mechanism of Fluid Displacement in Sands. Trans., AIME 146: 107-116. SPE-942107-G.
4. Cipolla, C.L., Carbo Ceramics, Lolon, E.P.,StrataGen Engineering, and Mayerhofer, M.J., Pinnacle – a Halliburton Service. 2009.
Reservoir Modeling and Production Evaluation in Shale-Gas Reservoirs. IPTC 13185, Doha, Qatar, 7-9 December. doi:
10.2523/13185-MS.
SPE 135564 11

5. Cipolla, C.L., Lolon, E.P., StrataGen Engineering, Erdle, J.C., and Rubin, B., CMG. 2009. Reservoir Modeling in Shale-Gas
Reservoirs. Paper SPE 125530-MS presented at the SPE Eastern Regional Meeting, Charleston, West Virginia, USA, 23-25 September.
doi: 10.2118/125530-MS.
6. Civan, F. 1996. Convenient Formulations For Immiscible Displacement In Porous Media. Paper SPE 36701-MS presented at the 1996
SPE Annual Technical Conference and Exhibition, Denver, Colorado, USA, 6-9 October. doi: 10.2118/36701-MS.
7. Civan, F. 2006. Including Nonequilibrium Relaxation in Models for Rapid Multiphase Flow in Wells. SPE Production & Operations,
21 (1): 98-106. SPE-90583-PA. doi: 10.2118/90583-PA.
8. Downar-Zapolski, P., Bilicki, Z., Bolle, L., and Franco, J. 1996. The Non-equilibrium Relaxation Model for One-dimensional Flashing
Liquid Flow. Int. J. Multiphase Flow 22 (3): 473-483.
9. Javadpour, F. 2009. Nanopores and Apparent Permeability of Gas Flow in Mudrocks (Shales and Siltstone). JCPT, 48 (8): 16-21,
August. doi: 10.2118/09-08-19-DA.
10. Kang, S.M., Fathi, E., Ambrose, R.J., Akkutlu, I.Y., and Sigal R.F. 2010. Carbon Dioxide Storage Capacity of Organic-rich Shales,
Paper SPE 134583 presented at the 2010 SPE Annual Technical Conference and Exhibition, Florence, Italy, 19-22 September.
11. Kulkarni, R., Watson, A.T., Nordtvedt, J.E., and Sylte, A. 1998. Two-phase Flow in Porous Media: Property Identification and Model
Validation. AIChE, 44 (11): 2337-2350. doi: 10.1002/aic.690441102.
12. McDougall, S.R., and Sorbie, K.S. 1995. The Impact of Wettability on Waterflooding: Pore-Scale Simulation. SPE Reservoir
Engineering 10 (3): 208-213. SPE-25271-PA. doi: 10.2118/25271-PA.
13. Michael, G., & Civan, F. 2008. Modeling Nonisothermal Rapid Multiphase Flow in Wells Under Nonequilibrium Conditions. SPE
Production & Operations 23 (2): 125-134. SPE-102231-PA. doi: 10.2118/102231-PA.
14. Quin, B. 2007. Numerical Study of Recovery Mechanisms in Tight Gas Recovery. In M.S. thesis. The University of Oklahoma,
Norman, Oklahoma, USA.
15. Silin, D., & Patzek, T. (2004). On Barenblatt's Model of Spontaneous Countercurrent Imbibition. Transport in Porous Media 54 (3):
297-322. doi: 10.1023/B:TIPM.0000003678.85526.b1.
16. Sondergeld, C.H., Ambrose, R.J., Moncrief, J., Micro-Structural Studies of Gas Shales. Paper SPE 131771-MS presented at the SPE
Unconventional Gas Conference, Pittsburg, Pennsylvania, USA, 23-25 February. doi: 10.2118-131771-MS.
17. Tyagi, M., Jenny, P., & Tchelepi, H. A. 2008. Stochastic modeling of non-equilibrium multiphase flow in porous media. PAMM 7 (1):
2100025-2100026. doi: 10.1002/pamm.200700219.
18. Welge, H.J. 1952. A Simplified Method for Computing Oil Recovery by Gas or Water Drive. Trans., AIME 195: 91-98. SPE-124-G.