Sei sulla pagina 1di 15

SPE 154062

CO
2
Injection and CO
2
WAG in Dipping Gas Condensate and Oil Reservoirs
Mansour Soroush, SPE, Norwegian University of Science and Technology (NTNU); Lars Hoier, SPE, Statoil; and
J on Kleppe, SPE, NTNU
Copyright 2012, Society of Petroleum Engineers

This paper was prepared for presentation at the Eighteenth SPE Improved Oil Recovery Symposium held in Tulsa, Oklahoma, USA, 1418 April 2012.

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
In this study, different scenarios of CO
2
injection in dipping gas condensate and oil reservoirs are investigated through
reservoir simulations. Both miscible and immiscible flooding conditions were investigated for a range of different injection gas
mixtures, and geological realizations.

We find particularly interesting results for miscible flooding of gas condensate systems below dewpoint pressure. Here,
dropped out condensate is the prime target for enhanced recovery projects and multi-contact miscibility could develop through
the combined condensing/vaporizing mechanism.

Different patterns of permeability variation with depth in layering scenarios with dip angle showed distinct different responses
on produced condensate. CO
2
WAG in partially depleted gas condensate reservoirs seems to have the same value of oil
recovery in early times, but ultimate recovery is different according to layering heterogeneity. In the case of pure CO
2

injection, both up-dip and down-dip, it was found that homogeneous layering showed highest recovery. Here developed multi-
contact miscible oil-bank is able to move and sweep condensate above it easier. Applying various gas injection mixtures of
CO
2
and C
1
combinations, the effect on ultimate recovery were studied. In this case CO
2
injection is above minimum
miscibility pressure (MMP) resulting in high recoveries.

CO
2
WAG in dipping oil reservoir was also studied extensively, based on injection pattern, MMP values and various layering
systems. CO
2
WAG in scenarios with increasingly trend of permeability with depth had the highest value of recovery. This is
because of prevention from early gas breakthrough in upper layers and good sweep efficiency in lower layers. Pure CO
2

injection with total same injection volume showed lower recovery. This may verify that gravity in WAG water injection period
is the most effective parameter in the case of down-dip WAG.

Introduction
Gas injection is an important EOR method for increasing oil recovery. Planning efficient injection scenarios depends on many
parameters. Most important parameters are precise evaluation of MMP in injection process, choosing proper injection gas
based on cost and availability of gas, technical concerns, scheduling the process and special behavior of injected gas in
companion with the reservoir fluids. CO
2
is also considered for EOR processes that could be injected either in miscible or
immiscible flooding scenarios.
Immiscible CO
2
injection has been successfully practiced in heavy-oil reservoirs. Two main mechanisms in immiscible CO
2

displacement are oil swelling and reducing oil viscosity. Low-pressure reservoirs and reservoirs with oil gravities less than
approximately 30API are typical candidates for immiscible CO
2
displacement (Whitson and Brule, 2000).

In higher pressure
injection above MMP, CO
2
phase behave from vapor-like to liquid-like CO
2
-richer phase. This phase extracts some
intermediate and heavy components of the reservoir oil or gas condensate and higher recovery is achievable (Whitson and
Brule, 2000).
CO
2
injection could be investigated in different reservoir conditions and some important parameters should be noticed.
Alavian and Whitson (2005) showed that there is good CO
2
IOR potential in naturally fractured reservoir. Oil recoveries from
CO
2
injection is reported in the range of 80-90% for the fractured reservoir that they studied. There are also lots of efforts done
on improving MMP measurements that affect directly the CO
2
injection process (Emera and Sarma 2005, and Johns et al.
2 SPE 154062
2010). Hoier and Whitson (2001) found that the best extrapolation method to find infinite grid block approximation in
slimtube simulation for MMP estimation is the Pade approximation (Van Dayke, 1978).
Although good recovery could be obtained in CO
2
injection process, some special problems with CO
2
injection should be
noticed. Whitson (2000) described these side effects that might have small to large consequences on the whole process. One of
these problems is multiphase behavior of the CO
2
-oil that is limited to special reservoir temperatures and CO
2
concentrations.
According to the reservoir condition, CO
2
in contact with oil could be in CO
2
-enriched vapour or CO
2
-enriched liquid phase.
Asphaltine precipitation could be a more serious problem because it happens in a wider range of pressure and CO
2

concentrations and this also could result in some wettability alternation in advance (Whitson and Brule, 2000). However
simulating different phenomena of CO
2
injection leads to better understanding of the process and thus better planning of
injection projects.

Flooding in Dipping Reservoirs
Flooding behavior in porous media depends on viscous and gravity forces. Viscous dominated flow takes place in open
systems with fluid injection or fluid production from that. Gravity dominated flow can occur both in closed or open systems.
The effect of gravity is considered in many studies including phase density differences and dipping reservoirs. In the case of
having an angle of inclination in the system the Darcy equation for flow of each phase will be,
u =
Kk
r

(
dP
dx
+pgsino) . (1)
In which is considered to be angle of inclination. is assumed to be positive if the flow is upward.
Total flux in the reservoir will be,
u
1
= u
o
+u
g
+u
w
....(2)
Assuming that gas and oil are flowing phases,
u
1
=
Kk
rg

g
[
dP
dx
+p
g
gsino +
Kk
rc

c
[
dP
dx
+p
o
gsino ......(3)
Fractional flow of gas is,
f
g
=
U
g
U
T
=
1+
(p
g
-p
o
)KR
o

o
U
T
1+
R
o

o
R
g

g
...(4)
Assuming a gravity number as below,
N
gv
=
pKgsInu
U
T
.....(5)
This dimensionless number is defining parameter for effect of gravity in dipping reservoir. Including the gravity number
the fractional flow equations can be modified as,

g
=
1+N
g
k
rc
1+
k
rc

c
k
rg

g
...(6)

o
=
1+N
g
k
rg
1+
k
rg

g
k
rc

c
....(7)
The continuity equation for two flowing phases is,

S
w,g
t
= -
0
w,g
x
.(8)
Defining normalized injection volume and using dimensionless forms of saturation and time, equations could be simplified
and applying initial and boundary conditions for the problem the proper solution may be obtained. Immiscible flow conditions
are considered for all cases. In miscible flooding, it is critical to understand the mechanisms as well as density and saturation
variations at reservoir conditions.

Condensing Vapourizing Mechanism
Several gas injection experiments verified with equation of state simulations were reported by Zick (1998). He concluded
that a mixture of condensing vaporizing gas drive mechanism is the controlling phenomena for several EOR processes by
enriched gas injection. Actually simulation runs based on his studies indicate that the traditional condensing vaporizing
concept is not occurring in many cases. Above MMP, the condensing/vapourizing mechanism can generate miscible
displacement. He also showed that the true MMP could be lower than what is predicted by the vapourized or condensing gas
drives mechanisms. This mechanism has been verified by many publications (Hoier and Whitson 2001, Uleberg and Hoier
2002). These displacements are characterized by developments of a nearly miscible transition zone that is created by a zone of
oil in equilibrium with gas that has lost its intermediates (Zick, 1998). Fig. 1 shows the concept of condensing/vapourizing
mechanism. In this study, simulations are based on the concept developed by Zick (1998).
SPE 154062 3

Fig. 1Oil and gas densities based on condensating/vapourizing mechanism. Net vaporization of heavy component at upstream of
the front and net condensation of intermediate component at downstream of the front

Model Description
A 2u 2u 1S grid system in x, y and z direction, respectively, is considered for a 3D conceptual simulation model. The
reservoir size is 3500 ft by 3500 ft horizontally and has a thickness of 60 ft. Three layers are selected for more detailed studies.
Layer 1 is from grid block 1 to grid block 5, layer 2 is from grid block 5 to 10 and layer 3 is from grid block 10 to 15 in z
direction. Two different sets of simulations including an oil reservoir and a gas condensate model are used in this study. The
oil reservoir model is initialized at a depth of 8335 ft with initial pressure of 3984.3 psia and an initial oil saturation of 0.8,
with porosity of 0.3. The Peng-Robinson equation of state at 16u has been used for PVT modeling of this simulation.
Another model with the same geology is used for the gas condensate system. The gas condensate model is initialized at a depth
of 7500 ft and an initial pressure of 3550 psia, with an initial water saturation of 0.16. Table 1 shows the composition that has
been used for oil reservoir fluid and injection gas, which is used in one of the injection scenarios to compare with CO
2

injection. Gas condensate composition is given in Table 2.
Figs. 2-3 show capillary pressure and relative permeability curves for the oil reservoir model. Corresponding curves for the
gas condensate reservoir are plotted in Figs. 4-5. One injector and one producer are located on the sides of the both reservoir
models and completed in all layers. Table 3 briefly describes the model properties for the oil and gas condensate reservoirs.
Table 4 shows geological scenarios that are considered for this study. A dip angle of 10 degree in the direction of the x axis is
considered with down-dip injection and production from up-dip of the reservoir. In some scenarios, up-dip injection is
considered as well. It is also possible to consider other dip angles and well patterns, but in this study CO
2
injection behavior is
the main target.

Table 1Oil reservoir composition,
Composition Mole fraction(oil) Mole fraction(injection gas)
CO
2
0 0
C
1
0.5 0.77
C
3
0.03 0.2
C
6
0.07 0.03
C
10
0.2 0
C
15
0.15 0
C
20
0.05 0

Table 2Gas condensate composition
Composition Mole fraction
CO
2
0.0121
N
2
0.0194
C
1
0.6599
C
2
0.0869
C
3
0.0591
C
4-6
0.0967
C
7
P
1
0.0472
C
7
P
2
0.0153
C
7
P
3
0.0034
200
250
300
350
400
450
500
550
600
650
700
0.5 0.55 0.6 0.65 0.7 0.75
Nor mal i zed Lengt h f r om Inj ect i on Poi nt
P
h
a
s
e

d
e
n
s
i
t
i
e
s
,

k
g
/
m
3
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
N
o
r
m
a
l
i
z
e
d

o
i
l

s
a
t
u
r
a
t
i
o
n
,

f
r
a
c
t
i
o
n
Oil density
Gas density
So
4 SPE 154062
Table 3Reservoir properties
Gas condensate Oil
Grid block dimensions (ft) SSuu SSuu 6u SSuu SSuu 6u
Initial depth (ft) 7500 8335
Initial pressure (psia) 3550 3984.3
Saturation pressure (psia) 3550 2302
Layer 1 Grid blocks 1 to 5 Grid blocks 1 to 5
Layer 2 Grid blocks 5 to10 Grid blocks 5 to10
Layer 3 Grid blocks 10 to15 Grid blocks 10 to15
Completion All layers All layers
Porosity 0.3 0.3
Swi 0.16 0.2

Table 4Scenario definitions
Scenario 1 (S1)
K
x
=K
y
, mD (all layers) 100
K
z
, mD (all layers) 50
Scenario 2 (S2 - Heterogen layering, decreasing k with depth)
K
x
=K
y
, mD layer1 200
K
x
=K
y
, mD layer2 100
K
x
=K
y
, mD layer3 50
K
z
(all layers) 50
Scenario 3 (S3 - Heterogen layering, increasing k with depth)
K
x
=K
y
, mD layer1 50
K
x
=K
y
, mD layer 2 100
K
x
=K
y
, mD layer 3 200
K
z
, mD (all layers) 50

Fig. 2Water-oil capillary pressure and relati ve permeability,
oil reservoir.
Fig. 3Gas-oil capillary pressure and relati ve permeability, oil
reservoir.

Fig. 4Water-oil capillary pressure and relati ve permeability,
gas condensate reservoir.
Fig. 5Gas-oil capillary pressure and relati ve permeability,
gas condensate reservoir.
0
5
10
15
20
25
30
35
40
45
50
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1
K
r
S
w
Krw
Kro
Pcwo
P
c
,

p
s
i
a
0
5
10
15
20
25
30
35
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1
K
r
S
l
Krg
Kro
Pcgo
P
c
,

p
s
i
a
0
10
20
30
40
50
60
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1
K
r
S
w
krw
kro
Pcwo
P
c
,

p
s
i
a
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 0.2 0.4 0.6 0.8 1
K
r
S
l
krg
kro
SPE 154062 5
Simulation Scenarios and Results
A. Oil Reservoir Model
CO
2
injection in gas cycles of WAG scenarios and also continuous CO
2
injecting are considered for the oil reservoir
model. Different layering scenarios in a 10-degree dipping reservoir, with miscible and immiscible cases of CO
2
injection are
compared. Finally, rich gas injection is modeled and compared with the cases of CO
2
injection.
WAG schedules are defined and started just after opening the production well by a rate of . The first WAG
cycle starts after six months of water injection with rate of followed by six months of injecting CO
2
with a rate
of . These cycles repeatedly continue as far as the constraints for well control are fulfilled. Well restrictions are
an injection bottomhole pressure (BHP) of 10000 psia and a production BHP of 1000 psia. Three geological scenarios for the
dipping reservoir are considered for this study, as described in Table 4. Scenario 1 includes three layers with the same
permeability, scenario 2 includes a trend of decreasing permeability in deeper layers and scenario 3 includes a trend of
increasing permeability in deeper layers. It has been attempted to determine the effect of injecting CO
2
in all these cases and to
evaluate the influencing parameters and heterogeneities on the result.
CO
2
injection MMP has been obtained from slimtube simulation and the Pade approximation (Van Dayke, 1978) is used to
predict the infinite grid block results. Fig. 6 shows different MMP values that are calculated by PVT modeling and also
slimtube simulations with different grid sizes. The best result is obtained from three highest number of grid blocks averaging,
by the Pade approximation that is compatible with the PVT model result and this value is chosen for true MMP for this study.
In Fig. 6 the effect of grid refinement on the result of slimtube simulation is shown, and the result is estimated for the case of
an infinite number of grid blocks. By this method, reliable MMP values could be collected from the calculations.


Fig. 6MMP calculation for CO
2
injection in oil reservoir (GB: number of grid blocks, SL: straight line approximation, Pol: polynomial
degree two approximation)

As it can be observed in Fig. 7, after several cycles of CO
2
WAG in the model, the average reservoir pressure is
approaching the MMP of CO
2
injection for all the CO
2
WAG scenarios and then the reservoir is pressured. In this case
reservoir pressure will be above MMP and saturation pressure most of the time. It has been shown in Fig. 8 that the highest oil
recovery is achieved in scenario 3 and the lowest is for scenario 2. Increasing the permeability trend to the deeper layers
affects directly the ultimate recovery of the system.

1300
1500
1700
1900
2100
2300
2500
2700
2900
GB 100 GB 200 GB 500 GB
2500
GB
5000
Infinity
(SL)
Infinity
(Pol)
Infinity
(Pade 5
Points)
Infinity
(Pade 3
Points)
PVT
Modeling
Grid Size Effect 2700 2475 2350 2250 2150 1950 1920 1922 1918 1892.59
M
M
P
,

p
s
i
a
Grid size effect
6 SPE 154062

Fig. 7Average reservoir pressure, CO
2
WAG. Fig. 8Oil recovery, CO
2
WAG.

To have better understanding of the injection slug movement, saturation maps are plotted at different times of the
simulation. Fig. 9 shows the saturation profile after 13 years of WAG cycles for scenario 1. The cross-sections for different
layering are plotted in Figs. 10-12 for scenarios 1, 2 and 3 at the same time (dip angle is exaggerated in plotting to have better
display). Better sweep efficiency occurred in the case of scenario 3 where the permeability trend is increasing with depth. This
is because of gravity effects in water cycles that tend to sweep lower parts and the lower permeability of shallower layers
preventing gas fingering and early breakthrough.


Fig. 9Oil saturation after 13 years, scenario 1.


Fig. 10Oil saturation after 13 years, scenario 1.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 2000 4000 6000 8000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1
PAVG S2
PAVG S3
MMP=1918 psia
0
10
20
30
40
50
60
70
80
90
100
0 2000 4000 6000 8000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC S1
OILREC S2
OILREC S3
SPE 154062 7


Fig. 11Oil saturation after 13 years, scenario 2.


Fig. 12Oil saturation after 13 years, scenario 3.

The same geology model is considered for continuous CO
2
injection. CO
2
injection is started by a rate of 14uuu mcfu
and the production starts with rate of 1uuuu stbu. Fig. 13 shows how the average reservoir pressure is changing with time
due to CO
2
injection below MMP. Pressure stabilizes around 1250 psia. In this case scenario 3 has the best recovery close to
the results of scenario 1, but scenario 2 has lower recovery due to higher values of permeability in upper layer that makes CO
2

bypass there.
Ultimate recovery is lower in pure CO
2
injection compared to CO
2
WAG. Economical analysis based on availability of
CO
2
and water and also investigation of production per volume injected will result in the best scenario selection. Fig. 14
compares oil recovery from each of the scenarios.

8 SPE 154062

Fig. 13Average reservoir pressure, CO
2
injetion. Fig. 14Oil recovery, CO
2
injection.

Then it was attempted to inject CO
2
from up-dip and produce from down-dip to evaluate the differences. Fig. 15 shows
trends of average reservoir pressure due to CO
2
injection and in Fig. 16 the oil recovery obtained in different scenarios is
shown. A higher volume of oil is recovered in this case and gravity helps down-dip production of the oil.

Fig. 15Reservoir pressure, up-dip CO
2
injection. Fig. 16Oil recovery, up-dip CO
2
injection.

To obtain more results on the effect of miscibility, CO
2
is also injected in WAG scenarios at immiscible conditions. The
same geology and WAG plan is set with a gas injection rate of Suuu mcfu and a water injection rate of Suuu stbu. WAG
scenarios start after two years of production with a rate of 1uuuu stbu. As it is demonstrated in Fig. 17, CO
2
is injected as an
immiscible gas and becomes closer to the miscible condition at late simulation times. Fig. 18 shows oil recovery of the three
scenarios based on immiscible CO
2
injection. Recovery is in the lower range compared to CO
2
WAG because of lower
injection rates and the same trend of production development is observed in three different layering scenarios. WAG cycles in
an immiscible case are performed in lower injection rates and then well controlling constrains continue to be valid longer.


Fig.17Average reservoir pressure, immiscible CO
2
WAG. Fig. 18Oil recovery, immiscible CO
2
WAG.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 3000 6000 9000 12000 15000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1
PAVG S2
PAVG S3
MMP=1918 psia
0
10
20
30
40
50
60
70
80
90
0 3000 6000 9000 12000 15000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, Days
OILREC S1
OILREC S2
OILREC S3
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 3000 6000 9000 12000 15000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1
PAVG S2
PAVG S3
MMP=1918 psia
0
10
20
30
40
50
60
70
80
90
0 3000 6000 9000 12000 15000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC S1
OILREC S2
OILREC S3
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 4000 8000 12000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1
PAVG S2
PAVG S3
MMP=1918 psia
0
10
20
30
40
50
60
70
80
0 3000 6000 9000 12000 15000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC S1
OILREC S2
OILREC S3
SPE 154062 9
Rich gas injection is considered in the WAG cycles (composition is given in Table 2) and is compared with CO
2
injection
results. The difference between MMP for both injection plans characterizes these cases. Fig. 19 shows average reservoir
pressure of three scenarios due to injecting rich gas. A higher value of MMP is determined for this case through PVT modeling
and is verified with slimtube simulation. The higher MMP makes the process to be at an immiscible condition in early
simulation times. Fig. 20 shows the values of oil recovery in different scenarios and almost the same trend is observed for all
the scenarios due to rich gas injection. A comparison of CO
2
and rich gas injection in WAG, pressure and oil recovery in
scenario 1 is presented in Figs. 21-22. A slightly higher value of produced oil is obtained using CO
2
as injected gas in WAG,
mainly at a late time of the simulation and this is related to MMP difference.


Fig. 19Average reservoir pressure, rich gas WAG. Fig. 20Oil Recovery, rich gas WAG.

Fig. 21Average reservoir pressure, rich gas and CO
2
WAG.

Fig. 22Cumulati ve oil production, rich gas and CO
2
WAG.
Different injection plans including CO
2
WAG, rich gas WAG, and up-dip and down-dip CO
2
injection are compared and
results are presented in Figs. 23-24. With the same amount of CO
2
injected, CO
2
WAG yields highest values of recovery in
shorter time, but due to the higher injection pressure, gas will break through and well constrains are not valid anymore, while
in the case of CO
2
injection breakthrough time is later and maximum possible recovery is related to up-dip CO
2
injection.
Table 5 demonstrates a brief comparison of these injection scenarios in homogeneous layers dipping reservoir.

Fig. 23Average reservoir pressure, injection scenarios Fig. 24Oil recovery, injection scenarios
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 2000 4000 6000 8000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1 Rich Gas inj.
PAVG S2 Rich Gas Inj.
PAVG S3 Rich Gas Inj.
MMP=2842.13 psia
0
10
20
30
40
50
60
70
80
90
0 2000 4000 6000 8000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC S1-Rich Gas Inj.
OILREC S2 Rich Gas inj.
OILREC S3 Rich Gas Inj.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 2000 4000 6000 8000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1 Rich Gas
WAG
PAVG S1 CO2 WAG
0
5000
10000
15000
20000
0 10000 20000 30000 40000 50000 60000
C
u
m
u
l
a
t
i
v
e

o
i
l

p
r
o
d
u
c
e
d
,
M
S
T
B
Cumulative gas injected, MMCF
CUMOIL S1 Rich Gas
WAG
CUMOIL S1 CO2 WAG
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 2000 4000 6000 8000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG CO2 WAG
PAVG Rich Gas WAG
PAVG CO2 Inj. UP-Dip
PAVG CO2 Inj Down-Dip
0
10
20
30
40
50
60
70
80
90
0 3000 6000 9000 12000 15000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC CO2 WAG
OILREC Rich Gas WAG
OILREC CO2 Inj. UP-Dip
OILREC CO2 Inj. Down-Dip
10 SPE 154062

Table 5comparing ultimate recovery and produced oil in scenario 1 (homogeneous layers)
Injection plan Injection location Miscibility condition
Ultimate Recovery,
percent
Produced oil*, MSTB
CO
2
WAG down-dip miscible 79.22 19882.36
CO
2
injection down-dip immiscible 78.95 16845.88
CO
2
injection up-dip immiscible 85.15 18246.62
CO
2
WAG down-dip immiscible 65.44 N.A. **
Rich gas WAG down-dip near miscible 76.57 19219.07
*Produced oil at 50000 MMCF gas injected ** 16425.5 MSTB recovered after 20748 MMCF gas injected, then simulation stopped due to very low production BHP

B. Gas Condensate Model
Another model is constructed with the same geological properties to investigate the process with presence of gas
condensate. Table 2 shows the gas condensate composition that is used for initialization of the model and Peng-Robinson EOS
at 2uu is set for the PVT model. The same well location is considered for this case but different well constrains have been
used. BHP for the production well is set at 500 psia and the injection BHP is set at 4000 psia.
In the gas condensate case it has been attempted to run the same WAG scenario using CO
2
as an injection gas. Then in
order to evaluate the effect of pure CO
2
injection, C
1
combinations are also considered. Increasing enrichment levels of CO
2
in
CO
2
-C
1
combination will result in decreasing MMP as it is plotted in Fig. 25. After a critical enrichment level (E
-
) of 0.1015
of C
1
, the combination will be a condensing-vaporizing miscibility mechanism and MMP gradually decreases while mixing
more CO
2
. According to calculation for the case of pure CO
2
injection, NNP = 174S.u2 psia is obtained while with mixing
80 percent C
1
, MMP is 3051.82 psia and for pure C
1
the calculated MMP is 3480.81 psia. This combination of injection gas
has been used in the simulation to verify the effects of different injection gases.
The same geological scenario is used and in all the cases after two years of gas production by the rate of 62uu mcfu,
WAG cycles start with CO2 injection rate of 7Suu mcfu followed by water injection rate of 7Suu stbu. All these cases are
flooding of gas condensate systems below dewpoint pressure. Here, dropped out condensate is the prime target for enhanced
recovery and multi-contact miscibility could develop through the combined condensing/vaporizing mechanism.


Fig. 25CO
2
-C
1
enrichment level effect on MMP

It can be observed in Fig. 26 that all of the scenarios occur above MMP and miscible injection condition is maintained. The
production continues to the end of the WAG plan and again scenario 3 has the highest value of ultimate condensate production
while it is very close to scenario 1 and scenario 2 is the lowest one while this time the values are almost the same at short
simulation time of the WAG. More than 80% of condensate is recovered in scenario 3 and scenario 1 while it is around 75% in
scenario 2 after 30 years of WAG cycles. Fig. 27 shows condensate recovery by CO
2
WAG in the gas condensate model.

1500
2000
2500
3000
3500
0 20 40 60 80 100
M
M
P
,

p
s
i
a
CO
2
Enrichment Level
CO
2
-C
1
enrichment effect on MMP
E*=0.1015C1
SPE 154062 11

Fig. 26Average reservoir pressure, CO
2
WAG in gas
condensate reservoir.
Fig. 27Oil recovery, CO
2
WAG in gas condensate reservoir.

All the process is repeated again with pure CO
2
injection with rate of 8uuu mcfu. Reservoir pressure response is plotted
in Fig. 28, and again the process is occurring above MMP and miscible condition is achieved. Fig. 29 is condensate recovery
by pure CO
2
injection. More than 80% condensate recovery is achieved by this injection plan and scenario 1 that is
homogeneous layers has the best recovery that is showed in Fig. 29. Scenario 2 is moderate case and scenario 3 is the worst
one. Here miscible oil bank is able to move and sweep condensate above it easier in a smooth path in homogenous layering
system. In scenario 3, the bank could penetrate from lower layers and best efficiency is not achievable. Multi-contact miscible
oil-bank could be seen in Fig. 30 which is a plot of oil saturation after 10 years of CO
2
injection. The condensing vaporizing
mechanism will result in an oil bank creation ahead of the front. This phenomenon is presented by Zick (1988) and discussed
by Whitson and Hoier (2001). CO
2
is also injected from up-dip, Fig. 31 shows the pressure and Fig. 32 shows condensate
recovery. The same trend could be observed in this case.


Fig. 28Average reservoir pressure, CO
2
injection in gas
condensate reservoir.
Fig. 29Oil recovery, CO
2
injection in gas condensate
reservoir.

0
500
1000
1500
2000
2500
3000
3500
4000
0 4000 8000 12000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1
PAVG S2
PAVG S3
MMP=1745.02 psia
0
10
20
30
40
50
60
70
80
90
0 4000 8000 12000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC S1
OILREC S2
OILREC S3
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 5000 10000 15000
A
v
e
r
g
a
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1
PAVG S2
PAVG S3
MMP=1745.02 psia
0
10
20
30
40
50
60
70
80
90
0 5000 10000 15000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC S1
OILREC S2
OILREC S3
12 SPE 154062

Fig. 30Oil saturation in scenario 1, CO
2
injection in gas condensate reservoir and oil bank creation.


Fig. 31Average reservoir pressure, up-dip CO
2
injection in
gas condensate reservoir.
Fig. 32Conensate recovery, up-dip CO
2
injection in gas
condensate reservoir.

To evaluate the effect of different combinations of injection gases, C
1
, CO
2
and a mixture of 80% CO
2
and 20% of C
1
are
used for gas injection cycles. Fig. 33 shows the average reservoir pressure trend in these combinations and the MMP values. In
the C
1
injection case, the process is always below MMP and immiscible gas injection is performed while in CO
2
injection the
system is always above MMP. In the case of a combination mixture of 80% CO
2
, the process is in a near miscible condition.
Condensate recovery can be observed in Fig. 34. In the pure CO
2
injection scenarios, recovery tend to be lower than other
cases in short term injection periods, while it exceeds and reaches the highest ultimate recovery at the end of the process.
WAG with CO
2
, WAG with C
1
, and up-dip and down-dip CO
2
injection are compared and presented in Figs. 35-37. For
this reservoir condition, the highest ultimate condensate recovery could be achieved while performing CO
2
WAG and C
1

WAG results in the lowest value of ultimate condensate recovery. In Table 6 a brief comparison of all continuous injection
and WAG scenarios in gas condensate reservoir is showed. All the cases in Table 6 are from scenario 1 that is homogeneous
layering system. To be able to compare the cases economically, produced condensate at the same injected gas is also provided
in Table 6.

0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 5000 10000 15000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG S1
PAVG S2
PAVG S3
MMP=1745.02 psia
0
10
20
30
40
50
60
70
0 5000 10000 15000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC S1
OILREC S2
OILREC S3
SPE 154062 13

Fig. 33Reservoir pressure, injecting C
1
, CO
2
and
combination in gas condensate reservoir.
Fig. 34Condensate recovery, injecting C
1
, CO
2
and
combinationin gas condensate reservoir.
Fig. 35Reservoir pressure, comparing injection plans in gas
condensate reservoirs.
Fig. 36Oil production, different injection plans.

Fig. 37Oil recovery, different injection scenarios.

Table 6Comparing ultimate recovery and produced condensate in scenario 1 (homogeneous layers)
Injection plan Injection location Miscibility condition
Ultimate condensate
recovery, percent
Produced
condensate*, MSTB
CO
2
WAG down-dip miscible 81.96 2290.1
CO
2
injection down-dip miscible 81.76 1570.59
CO
2
injection up-dip miscible 60.40 1333.33
C
1
WAG down-dip immiscible 57.59 1608.88
80%C
1
-20%CO
2
WAG down-dip near-miscible 62.53 1748.64
*Produced condensate at 38000 MMCF gas injected

0
500
1000
1500
2000
2500
3000
3500
4000
0 3000 6000 9000 12000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG C1
PAVG CO2
PAVG 80C1-20CO2
C
1
MMP=3480.81 psia
80% C
1
-20%CO
2
MMP=3051.81 psia
CO
2
MMP=1745.02 psia
0
10
20
30
40
50
60
70
80
90
0 3000 6000 9000 12000
C
o
n
e
n
s
a
t
e

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC C1
OILREC CO2
OILREC 80C1-20CO2
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 4000 8000 12000
A
v
e
r
a
g
e

p
r
e
s
s
u
r
e
,

p
s
i
a
Time, days
PAVG WAGCO2
PAVG WAGC1
PAVG CO2
PAVG CO2 UP-Dip
CO
2
MMP=1745.02 psia
C
1
MMP=3480.81 psia
0
500
1000
1500
2000
2500
0 30000 60000 90000 120000
C
u
m
u
l
a
t
i
v
e

o
i
l

p
r
o
d
u
c
e
d
,

M
S
T
B
Cumulative gas injected, MMCF
CUMOIL WAGCO2
CUMOIL WAGC1
CUMOIL CO2
CUMOIL CO2 UP-Dip
0
10
20
30
40
50
60
70
80
90
0 3000 6000 9000 12000 15000
O
i
l

r
e
c
o
v
e
r
y
,

%
Time, days
OILREC WAGCO2
OILREC WAGC1
OILREC CO2
OILREC CO2 UP-Dip
14 SPE 154062
Conclusion
A conceptual numerical simulation is carried out using a dipping reservoir model with both oil and gas condensate fluids.
Different scenarios are compared and investigated through CO
2
EOR processes. According to the results of this study, the
following observations are made:
1. CO
2
injection in both the oil and the gas condensate dipping reservoirs could be achievable and good values of oil
recovery are obtained.
2. CO
2
WAG in dipping oil reservoir results in high recoveries based on injection pattern, MMP values and layering
systems. CO
2
WAG in a reservoir with increasing permeability trend with depth (scenario 1) had the highest
values of recovery. This is because of prevention from early gas breakthrough in upper layers and good sweep
efficiency in lower layers.
3. In the oil reservoirs, it has been observed that immiscible up-dip continues CO
2
injection could achieve highest
recovery while higher oil production could be obtained in shorter time with the same value of gas injected in
miscible CO
2
WAG.
4. CO
2
WAG in gas condensate systems seems to have the same value of condensate recovery in early times but
ultimate recovery trend is almost the same as the oil reservoir cases according to layering heterogeneity.
5. In the case of pure CO
2
injection up-dip and down-dip, the homogeneous layering system showed highest possible
recovery. Here, the oil bank is able to move and sweep the condensate above it easier and in a smooth path in the
homogenous layering system. In scenario 3 the bank could penetrate from lower layers and the best efficiency is
not achievable.
6. CO
2
injection is compared to the C
1
and C
1
-CO
2
injection combinations and this leads to higher variation in
ultimate recovery in the case of the gas condensate model. In this case, CO
2
injection is far above MMP and good
recovery is achieved compared to other cases.
7. In all the dipping reservoir injection cases, heterogeneities can play both positive and negative role on the
recovery based on the injection fluid type and schedule and it should be investigated individually in each case.

Nomenclature
= roctionol low, roction
K = obsolutc pcrmcobility, l
2
, m
k

= rcloti:c pcrmcobility
Pcgo = gos oil copillory prcssurc,
m
lt
2 ,
, psio
Pcwo = wotcr oil copillory prcssurc,
m
lt
2 ,
, psio
S = soturotion, roction
u = lux,
l
t
, ,
m
s

g = gro:irty occcrlorotion,
l
t
2
, ,
m
s
2 ,


Greek
p = Jcnsity,
m
l
3 ,
,
kg
m
3
,
p = :iscosity,
m
lt
,
, cp
= porosity, pcrccnt

Abbreviations
CuH0II = cumuloti:c oil proJuccJ, l
3
, HSIB
CuH0I = cumuloti:c gos in]cctcJ, l
3
, HHCF
0IIREC = oil rcco:cry, pcrccnt
S1, S2 onJ S = sccnorio numbcrs
PAI0 = o:crogc rcscr:oir prcssurc,
m
lt
2 ,
, psio
wA0 = wotcr oltcrnoting gos in]cction

Aknowledgement
The authors wish to express special appreciation to the research center BIGCCS (International CCS Center) for hosting the
research, and to STATOIL for the financial grant to the publication.

References
Alavian S.A., Whitson C.H. 2005. CO
2
IOR Potential in Naturally Fractured Haft Kel Field, Iran, International Petroleum Technology
Conference, held in Doha, Qatar 21-23 November, IPTC 10641
Emera M.K., Sarma H.K. 2005, Genetic Algorithm (GA)-Based Correlations Offer More Reliable Prediction of Minimum Miscibility
Pressures (MMP) Between the Reservoir Oil and CO
2
or Flue Gas, presented at Petroleum Society 6
th
Canadian international
Petroleum Conference, Calgary, Alberta.
SPE 154062 15
Grenwalder M., Clemens T. 2008. Immiscible Gas Injection: Challenges, Example of 16
th
TH Horizon, Austria, presented at SPE
Europec/EAGE annual conferences in Rome, Italy, 9-12 J une. SPE 113504
Hoier, L. 1997. Miscibility Variation in Compositional Grading Petroleum Reservoirs, PhD thesis, NTNU, Trondheim, Norway
Hoier, L. and Whitson C.H. 2001. Miscibility Variation in Compositionally Grading Reservoirs, SPE Reservoir Evaluation and Engineering
J ournal
Karim F., Berzins T.V., Schenewerk P.A., Bassiouni Z.A., and Wolcott J .M. 1992. Light Oil Recovery from Cyclic CO
2
Injection: Influence
of Drive Gas, CO
2
Injection Rate, and Reservoir Dip, presented at SPE Rocky Mountain regional meeting held in Casper, Wyoming,
May 18-21, SPE 24336
Uleberg K. Hoier L. 2002. Miscible Gas Injection in Fractured Reservoirs, presented at SPE/DOE improved oil recovery symposium held in
Tulsa, Oklahoma, US, 13-17 April. SPE 75146
Van Dyke, M. 1978. Perturbation Methods in Fluid Mechanic Parabolic Press, 2
nd
edition, 205-207
Whitson C.H., and Brule M.R. 2000. Phase Behaviour, Monograph, SPE of AIME, Richardson, Texas
Zick A.A., 1986. A Combined Condensing Vapourising Mechanism in the Displacement of Oil by Enriched Gases, presented at 61
st
Annual
technical conference of SPE held in New Orleans, LA, US, SPE 15493

Conversion Factor
bbl 1.S8987S E -u1 = m
3

t S.u48 E -u1 = m
t
3
2.8S168S E -u2 = m
3

psi 6.8947S7 E +uu = kPo
API 141.S(1S1.S +API) =
g
cm
3
,
(-S2)1.8 E -u2 = m
3

Potrebbero piacerti anche