Petroleum Science and Technology

ISSN: 1091-6466 (Print) 1532-2459 (Online) Journal homepage: http://www.tandfonline.com/loi/lpet20

Adomian Decomposition of Buckley-Leverett

Equation with Capillary Effects

S. Mustafiz , S. H. Mousavizadegan & M. R. Islam

To cite this article: S. Mustafiz , S. H. Mousavizadegan & M. R. Islam (2008) Adomian

Decomposition of Buckley-Leverett Equation with Capillary Effects, Petroleum Science and
Technology, 26:15, 1796-1810, DOI: 10.1080/10916460701426049

To link to this article: http://dx.doi.org/10.1080/10916460701426049

Published online: 10 Oct 2008.

Submit your article to this journal

Article views: 83

View related articles

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=lpet20

Download by: [Memorial University of Newfoundland] Date: 05 June 2016, At: 17:20
 Petroleum Science and Technology, 26:17961810, 2008
Copyright Taylor & Francis Group, LLC
ISSN: 1091-6466 print/1532-2459 online
DOI: 10.1080/10916460701426049

Adomian Decomposition of Buckley-Leverett

Equation with Capillary Effects
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

S. Mustafiz,1;2 S. H. Mousavizadegan,1 and M. R. Islam1

1 Department of Civil and Resource Engineering, Dalhousie University,
Halifax, Nova Scotia, Canada
2 Currently with Alberta Research Council, Edmonton, Alberta, Canada

Abstract: Petroleum reservoir engineering problems are known to be inherently

nonlinear. Consequently, solutions to the complete multiphase flow equations have
been principally attempted with numerical methods. However, simplified forms of the
problem have been solved some 60 years ago, when the Buckley-Leverett formulation
was introduced. Ever since that pioneer work, which neglected the capillary term, this
formulation has been widely accepted in the petroleum industry. By using the method
of characteristic, the multiphase one-dimensional fluid flow was solved. However,
the resulting solution was a triple-valued one for a significant region. For decades,
the existence of multiple solutions was considered to be the result of nonlinearity.
Buckley and Leverett introduced shock utilizing the concept of material balance, and,
two decades later, when numerical solutions were possible, it was discovered that
the triple-value problem disappeared if the complete flow equation, including the
capillary pressure form, is solved. Numerical methods, however, are not free from
linearization. In fact, every numerical solution imposes linearization at some point of
the solution scheme. Therefore, a numerical technique cannot be used to definitively
state the origin of multiple solutions. In this article, a semi-analytical technique,
the Adomian decomposition method (ADM), capable of solving nonlinear partial
differential equations without any linearizing assumptions, is used to unravel the true
nature of the one-dimensional, two-phase flow. Results show that the Buckley-Leverett
shock is neither necessary nor accurately portrayed in the displacement process. By
using the ADM, the solution profile observed through numerous experimental studies
was rediscovered. This article opens up an opportunity to seek approximate but close
to exact solutions to the multiphase flow problems in porous media.

Keywords: Adomian decomposition, Buckley-Leverett solution, capillary, nonlinear-

ity, waterflood

Address correspondence to S. Mustafiz, Heavy Oil and Oil Sands, Alberta

Research Council, 250 Karl Clark Road, Edmonton, AB, T6N 1E4, Canada. E-mail:
shabbir.mustafiz@arc.ab.ca

1796
 Adomian Decomposition of Buckley-Leverett Equation 1797

INTRODUCTION

Since Adomian (1984) proposed his decomposition technique, the Ado-

mian decomposition method (ADM) has gained significant interest among
researchers, particularly in the fields of physics and mathematics. They
applied this method to many deterministic and stochastic problems (Adomian,
1986, 1994; Eugene, 1993). In this method, the governing equation is
transformed into a recursive relationship, and the solution appears in the
form of a power series. The ADM has emerged as an alternative method to
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

solve various mathematical models, including algebraic, differential, integral,

integro-differential, partial differential equations (PDEs) and systems, higher-
order ordinary differential equations, and others.
Guellal and Cherruault (1995) utilized the method to solve an elliptical
boundary value problem with an auxiliary condition. Laffez and Abbaoui
(1996) used it in modeling thermal exchanges associated with drilling wells.
The application of this method is also noticeable in medical research, in
which the decomposition technique is used to solve differential systems of
equations in modeling of HIV immune dynamics (Adjedj, 1999). Wazwaz
(2001) and Wazwaz and El-Sayed (2001) reported that the ADM could be
useful in solving problems without considering linearization, perturbation,
or unjustified assumptions that may alter the nature of the problem under
investigation. Also, there have been suggestions that this method can be
more advantageous over numerical methods by providing analytic, verifiable,
rapidly convergent approximations, which add insight into the character and
the behavior of the solution as obtainable in the closed-form solution (El-
Sayed and Abdel-Aziz, 2003). In particular, they noticed the strength of the
ADM in handling nonlinear problems in terms of rapid convergence. Biazar
and Ebrahimi (2005) also expressed a similar notion about rapid convergence
and further added the advantage of the technique in terms of considerable sav-
ings in computation time when they attempted to solve hyperbolic equations.
Even though petroleum problems are some of the most intriguing can-
didates for the ADM, there have been sparse attempts to utilize the method
in the petroleum problems. Only recently, Mustafiz et al. (2005) reported
the decomposition of the non-Darcy flow equations in porous media. In a
different paper, Mustafiz and Islam (2005) transformed the diffusivity equa-
tions in well-test applications into canonical forms. There is some success
with Adomian method in modeling drilling activities, such as perforation
by drilling (Rahaman et al., 2007) specified drilling operations (Biazar and
Islam, in press).
The Buckley-Leverett (1942) analysis is considered to be the first pioneer
work in the study of linear displacement of a fluid by another fluid. The
solution of their displacement study on two-phase fluid excluded the effect
of capillary and gave multiple results for saturation at a given position.
Realizing the fact that such co-existence of multiple saturation values is
physically unrealistic, Buckley and Leverett (1942) used the fundamental
 1798 S. Mustafiz et al.

concept of material balance to explain the shock. However, the lack of

theoretical justification of shock was a major constraint in understanding
the displacement phenomenon more vividly than before, and would not be
reevaluated until the next significant work published by Holmgren and Morse
(1951). They utilized the Buckley-Leverett theory to calculate the average
water saturation at breakthrough and explained dispersion as a consequence
of capillary effects. Welge (1952) continued the effort in finding the average
saturation at water breakthrough in an oil reservoir. He also found that the
method of tangent construction by Terwilliger et al. (1951) was equivalent
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

to the shock proposed by Buckley and Leverett.

To solve the displacement equation, including capillary as well as gravity,
Fayers and Sheldon (1959) attempted a Lagrangian approach. They, however,
did not succeed to determine the time required to obtain a particular satu-
ration, which later was explained by Bentsen (1978), revealing the fact that
the distance traveled by zero saturation is governed by a separate equation.
Bentsen also noted that, at slower injection rates, the input boundary condition
of constant normalized saturation that Fayers and Sheldon used was incorrect
in formulation. Also, there have been numerical investigations in the past to
solve the displacement equation. Hovanessian and Fayers (1961) were able
to avoid profiles of multiple valued saturations by considering the capillary
pressure.
The Adomian decomposition method is applied to solve the nonlinear
Buckley-Leverett equation. The solution for the water saturation is expressed
in a series form. The base element of the series solution is obtained using
the solution of the linear part of the Buckley-Leverett equation without the
effect of the capillary pressure using the characteristic method. The other
elements of the series solution are obtained recursively using the Adomian
polynomial. The modification of the ADM in selection of the base element
of the series solution makes the ADM feasible in solution of the nonlinear
Buckley-Leverett equation. The computation is carried out for a reservoir of
certain properties and initial conditions.

GOVERNING EQUATIONS

The Buckley-Leverett equation is given by

@Sw q @fw @Sw

C D 0; (1)
@t A @Sw @x

where fw is expressed in the form

0 1
  
B 1 C 1 C Akkr o @Pc
C
fw D B .w o /g sin : (2)
@ kr o w A qo @x
1C
kr w o
 Adomian Decomposition of Buckley-Leverett Equation 1799

This equation indicates that the fractional flow rate of water depends on
reservoir characteristics, water injection rate, viscosity, and direction of flow.
The effect of capillary pressure, Pc , which appears in the fractional flow
equation, on saturation profiles is important, since these profiles affect the
ultimate economic oil recovery (Bentsen, 1978). The ratio of effective per-
meability to viscosity is defined as the mobility, which is shown for water
and oil, respectively.

kkr w kkr o
w D and o D : (3)
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

w o

The mobility ratio of oil to water is defined by

o kr ow
M D D : (4)
w kr w o

The assumptions associated with the Buckley-Leverett equation, according to

Ertekin et al. (2001) are:

 Oil and water phases are assumed incompressible;

 The porous medium is assumed incompressible, which implies that porosity
is constant;
 Injection and production are taken care of by means of boundary condi-
tions, which indicate that there are no external sink or source in the porous
medium;
 The cross-sectional area that is open to flow is constant;
 The saturation-constraint equation for two-phase flow is valid; and
 The fractional flow of water is dependent on water saturation only.

The expression of the fractional flow rate of water in Eq. (2) suggests
that Eq. (1) is a nonlinear differential equation, and it is due to the effect of
capillary pressure. In the simplest case of horizontal flow and neglecting the
effects of capillary pressure variation along the reservoir, the expression for
fw in Eq. (2) is simplified to a linear differential equation, which is given by

1
fw D : (5)
1CM

The water saturation distribution along the reservoir can be found at

different time steps by knowing the injection flow rate, the initial water
saturation distribution, and the variation of fractional water flow rate. The
water is normally the wetting fluid in the water-oil two phase flow systems.
The relative permeability of the water and oil for a specific reservoir are ob-
tained with the drainage and imbibitions process on the core in the laboratory.
However, in this paper, the variation of relative permeability to water and oil
as a function of water saturation are obtained using the following empirical
 1800 S. Mustafiz et al.

relationships, respectively.
n1
kr w D 1 Swn ; (6.1)

kr o D 2 .1 Swn /n2 ; (6.2)

where the normalized water saturation, Swn , is defined as

Sw Swi
Swn D : (7)
1 Swi Sor
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

If the effects of capillary pressure are included in a horizontal reservoir, Eq.

(2) takes the form:
 
1 Akkr o @Pc
fw D 1C : (8)
kr o w qo @x
1C
kr w o
Capillary pressure at any point is directly related to the mean curvature of
the interface, which, in turn, is a function of saturation (Leverett, 1941).
Therefore, it can be safely assumed that capillary pressure is only a function
of water saturation. By applying the chain rule,
@Pc @Pc @Sw
Pc D f .Sw / ! D ; (9)
@x @Sw @x
and by incorporating Eqs. (8) and (9) in Eq. (1), the following partial differ-
ential equation is obtained:
0 1
@Sw q @ B 1 C @Sw
C B C
@t A @Sw @ kr ow A @x
1C
kr w o
0 1
 2
kkr o @ B 1 C @Pc @Sw
C B C
o @Sw @ kr ow A @Sw @x
1C
kr w o
0 1
C @kr o @Pc @Sw 2
 
k B 1
C B C
o  @ kr ow A @Sw @Sw @x
1C
kr w o
0 1
C @2 Pc @Sw 2
 
kkr o B 1
C B C D 0: (10)
o  @ kr ow A @Sw2 @x
1C
kr w o
 Adomian Decomposition of Buckley-Leverett Equation 1801

Equation 10 is a nonlinear partial differential equation, and the nonlinearity

arises because of the inclusion of capillary pressure in it.

Adomian Decomposition of Buckley-Leverett Equation

In the Adomian decomposition method, the solution of a given problem is

considered as a series solution. Therefore, the water saturation is expressed
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

as
1
X
Sw .x; t/ D Swn : (11)
nD0

If a functional equation is taken into account for water saturation, it can be

written in the form:

1
X 1
X
Swn D f .x; t/ C An ; (12)
nD0 nD0

where f .x; t/ is a given function and An .Sw0 ; Sw1 ; Sw2 ; : : : : ; Swn / or sim-
ply An s are called Adomian polynomials. The Adomian polynomials are
expressed as
" 1
!#
1 dn X
i
An D N  Swn .x; t/i (13)
n dn i D0 D0

where N is an operator and  is a parameter. The elements of the series

solution for Swn .x; t/ are obtained recursively by

Sw0 .x; t/ D f .x; t/;

Sw1 .x; t/ D A0 ;

::: (14)

Swk .x; t/ D Ak 1

::::

By this arrangement, the linear and nonlinear part of the functional Eq. (12)
is replaced by a known function using the recursive Eq. (14). By integrating
Eq. (10) with respect to t
 1802 S. Mustafiz et al.
2 0 1 3
t 6 q @ B 1 C @Sw 7
Z
Sw D Sw .x; 0/ 6 B C 7 dt
0
4 A @Sw @ kr ow A @x 5
1C
kr w o
2 0 1
t  2
6 kkr o @ B 1 C @Pc @Sw
Z
6 B C
0
4 o @Sw @ kr ow A @Sw @x
1C
kr w o
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

0 1
 2
k B 1 C @kr o @Pc @Sw
C B C
o  @ kr o w A @Sw @Sw @x
1C
kr w o
0 1 3
2
C @2 Pc

kkr o B 1 @Sw 7
7 dt:
C B C (15)
o  @ kr ow A @Sw2 @x 5
1C
kr w o

Comparing Eq. (15) with Eq. (12), and taking into account Eq. (14), the
elements of the water saturation series are obtained:
2 0 1 3
t 6 q @ B 1 C @Sw0 7
Z
Sw0 .x; t/ D Sw .0; x/ 6 B C 7 dt
0
4 A @Sw0 @ kr o w A @x 5
1C
kr w o
2 0 1
t 6 kkr o @ B 1 C @Pc
Z
Sw1 .x; t/ D 6 B C
0
4 o @Sw0 @ kr ow A @Sw0
1C
kr w o
0 1
k B 1 C @kr o @Pc
C B C
o  @ kr ow A @Sw0 @Sw0
1C
kr w o
0 1 3
C @2 Pc 7 @Sw0 2
 
kkr o B 1
C B C
2 5
7 dt
o  @ kr o w A @Sw0 @x
1C
kr w o
 Adomian Decomposition of Buckley-Leverett Equation 1803
2 0 1
t
d 6 kkr o @ 1 @Pc
Z
B C
Sw2 .x; t/ D 6 B C
d 0
4 o  @.Sw0 C Sw1 / @ kr o w @.Sw0 C Sw1 /
A
1C
kr w o
0 1
k B 1 C @kr o @Pc
C B C
o  @kr ow @.Sw0 C Sw1 / @.Sw0 C Sw1 /
A
1C
kr w o
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

0 1 3
@2 Pc 7 @.Sw0 C Sw1 / 2
 
kkr o B 1 C
C B C 7 dt
o  @ kr o w A @.Sw0 C Sw1 /2 5 @x
1C
kr w o
2 0 1
t6
1 d2
Z
6 kkr o @ 1
B C
Sw3 .x; t/ D 2
B C
2 d 0 4 o  @.Sw0 C Sw1 C Sw2 / @ k 
ro w
A
1C
kr w o

@Pc

@.Sw0 C Sw1 C Sw2 /
0 1
k B 1 C @kr o @Pc
C B C
o  @ kr o w A @.Sw0 C Sw1 C Sw2 / @.Sw0 C Sw1 C Sw2 /
1C
kr w o
0 1 3
kkr o B 1 C @2 Pc 7
C B C 7
o  @ kr o w @.Sw0 C Sw1 C Sw2 / 5
A 2
1C
kr w o
2
@.Sw0 C Sw1 C 2 Sw2 /

 dt
@x

::: ::: ::: ::: ::: ::: ::: ::: (16)

RESULTS AND DISCUSSION

The method is applied to a given reservoir with the initial water saturation
of 0.18. This value is corresponding to the irreducible water saturation of
the reservoir. Water is injected into the reservoir with a linear flow rate of
1 ft/day. The oil and water viscosities are 1.73 cp and 0.52 cp, respectively.
 1804 S. Mustafiz et al.
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

Figure 1. The variation of the water and oil relative permeability, the fractional water
flow rate, and the differentiation of fractional water flow rate with the water saturation.

The flow of the displaced phase (oil) ceases at Soc = 0.1. The porosity of the
medium is 0.25, with an absolute permeability of k = 10 md.
The normalized water saturation and the water and oil relative perme-
ability are obtained using
Sw 0:18 4
Swn D ; kr w D 0:59439Swn ; kr o D .1 Swn /2 : (17)
1 0:1 0:18
A typical plot of variation of relative permeability to water, kr w , rel-
ative permeability to oil, kr o, fractional flow curve, fw , and its derivative,
dfw =dSw , are shown in Figure 1. The capillary pressure data are also known
and are shown in Table 1.
The solution for the first base element of the series solution of the water
saturation, Sw0 , is obtained through the solution of the first equation in Eq.
(16). It can be written, by integrating from both sides of the first equation in
Eq. (16) and using Eq. (4), that
 
@Sw0 q @ 1 @Sw0
C D 0: (18)
@t A @Sw0 1 C M @x

Table 1. Capillary pressure data

Swn Pc , [atm] Swn Pc , [atm] Swn Pc , [atm] Swn Pc , [atm]

0.00 3.9921 0.11 1.2036 0.30 0.3600 0.65 0.0550

0.01 3.5853 0.15 0.8745 0.36 0.2699 0.72 0.0350
0.02 3.1987 0.18 0.7010 0.42 0.1980 0.87 0.0100
0.05 2.2577 0.21 0.5709 0.48 0.1450 0.95 0.0027
0.08 1.6209 0.25 0.4592 0.56 0.0920 1.00 0.0000
 Adomian Decomposition of Buckley-Leverett Equation 1805

This equation suggests that Sw0 is constant along a direction that is called
characteristic direction. The characteristic direction can be obtained by
   
dx q @ 1
D ; (19)
dt Sw0 A @Sw0 1CM t

that is the Buckley-Leverett frontal advance equation. Integrating respect to

time, the distribution of Sw0 is found in the form of
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

 
qt @ 1
x.t; Sw0 / D : (20)
A @Sw0 1 C M t

The solution of Eq. (20) gives the variation of the Sw0 along the reservoir at
certain time. The distribution of Sw0 along the reservoir is obtained based on
the definition of the mobility ratio in Eq. (4) and the relations for the relative
permeability of water and oil in Eq. (17). It corresponds to the variation of
the water saturation when the effect of the capillary pressure is ignored. The
computations are carried out at different time of t D 0.5, 1, 2, 5, and 10 [days].
The solution for the water saturation without the capillary pressure effect
shows the unrealistic physical situation that Buckley-Leverett mentioned in

Figure 2. The water saturation distribution with and without the effect of capillary
pressure using ADM.
 1806 S. Mustafiz et al.

their pioneer paper, with multiple saturations at each distance (x-position),

as given partly in Figure 2. In order to avoid multiple saturation values at a
particular distance, a saturation discontinuity at a distance, xf is generally
created in such a way that the areas ahead of the front and below the curve
are equal to each other.
The other elements of the series solution for the water are obtained recur-
sively using Eq. (16) and the solution for Sw0 . The solution for Sw1 is obtained
using the solution of Sw0 at a certain point for a given time. To induce the
nonlinear dependence of capillary on saturation, during decomposition, the
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

capillary pressure and its derivatives are approximated by using the cubic
spline, as shown in Figure 3. The interpolating splines are preferred over
interpolating polynomials as they do not suffer from oscillations between the
knots and are smoother and more realistic than linear splines (Ertekin et al.,
2001). Such technique is also used to observe the effects of linearization in
pressure in reservoir flow equations (Mustafiz et al., 2006). The solutions
for Sw2 and Sw3 are obtained recursively using Eq. (16). The computations
show that the series solution converges very fast and it is not necessary to go
further than Sw3 .
The distribution of the water saturation along the reservoir is given in
Figure 2. The results of the water saturation without the effect of the capillary
pressure are also depicted in Figure 2 for the sake of comparison. This figure

Figure 3. The capillary pressure variation and its first and second derivatives as a
function of the water saturation.
 Adomian Decomposition of Buckley-Leverett Equation 1807

shows that by considering the effects of capillary pressure, it is possible

to avoid unrealistic multiple saturation values. Moreover, the decomposition
approach shows notable prediction of saturation profiles along the length. The
gradual and mild changes observed in the saturation profile here are perhaps
less severe than the conventional prediction of shock fronts.

CONCLUSIONS
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

To unravel the lure of finding a justifiable answer to the shock front, which
is often less severe in the experiments than in the theoretical postulations, the
Adomian decomposition method is investigated to solve the Buckley-Leverett
equation. The solution by the ADM also assures that the nonlinear dependence
of capillary pressure on saturation is maintained. The term, Sw0 .x; t/; is
prepared through the method of characteristics, which subsequently becomes
useful for the purpose of evaluation of the other elements in the water
saturation series. The saturation at a point and at a given time is evaluated
through recursion. The term Sw1 .x; t/ is calculated using the solution of
Sw0 .x; t/ at any distance and time. The solutions for Sw2 .x; t/ and Sw3 .x; t/
are obtained using the results of Sw1 .x; t/ and Sw2 .x; t/; respectively. The
computation is done up to the term Sw3 .x; t/; as it is found that the series
solution converges very quickly.
The most significant observation is that, by considering the effects of
capillary pressure, it is possible to avoid unrealistic multiple saturation values.
Moreover, the saturation profiles along the length are noted with gradual and
mild changes as predicted by the decomposition approach. Perhaps the slope
is less severe than predicted by the conventional finite difference approach;
a study remains for the future.

ACKNOWLEDGMENTS

The authors gratefully acknowledge the research grant received through the
Atlantic Innovation Fund (AIF), Canada. Mustafiz would also like to thank
the Killam Foundation for its financial support.

REFERENCES

Adjedj, B. (1999). Application of the decomposition method to the under-

standing of HIV immune dynamics. Kybernetes 28:271283.
Adomian, G. (1984). A new approach to nonlinear partial differential equa-
tions. J. Math. Anal. Appl. 102:420434.
 1808 S. Mustafiz et al.

Adomian, G. (1986). Nonlinear stochastic operator equations. San Francisco:

Academic Press.
Adomian, G. (1994). Solving frontier problems of physics: The decomposi-
tion method. Boston: Kluwer.
Bentsen, R. G. (1978). Conditions under which the capillary term may be
neglected. J. Can. Pet. Technol. 17:2530.
Biazar, J., and Ebrahimi, H. (2005). An approximation to the solution of
hyperbolic equations by Adomian decomposition method and comparison
with characteristics method. Appl. Math. Comput. 163:633638.
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

Biazar, J., and Islam, M. R. (2007). The Adomian decomposition method

for the solution of transient energy equation in rocks subjected to laser
irradiation. Iranian J. Sci. Tech. Transaction A 30(A2):201212.
Buckley, S. E., and Leverett, M. C. (1942). Mechanism of fluid displacement
in sands. Trans. AIME 146:187196.
El-Sayed, S. M., and Abdel-Aziz, M. R. (2003). A comparison of Adomians
decomposition method and wavelet-Galerkin method for solving integro-
differential equations. Appl. Math. Comput. 136:151159.
Ertekin, T., Abou-Kassem, J. H., and King, G. R. (2001). Basic Applied
Reservoir Simulation. SPE textbook series, vol. 7. Richardson, TX: SPE.
Eugene, Y. (1993). Application of the decomposition method to the solution
of the reaction-convection-diffusion equation, Appl. Math. Comput. 56:1
27.
Fayers, F. J., and Sheldon, J. W. (1959). The effect of capillary pressure
and gravity on two-phase fluid flow in a porous medium. Trans. AIME
216:147155.
Guellal, S., and Cherruault, Y. (1995). Application of the decomposition
method to identify the distributed parameters of an elliptical equation.
Math. Comput. Model. 21:5155.
Holmgren, C. R., and Morse, R. A. (1951). Effect of free gas saturation on
oil recovery by waterflooding, Trans. AIME 192:135140.
Hovanessian, S. A., and Fayers, F. J. (1961). Linear water flood with gravity
and capillary effects. SPEJ March:3236.
Laffez, P., and Abbaoui, K. (1996). Modeling of the thermic exchanges
during a drilling: Resolution with Adomians decomposition method.
Math. Comput. Model. 23:1114.
Leverett, M. C. (1941). Capillary behaviour in porous solids. Trans AIME
142.
Mustafiz, S., and Islam, M. R. (2005). Adomian decomposition of two-phase,
two-dimensional non-linear PDEs as applied in well testing. Proceedings
of the 4th International Conference on Computational Heat and Mass
Transfer, Paris-Cachan, May 1720.
Mustafiz, S., Biazar, J., and Islam, M. R. (2005). An Adomian decomposition
solution to the modified Brinkman model (MBM) for a two-dimensional,
one-phase flow of petroleum fluids. Proceedings of the CSCE 33rd An-
nual Conference, Toronto, ON, Canada, June 24.
 Adomian Decomposition of Buckley-Leverett Equation 1809

Mustafiz, S., Mousavizadegan, S. H., and Islam, M. R. (2008). The effects

of linearization on solutions of reservoir engineering problems. Pet. Sci.
Technol. 26:12241246.
Rahman, M. A., Mustafiz, S., Biazar, J., Koksal, M., and Islam, M. R. (2007).
Investigation of a novel perforation technique in petroleum wells
perforation by drilling. J. Franklin Inst. 344:777789.
Terwilliger, P. L., Wilsey, L. E., Hall, H. N., Bridges, P. M., and Morse, R. A.
(1951). An experimental and theoretical investigation of gravity drainage
performance. Trans. AIME 192:285296.
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

Wazwaz, A. M. (2001). A new algorithm for calculating Adomian polyno-

mials for nonlinear operators. Appl. Math. Comput. 111:3351.
Wazwaz, A. M., and El-Sayed, S. M. (2001). A new modification of Adomian
decomposition method for linear and nonlinear operators. Appl. Math.
Comput. 122:393405.
Welge, H. J. (1952). A simplified method for computing oil recovery by gas
or water drive. Petrol. Trans. AIME 195:9198.

NOMENCLATURE

A cross-sectional area in the x-direction

An Adomian polynomial
fw fractional flow of water
g acceleration of gravity
k absolute permeability
kr o relative permeability to oil
kr w relative permeability to water
M mobility ratio of oil to water
N operator in the expression (Eq. (14)) for Adomian polynomial
n1 empirical constant in the relationship between relative permeability
to water and normalized water saturation
n2 empirical constant in the relationship between relative permeability
to oil and normalized water saturation
Pc capillary pressure
q total fluid flow rate
Sor residual oil saturation
Sw water saturation
Swi irreducible water saturation
Swn normalized water saturation
t time
x distance
xf location of the front of saturation discontinuity in the x-direction
angle of inclination of reservoir (positive for the x axis rotated in
the counter clockwise direction from the horizon)
 1810 S. Mustafiz et al.

1 constant coefficient in the relationship between relative permeability

to water and normalized water saturation
2 constant coefficient in the relationship between relative permeability
to oil and normalized water saturation
 parameter in the expression (Eq. (14)) of Adomian polynomial
o mobility of oil
w mobility of water
o viscosity of oil
w viscosity of water
Downloaded by [Memorial University of Newfoundland] at 17:20 05 June 2016

o density of oil
w density of water
' porosity