Está en la página 1de 13

52

Iranian Journal of Chemical Engineering


Jol. 3, No. 1 (hinter), 2006, IAChE
A Sequential Formulation for Compositional Reservoir Simulation
Using Peng Robinson Equation of State
A. Shahrabadi

, B. Dabir
Chemical Engineering Department, AmirKabir Universitv of Technologv, Tehran, Iran.
Abstract
In this paper, a simplified formulation for a compositional reservoir simulator is
presented. These tvpe of simulators are used when interphase mass transfer depends on
phase composition as well as pressure. The procedure for solving compositional model
equations is completelv described. The Peng Robinson equation of state is used for
preparing a compositional thermodvnamic program for equilibrium calculation,
propertv estimation and pseudo component determination. Another purpose of this
paper is to prepare an experimental apparatus for the displacement of oil bv gas
infection. In each test, oil recoverv as a function of infected pore volume was measured.
The application of the developed simulator to simulate the results of the oil recoverv
from slim tube experiments is also presented. Finallv, the model was run for a 2-D
reservoir. Acceptable trends were obtained from the model predictions.
Keywords: Compositional, Reservoir simulator, Slim tube, Equation of state, Peng
Robinson
- - E-mail: ashahrabadiyahoo.com
Introduction
Numerical reservoir simulation is a vital tool
in the petroleum industry. This technique can
be used in planning the exploration oI new
reservoirs, Iorecasting the behavior oI old
Iields, to make single well studies, and in
researching methods to improve hydrocarbon
recovery. At the beginning, the numerical
models were built by considering three pha-
ses in the reservoir (oil, gas and water), but
they were unable to treat the compositional
variations occurring in each phase. These
models are reIerred to as black oil models. In
this case, phase behavior is simply repre-
sented by B
o
(oil Iormation volume Iactor)
and R
s
(solution gas oil ratio), which are only
Iunctions oI pressure. Whenever it is inap-
propriate to use a black oil model, the oil and
gas must be described by more than two
pseudo-components. In this case, phase
behavior is represented by an equation oI
state and phase equilibrium relations, requi-
ring Ilash calculations|1|. In this case com-
positional reservoir simulators must be used.
Compositional reservoir simulators are im-
portant tools Ior predicting the perIormance
oI oil recovery methods whenever oil and gas
undergo vigorous mass transIer during the
recovery process. These processes include
nitrogen injection into a gas condensate or
volatile oil reservoir during the primary
production period, and CO
2
or enriched gas
injection Ior enhanced oil recovery. The
design oI these processes requires an accurate
Archive of SID
www.SID.ir
Shahrabadi, Dabir
Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1 53
prediction oI the vapor liquid equilibrium
between oil and gas. Generally this is done
by using an equation oI state. Recently cubic
equations oI state such as Peng Robinson and
Soave Redlich Kwong appear to be more
popular Ior the correlation oI Iluid properties.
Eor compositional simulation two basic so-
lution methods have been proposed. These
methods are the 'Newton-Raphson and
'Non Newton-Raphson methods. In the
Newton-Raphson method, the iterative tech-
nique speciIies how the pressure equation is
Iormed. In the Non Newton-Raphson me-
thod, the composition dependence oI certain
terms is neglected to Iorm the pressure
equation. Erom the Iormulation point oI
view, several compositional Iormulations
have been published in the literature. They
may be classiIied as IMPECS (Implicit in
Pressure, Explicit in Composition and Satu-
ration), semi implicit or IMPSEC (Implicit in
Pressure and Saturation, Explicit in Com-
position), sequential type and EIM (Eully
Implicit Method) Iormulation.
The Iirst compositional simulators oI general
applications were developed by Kazemi et
al.|2| and by Eussel and Eussel|3|. Young
and Stephenson|4| classiIied IMPECS type
into the Non-Newton and Newton method.
The Non-Newton approach was originally
proposed by Nelon|5|, other versions were
later presented by some authors. These
methods diIIer in the manner in which the
pressure equation is Iormed. The model
presented by Acs et al.|6| is a Non-Newton
Raphson method in which the pressure
equation includes the compositional eIIects.
Watts|7|, based on the ideas oI Acs et al.,
presented another approach Ior the sequential
implicit method to solve the compositional
Ilow diIIerence equations. An attempt was
made to combine the advantages oI IMPECS,
while retaining some oI the stability charac-
teristics oI the implicit method. Quandalle
and Savary|8| presented another method. Eor
multiphase Ilow, the grid block equations are
reduced to three equations with three primary
unknowns, P (pressure), S
o
(oil saturation)
and S
w
(water saturation). Having these para-
meters, the subset oI the thermodynamic
equation is arbitrarily decoupled and used.
Celso|9| presented a diIIerent approach that
was regarded as a semi implicit method. In
his method, the Newton iteration was applied
to the compositional Ilow equations that
result aIter considering, explicitly, in Ilow
terms, one iteration behind. Wang et al.|10|
presented a Iully implicit compositional
simulator Ior large scale reservoir simulation.
Their simulator uses a multi block, domain
decomposition approach. Recently, composi-
tional stream line simulators Ior the as-
sessment oI miscible/near miscible gas in-
jecttion perIormance are widely used. These
simulators have signiIicant potential to ac-
commodate the requirement Ior accurate and
reliable production Iorecasts. These require-
ments include high resolution descriptions oI
permeability, heterogeneity, and an appro-
priate representation oI phase behavior,
including a suIIicient number oI components
in the equation oI state.|11|. Numerous
authors have contributed to the development
oI stream line simulators|12-16|. This Paper
describes an iterative sequential composi-
tional Iormulation. Basically it uses the ideas
oI Nghiem et al.|17|, but in this work the
implicit transmiscibilities have been used.
Compositional model equations
The starting point Ior the compositional
Iormulation is the molar continuity equation
Ior any component i and water:
, 1, 2,...,
N
i
i n
c
i
t
6
c
= =
c
(1)
1
1
N
n
c
n
t
c
6
c
+
=
+
c
(2)
i
N denotes the moles oI component i per unit
Archive of SID
www.SID.ir
A Sequential Formulation for Compositional Reservoir Simulation Using Peng Robinson Equation of State
54 Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1
oI reservoir volume, and
1
c
n
N
+
is the moles
oI water per unit oI reservoir volume related
to the phase molar densities, saturations and
composition as Iollows:

N S x S v
o o g g
i i i
1 = +
1
c
n w w
N S 1
+
=

1 i nc
6 6
+
is the rate oI the accumulation oI
component i (water), equal to the divergence
oI the Ilux oI component i plus the rate oI
injection or production. Thus:
rg
ro
i o i o g i g i
o g
kk
kk
x v q 6
2 2
(
= V Vu + Vu +
(
(

1
c
rw
n w w w
w
kk
q 6
2
+
(
= V Vu +
(

In the above equations:
o o o
P h Vu = V V (7)
g o cog g
P P h Vu = V + V V (8)
w o cow w
P P h Vu = V V V (9)
1
, , ,
f f
c
g
C f o g w
g
7 = = (10)
The molar continuity equation Ior the hy-
drocarbon system is obtained by summing
equation 1 over n
c
hydrocarbon components:
N
t
6
c
=
c
In this equation:

N S S
o o g g
1 + =
rg
ro
o o g g t
o g
kk
kk
q 6
2 2
(
= V Vu + Vu +
(
(

1
c
n
t i
i
q q
=
=

Multiplying Equations 1,2 and 11 by grid


block volume (
b
J ) and writing in Iinite
diIIerence Iorm:


1 1 1 1
1 1 1
n n n n n n b
i i o i o o
n n n n n
g i o cog o i b
J
N N T x P h
t
T v P P h q J

+ + + +
+ + +
= A A A
A
+A A + A A +
(15)

1 1 1 n n n n n n b
w w w o cow w w b
J
N N T P P h q J
t

+ + +
= A A A A +
A
(16)


1 1 1
1 1
n n n n n b
o o o
n n n n
g o cog o t b
J
N N T P h
t
T P P h q J

+ + +
+ +
= A A A
A
+A A + A A +
(17)
In the above equations:
2
, , ,
rf
f
f t
kk
A
T C f o g w
L 2
= =
A
(18)
(3)
(4)
(5)
(6)
(11)
(12)
(13)
(14)
Archive of SID
www.SID.ir
Shahrabadi, Dabir
Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1 55
C
1
and C
2
are conversion constants that
depend on the units involved.
It is assumed that gas and oil are in phase
equilibrium at the end oI each time step. This
means that the interphase thermodynamic
exchange in the reservoir is rapid compared
with Iluid Ilow. The condition Ior ther-
modynamic equilibrium is established by the
equality oI the oil and gas phase chemical
potential or Iugacity Ior each component:
( , , ) , ,
o g
i i i i
f T P x f T P v = (19)
Erom a material balance on oil and gas
phases, the Iollowing equation is obtained:
, 1, 2,...,
i i i c
F: Lx Jv i n = + = (20)
The Iollowing equations are used Ior solving
equation 19:
1
1 0
c
n
i
i
x
=
=

1
1 0
c
n
i
i
v
=
=


1
1
0
1 ( 1)
c
n
i i
i
i
: K
J K
=

=
+

(23)
The deIinition oI saturation gives the con-
straint equation:
1
o g w
S S S + + = (24)
Each grid block has its own set oI the above
equations. The basic idea used in this work is
to reduce the number oI unknowns into one
unknown Ior each grid block in terms oI grid
block pressure.
Construction of pressure equation
Adding equations 16 and 17 results in the
pressure equation:


1
1 1
1 1 1 1
n n
n n n b
w w o o o
n n n n n n n
g o cog g w o cog
J
N N N N T P h
t
T P P h T P P

+
+ +
+ + + +
(
+ + = A A A

A
+A A +A A +A A +A

n
w b t w
h J q q A + +
(25)
The Iollowing approximations apply to the
Ilow terms:
1
, , ,
n l
f f
T T f o g w
+
= = (26)
1 n l
i i
x x
+
= (27)
1 n l
i i
v v
+
= (28)
l denotes the last iteration. The pressure
equation is solved by using the iterative bi-
conjugate gradient method Ior sparse matri-
ces|18|.
Composition and saturation equations
The mole Iraction oI each component is
calculated at each iteration by using the
pressure oI the same iteration. Eirst the
number oI moles oI component i is calculated
Irom equation 15 and then total moles will be
obtained Irom equation 17, and :
i
obtained
by the ratio oI N
i
to N.

1 1 1 1 1 1 l n n n n n n
i o i o o g i
b
t
N T x P h T v
J

+ + + + + +
A

= A A A + A


1 n n n n
o cog g t b i
P P h qJ N
+
(
A + A A + +

(29a)
(21)
(22)
Archive of SID
www.SID.ir
A Sequential Formulation for Compositional Reservoir Simulation Using Peng Robinson Equation of State
56 Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1

1 1 1 1 1 l n n n n n
o o o g o
b
t
N T P h T P
J

+ + + + +
A

= A A A + A A


n n n
cog g t
P h q J N
(
+A A + +

(29b)
1
1
1
l
l i
i l
N
:
N
+
+
+
= (29c)
water saturation is obtained Irom equation
16:

n
w
b w
n
w
n
cow
n
o
n
w
b
l
w
N
J q
h P P T
J
t
N +
(

+
A A A A A
=
+ +
+

1 1
1
(30)

1
1
1
l
l w
w l
o
N
S
1
+
+
+
=
(31)
AIter calculating compositions, Ilash calcula-
tion is perIormed on calculated compositions
in each grid block at block pressure and
temperature to obtain L and V. Then, gas and
oil saturations are calculated as:


1 1
1
1
1
1
l l
o w
l
g l
l
g o g
J S
S
J


+ +
+
+
+

=
+ +
1 1 1
1
l l l
o g o
S S S
+ + +
= (33)
Where:
g g
o o g g
S
J
S S


=
+
Solution Procedure
The Iollowing procedure to solve the above
equations over a time step must be used.
1- Solving the pressure equation (equation
25) by using the iterative bi-conjugate
gradient method Ior sparse matrix|18|.
2- Obtaining water saturation (equation
31) and composition oI each com-
ponent (equa-tion 29) in each grid
block using the block pressures com-
puted in Step 1.
3- PerIorming Ilash calculation on calcul-
ated composition in Step 2 at block
temperature and pressure in each grid
block to obtain L and V.
4- Calculating oil and gas saturations (eq-
uations 32, 33).
5- Updating
f
T , x
i
, v
i
, N and N
w
..
The above steps are repeated until conver-
gence is achieved.
Results
a- Slim tube simulation (one dimensional case)
The recovery curves obtained in this study
were determined by using the slim tube
apparatus. The speciIications oI the apparatus
are given in Table 1.
Table 1. SpeciIications oI slim tube
Initial Diameter (mm) 4.65
Length (m) 12.78
Pore Volume (cm
3
) 94.73
Porosity (%) 43.7
Permeability (Darcy) 4.850
Porous Media Glass beads
The tube was initially saturated with reser-
voir oil. Oil was displaced by gas at a
constant injection rate oI 15 cm
3
/sec. The oil
sample used in this work was obtained Irom
an Iranian oil reservoir. Also, a gas sample
Irom a gas reservoir oI Iran was used as the
displacing agent. The compositional analysis
oI oil and gas, as well as the characteristics oI
the plus Iraction are given in Tables 2 and 3.
Bubble point pressure and swelling data are
(34)
(32)
Archive of SID
www.SID.ir
Shahrabadi, Dabir
Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1 57
also included. About a 1.2 pore volume oI
gas was injected in each test. At the end oI
each step produced, the volume oI oil and gas
were measured. The recovery Iactor was
calculated Irom the measured volumes oI the
produced oil and gas. This study is perIormed
to determine the eIIect oI interIacial tension
(IET) on simulating results in slim tube
experiments. Eive displacement tests at diI-
Ierent pressures were perIormed and the oil
recovery as a Iunction oI the injected pore
volume was recorded. The inIluence oI in-
terIacial tension on the results oI the deve-
loped model was examined in two simulation
runs with diIIerent relative permeability cur-
ves. The Iirst simulation was perIormed with
Iixed relative permeability curves (im-
miscible condition), while the variable rela-
tive permeability curves as a Iunction oI both
saturation and interIacial tension were used
(miscible condition), in the second run. To
include interIacial tension (8 ) in relative
permeabilities, the Iollowing curves were
used|17|.
Table 2. Compositional analysis oI oil, gas sample, swelling and saturation data
Archive of SID
www.SID.ir
A Sequential Formulation for Compositional Reservoir Simulation Using Peng Robinson Equation of State
58 Iranian 1ournal of Chemical Engineering, Jol. 2, Ao. 3
Table 3. Characteristics oI plus Iraction
Molecular Weight 194
SpeciIic Gravity 0.8897

1
g o
a r a r
ro ro o
k e k e S

= +

1
g g
a r a r
rg rg g
k e k e S

= +
In the above equations, r
8
8
-
= and ,
g o
a a are
positive numbers and used as adjustable
parameters. InterIacial tension between vapor
and liquid phases at 20,678 kPa (3000 psia)
and 83
o
C (181
o
E) at equilibrium was con-
sidered as base IET (8
-
). Generally, in a two
phase system, as the IET between phases
decreases, the curvature as well as the
residual saturation decreases. In the extreme
case oI zero IET, the relative permeability
curves become two straight lines and the
value oI the relative permeability oI each
phase will be equal to the value oI its
saturation. The Iollowing equations were
used Ior ,
rg ro
k k :
2
rg g
k S =
(37)
3
0.2
0.8
o
ro
S
k
| |
=
]
\
(38)
A one dimensional compositional model
composed oI 30 grid blocks was employed to
simulate slim tube experiments. Eigures 1
through 5 show the predicted oil recovery
versus injected gas volume Ior the two
simulated runs with diIIerent relative per-
meability curves at diIIerent pressures. The
experimental data are also included in the
Eigures. Generally, recovery versus pore
volume injection is a well-known plot to
illustrate, evaluate and compare the results oI
laboratory Ilood tests. The produced and in-
jected volumes are the same beIore break-
through, and so the resulting curve is a
straight-line. Then, it starts to deviate Irom
the straight trajectory. The slope oI the curve
portion decreases with increasing injected gas
volume, Iinally leveling oII. The point where
it starts to deviate Irom the straight line and
the level at which it becomes horizontal, is oI
great importance in making comparisons.
Figure 1. Recovery curve vs. PV injected at pressure 13,786 kPa (2000 psia)
(35)
(36)
O
i
I

R
e
c
o
v
e
r
y

(

%

)
PV Injected ( % )
Archive of SID
www.SID.ir
Shahrabadi, Dabir
Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1 59
Figure 2. Recovery curve vs. PV injected at pressure 20,678 kPa (3000 psia)
Figure 3. Recovery curve vs. PV injected at pressure 34,464 kPa (5000 psia)
Figure 4. Recovery curve vs. PV injected at pressure 41,357 kPa (6000 psia)
O
i
I

R
e
c
o
v
e
r
y

(

%

)
PV Injected ( % )
O
i
I

R
e
c
o
v
e
r
y

(

%

)
PV Injected ( % )
O
i
I

R
e
c
o
v
e
r
y

(

%

)
PV Injected ( % )
Archive of SID
www.SID.ir
A Sequential Formulation for Compositional Reservoir Simulation Using Peng Robinson Equation of State
60 Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1
Figure 5. Recovery curve vs. PV injected at pressure 44,804 kPa (6500 psia)
Comparisons show that the results Irom the
run with Iixed relative permeability curves
underestimate the oil recoveries, while the
run with interIacial tension in relative per-
meabilities improves the predictions. The
diIIerence between ultimate oil recoveries
obtained using Iixed relative permeability
and IET included in curves, is little at low
pressure, but signiIicant at high pressures.
This is due to the Iact that low pressure
corresponds to the immiscible zone, and
hence IET has a low eIIect on the recovery
results. ThereIore, including IET in relative
permeability curves can not improve the
predictions at low pressures, but as pressure
increases, the predictions show a good match
with the experimental results in a closer
margin. In other words, high pressure cor-
responds to the miscible zone, and so IET
plays a key role in oil recovery. ThereIore,
including IET in relative permeability curve
improves the results oI the model drastically
at high pressure regions. The eIIect oI
pressure on the recovery can be easily seen
while looking at Eigure 6, showing per-
Iormance curves at various pressures. The
greater the injection pressure, the higher the
ultimate recovery.
Figure 6. Oil recovery curves at diIIerent pressures
O
i
I

R
e
c
o
v
e
r
y

(
%

)
PV Injected ( % )
O
i
I

R
e
c
o
v
e
r
y

(

%

)
PV Injected ( % )
Archive of SID
www.SID.ir
Shahrabadi, Dabir
Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1 61
b-Areal run (two dimensional case)
The pattern shown in Eigure 7 was con-
sidered Ior the gas injection process. This
pattern was simulated using an areal 10 10
grid system. Two injectors and two producers
were considered. The bottom hole pressure
Ior injectors and producers were 38,600 kPa
(5600 psia) and 32,397 kPa (4700 psia)
respectively.
Figure 7. The pattern used in areal run
Additional required data are given in Table 4.
Table 4. Required data Ior areal run
Length oI reservoir,m (It) 900 (2951)
Width oI reservoir, m (It) 900 (2951)
Thickness, m (It) 6 (20)
Permeability (mdarcy) 40
Porosity (%) 18
Initial reservoir pressure , kPa
(psia)
34,465(5000)
Reservoir temperature ,
o
C (
o
E) 83 (181)
Initial water saturation 0.4
Eigure 8 shows the gas saturation contours at
1840 days aIter gas injection. In this Iigure,
no miscible region was Iound and thereIore
the process was immiscible.
Figure 8. Gas saturation contours at 1840 days aIter
gas injection
The pressure and methane mole Iraction
distributions oI the reservoir are shown in
Eigures 9 and 10. These curves conIirm each
other and show good trends compared with
anticipated real reservoir behavior. Hence,
the application oI the developed model Ior
the prediction oI reservoir behavior seems to
be successIul.
Figure 9. Pressure contours at 1840 days aIter gas
injection
1/4 Producer 1/4 Injector
Archive of SID
www.SID.ir
A Sequential Formulation for Compositional Reservoir Simulation Using Peng Robinson Equation of State
62 Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1
Figure 10. Methane mole Iraction contours at 1840
days aIter gas injection
Conclusion
In this work the Iormulation procedure oI a
three phase compositional reservoir model
and the method oI solving compositional
model equations were described. Also, an
experimental setup using a slim tube was
prepared and some displacement tests were
carried out. The oil recovery at diIIerent
injected volumes was measured during each
test. The model predictions were then com-
pared against measured laboratory re-covery
data. The eIIect oI including IET in relative
permeability curves on the outputs Irom the
model was also studied. Com-parison bet-
ween model outputs with and without,
including IET in relative per-meability
curves, indicated that having IET in relative
permeability curves has no eIIect (rather a
minor eIIect) on model results at low pres-
sures, while it was considerable at high
pressures. Acceptable results were obtained
Irom the areal run with respect to the
anticipated real reservoir behavior. Con-
sequently, the comparison oI the slim tube
results with the model predictions and areal
run predictions, indicate the possibility oI
employing this model Ior the reservoir
behavior prediction.
Nomenclature
A area perpendicular to Ilow ( ft
2
)
B
o
oil Iormation volume Iactor (bbl/STB)
C
1
conversion Iactor in equation 10
C
2
conversion Iactor in equation 18
F Ieed
f
i
f Iugacity oI component i in phase f
(atm)
g gravitational acceleration (ft/s
2
)
h depth (It)
k absolute permeability (mdarcy)
rf
k relative permeability oI phase f (im-
miscible condition)
rf
k relative permeability oI phase f (mis-
cible condition)
i
K K-value oI component i
L ratio oI moles in liquid phase to the
total number oI moles in the mixture
L
t
length ( ft )
N moles oI total hydrocarbon per unit oI
reservoir volume (
3
lb mole
ft

)
i
N moles oI component i per unit oI
reservoir volume (
3
lb mole
ft

)
1
c
n
N
+
moles oI water per unit oI reservoir
volume (
3
lb mole
ft

)
P pressure oI mixture at equilibrium
(psia)
f
P pressure oI phase f (psia)
cog
P oil / gas capillary pressure (psia)
cow
P water / oil capillary pressure (psia)
i
q molar injection / production rate oI
component i per unit volume
(
3
.
lb mole
ft dav

)
t
q molar injection / production rate oI
hydrocarbon per unit volume
(
3
.
lb mole
ft dav

)
Archive of SID
www.SID.ir
Shahrabadi, Dabir
Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1 63
w
q molar injection / production rate oI
water per unit volume (
3
.
lb mole
ft dav

)
r ratio oI IET to base IET
R
s
gas oil ratio (SCE/STB)
f
S saturation oI phase j
t time (day)
T temperature (
o
R)
f
T transmissibility oI phase j (
.
lb mole
psia dav

)
V ratio oI moles in vapor phase to the
total number oI moles in the mixture
b
J block volume (
3
ft )
i
x mole Iraction oI component i in oil
phase
i
v mole Iraction oI component i in gas
phase
i
: mole Iraction oI component i in
hydrocarbon
Greek Symbols
i
6 the rate oI accumulation oI component
i (
3
.
lb mole
ft dav

)
1
c
n
6
+
the rate oI accumulation oI water
(
3
.
lb mole
ft dav

)
6 the rate oI accumulation oI total
hydrocarbon (
3
.
lb mole
ft dav

)
1 porosity
f
speciIic weight oI phase j (
3
lb
ft
)
f
2 viscosity oI phase f (cP)
f
7 mass density oI phase f (
3
lb
ft
)
f
molar density oI phase f (
3
lb mole
ft

)
A diIIerence operator
V gradient operator
V divergence operator
f
u potential oI phase f (psia)
8 interIacial tension (
dvne
cm
)
8
-
base interIacial tension (
dvne
cm
)
References
1. Cao, H., 'Development oI Techniques Ior
General Purpose Simulators, Ph.D. Dis -
sertation, StanIord University, June 2002.
2. Kazemi, H., Vestal, C.R., and Shank, G.D.,
'An EIIicient Multicomponent Numerical
Simulator, Soc. Pet. Eng. J., 355 (Oct.
1978).
3. Eussell, L. T., and Eussell, D. D., 'An
Iterative Technique Ior Compositioml Reser -
voir Models, Soc. Pet. Eng. J., 211(Aug.
1979).
4. Young, L. C., and Stephenson, R. E., 'A
Generalized Compositional Approach Ior
Reservoir Simulation, Soc. Pet. Eng. J., 728
(Oct. 1983).
5. Nelon, J. S., 'Numerical Simulation oI
Compositional Phenomena in Petroleum
Reservoir, paper SPE 4274 presented at 3
rd
SPE Symposium on Reservoir Simulation,
Houston, Texas, (1973).
6. Acs, G., Doleschall, S., and Erakas, E.,
'General Purpose Compositional Model,
Paper SPE 10515 presented at 6
th
SPE
Symposium on Reservoir Simulation, New
Orleans, (1982).
7. Watts, J.M., 'A Compositional Eormulation
oI the Pressure and Saturation Equations,
Paper SPE 12244 presented at 7
th
SPE
Symposium on Reservoir Simulation, San
Erancisco, (1983).
8. Quandalle, P., and Savary, D., 'An Implicit
in Pressure and Saturations Approach to
Eully Compositional Simulation, paper SPE
18423 presented at Symposium on Reservoir
Simulation, Houston,Texas, (1982).
9. Celso Branco, M., and Rodriguez, E., 'A
Semi Implicit Eormulation Ior Compositional
Reservoir Simulation, SPE Advanced Tech -
nology Series, 4(1), 171(1995).
10. Wang, P., Balay, S., Sepehrnoori, K.,
Wheeler, J., Abate, J., Smith, B., and Pope,
G.A., 'A Eully Implicit Parallel EOS
Archive of SID
www.SID.ir
A Sequential Formulation for Compositional Reservoir Simulation Using Peng Robinson Equation of State
64 Iranian 1ournal of Chemical Engineering, Jol. 3, Ao. 1
Compositional Simulator Ior Large Scale
Reservoir Simulation, paper SPE 51885
presented at the 15
th
Symposium on Reser-
voir Simulation. Houston, Texas, (1997).
11. Jessen, K., and Orr Jr., E.M., 'Gravity
Segregation and Compositional Reservoir
Simulation, paper SPE 89448 presented at
the 14
th
symposium on improved oil reco-
very, Tulsa, Oklahoma, (2004).
12. Datta-Gupta, A., and King, M.J., 'A Semi
Analytical Approach to Tracer Elow Mo -
deling in Heterogeneous Permeable Media,
Advances in Water Resources , 18, 9 (1995).
13. Thiele, M.R., Batycky R.P., Blunt, M.J., and
Orr Jr., E.M., 'Simulating Elow in Hetero-
geneous Media Using Streamtubes and
Streamlines, SPE Reservoir Engineering,
10(1), 5 (1996).
14. Crane, M., Bratvedt, E., Bratvedt, K., and
OluIsen, R., 'A Eully Compositional Strea -
mline Simulator, paper SPE 63156 pre -
sented at the SPE Annual Technical Con-
Ierence and Exhibition, Dallas, Texas,
(2000).
15. Ichiro, O., Datta-Gupta, A., and King, M.J.,
'Time Step Selection During Streamline
Simulation Via Transverse Elux Correction,
paper SPE 79688 presented at the SPE
Reservoir Simulation Symposium, Houston,
Texas, (2003).
16. Seto, C.J., Jessen, K., and Orr Jr., E.M.,
'Compositional Streamline Simulation oI
Eield Scale Condensate Vaporization by Gas
Injection, paper SPE 79690 presented at the
SPE Reservoir Simulation Symposium,
Houston, Texas, (2003).
17. Nghiem, L.X., Eong, D.K., and Aziz, K.,
'Compositional Modeling with an Equation
oI State, Soc. Pet. Eng. J., 687(Dec. 1981)
18. Press, W.H., Teukolsky, S.A., Vetterling,
W.T., and Elannery, B.P., 'Numerical Reci -
pes in Eortran 77, Cambridge University
Press 1992.
Archive of SID
www.SID.ir

También podría gustarte