Professional Documents
Culture Documents
SPE 18061
This paper was prepared for presentation at the 63rd Annual Technical Conference and Exhibition of the Society of Petroleum Engineers held in
Houston, TX, October 2-5, 1988.
This paper was selected for presentation by an SPE Program Committee following review of information. contained in an a~stract submit~ed by the
author(s). Contents of the paper, as presented, have not been reviewed ~~the Society ?f Petroleum Engmee.rs and ~re s~bJect to correction by the
author(s). The material, as presented, does not necessarily reflect any pos1t1on of the Soc1ety of Petroleum Engmeers, lts.offlcers, or ~e11_1bers. Pape~s
presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Perm1ss1on to copy 1s
restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment of
where and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836. Telex, 730989 SPEDAL.
r
miscibility, the oil recovery did not increase sig-
ABSTRACT nificantly beyond that achieved at the slim tube ME.
This paper presents the results of a composi- Simulation results for the
tional simulation study to investigate the relation- condensing/vaporizing system were also compared with
ship between the displacement mechanism, the minimum those predicted by a classical condensing mechanism.
enrichment requirement, and the oil recovery effi- These comparisons showed that: a) for a classical
ciency for a rich gas system. The simulation condensing mechanism, the thermodynamic and the slim
results indicated that the displacement mechanism tube methods predicted the same minimum enrichment
over a wide range of solvent enrichment was governed requirement; b) at low dispersion the two mechanisms
by a condensing/vaporizing process; however, the showed similar displacement behavior; and c) higher
relative importance of the condensing or vaporizing values of dispersion had a bigger impact in reducing
mass transfer processes, and their impact on the the displacement efficiency of the
overall displacement efficiency, was a function of condensing/vaporizing mechanism compared with the
the enrichment level. For displacements simulated classical condensing process.
at or above the slim tube minimum enrichment (ME)
level (determined from the recovery breakover point INTRODUCTION
on a recovery versus enrichment plot), recovery was
high and the process behaved very similar to a clas- Hydrocarbon miscible gas flooding has long been
sical condensing mechanism. Displacements simulated recognized as a viable enhanced recovery method by
below the slim tube ME were less efficient; however, the petroleum industry. To improve the economic and
both condensing and vaporizing mass transfer proc- technical success for projects involving rich gas
esses contributed to the overall recovery effi- injection, it is essential to have a good under-
ciency. At enrichment levels considerably above the standing of the displacement mechanism and minimum
slim tube ME, the effect of the condensing mass enrichment requirement for achieving an efficient
transfer process on the displacement became less displacement.
significant and the behavior approached that of a
classical vaporizing mechanism. Traditionally, 1 ' 3 - 6 the displacement mechanism
of rich gas drives has been assumed to be via a con-
The slim tube technique for determining the ME densing process in which the intermediate components
requirement based on simulation results was further of the solvent enriched the reservoir oil toward the
compared with two other computational methods: one point of miscibility. However, the work of Stalkup 7
based on Hutchinson and Braun's multiple contact in 1965 and recent investigations 8 ' 9 have challenged
mixing, 1 and the other based on displacement behav- this traditional concept for some rich gas'displace-
ior at the limit of zero dispersion. 2 For this rich ments. Zick 8 showed that a very efficient oil
gas system, the minimum enrichment level predicted recovery can be achieved through a combined
by these methods was significantly higher than the condensing/vaporizing mass transfer interaction
slim tube ME and corresponded to conditions of ther- between the reservoir fluid and the injected sol-
modynamic miscibility development by a classical vent. His study was based on limited phase behavior
vaporizing mechanism. Despite such a high enrich- experiments and equation of state (EOS) calculations
ment requirement needed to achieve thermodynamic for a rich gas system. Novosad et al., 9 using slim
tube displacements and EOS calculations, claimed
References and illustrations at end of paper. that many Canadian rich gas projects are not con-
35
AN INTERPRETATION OF THE DISPLACEMENT BEHAVIOR OF
2 RICH GAS DRIVES USING AN EQUATION-OF-STATE COMPOSITIONAL MODEL SPE 18061
36
SPE 18061 J. MANSOORI AND S. P. GUPTA 3
recently proposed by Stalkup. 2 The basic principle First, for all the rich gases in Figure 2, thermody-
of this technique 1s derived from the work reported namic miscibility is predicted to be achieved by a
previously by Dumore et al. 20 They ~hawed, using vaporizing mechanism. This is consistent with the
the method of characteristics, that .when a multiple predicted shape of the pressure-composition -(P-X)
contact process develops miscibility in a piston- diagram of Figure A-1 showing a critical point to
like displacement at the limit of zero dispersion, the left hand side of the maximum 2-phase pressure. 9
all the in-place oil should be recovered at exactly Secondly, as the enrichment level is increased,
one pore volume of the fluid injection. Since some thermodynamic miscibility is developed at a lower
level of numerical dispersion is always present in pressure. From Figure 2, the minimum enrichment
the finite difference approximation pf flow levels needed to achieve thermodynamic miscibility
equations, compositional simulators ca.n not exactly at 2500 and 3000 psi are about 68 and 58 mol% C2+,
predict the displacement behavior at zero disper- respectively. Compared with the predicted slim tube
sion. To overcome this numerical shortcoming, as ME's of 55 and 50 mol% C2+ at these pressures, it is
suggested by Stalkup, the simulated recovery at 1 PV observed that this rich gas system requires consid-
is calculated at several dispersion levels, and the erably higher enrichments (13 and 18 mol% higher C2+
results are extrapolated to zero dispersion. The at 2500 and 3000 psi, respectively) to achieve ther-
lowest enrichment level at which 100% oil is recov- modynamic miscibility than that predicted by the
ered after injection of .1 PV of the .displacing fluid slim tube method.
(at the limit of zero dispersion) determines the ME
requirement. The minimum enrichment calculations presented
above were also compared with the method of zero
A third method for determining the minimum dispersion proposed by Stalkup. 2 Figure 3 shows the
enrichment requirement widely used in the industry simulated recovery at 1 PV of solvent injection
is by conducting slim tube displacements for a given plotted against the square root of inverse Peclet
oil/rich gas system at several solvent enrichment number at various enrichment levels for the 2500 psi
levels. In this technique, the oil ~ecovery at displacement. Stalkup originally suggested using a
1.2 HCPVI is plotted versus solvent enrichment, and dimensionless gridblock size to represent dispersiv-
the point where recovery is high and it no longer ity. This was because in his simulations, the
increases significantly with increased enrichment numerical dispersion (governed by the number of
level determines the slim tube ME requirement. The gridblocks rather than the actual dispersion coeffi-
minimum enrichment requirement predicted by these cient which is used here) was used to account for
three methods are compared in this investigation for the effect of dispersion. Based on the extrapolated
the oil/rich gas system reported by Stalkup. 7 Other behavior shown in Figure 3, it is observed that a
empirical methods of determining solvent enrichment minimum enrichment of 68.5 mol% C2+ is required to
requirement, such as Benham I s corre 1 at1on,
• 3 are not
achieve 100% recovery at 1 PV of injection at the
included in this study. limit of zero dispersion. Hence, the criteria for
the development of thermodynamic miscibility based
DISCUSSION OF RESULTS on Hutchinson and Braun's 1 multiple contact mixing
agrees favorably with the method of zero dispersion.
1 - Effect of Gas Enrichment on Displacement Both of these techniques, however, predict ME
Behavior requirements significantly higher than the slim tube
technique. Similar results are expected for the
The compositions of the oil and the rich gas displacement at 3000 psi.
used in slim tube simulations were the same as oil 2
and rich gas 2 of Table 1 in Stalkup's paper. 7 The Although a significant difference exists
gas enrichment level was changed by adding or sub- between various methods of determining ME require-
tracting a 50/50 mixture of c2 and c3 from the base ment, it is noted from Figure 1 that increasing the
gas. Compositional simulations were then conducted solvent enrichment beyond the slim tube ME does not
at two pressures of 2500 and 3000 psi using gas significantly affect the oil recovery performance
enrichments in the range of 43 to 69 mol% C2+. A under slim tube conditions. The slim tube technique
Peclet number of 800 was used for all simulations to gives the smallest estimate of the solvent minimum
mimic slim tube conditions. Figure 1 shows the pre- enrichment requirement to achieve an efficient dis-
dicted oil recovery at 1.2 hydrocarbon pore volume placement for this rich gas system.
injected (HCPVI) at two displacement pressures of
2500 and 3000 psi plotted as a function of the 2 - Comparison With Experimental Data
enrichment level. It is observed that recovery
first increases steeply with enrichment and then Stalkup 7 also reported the results of a limited
levels of£. The intersection of th• two line seg- slim tube tests conducted in a sand-packed tube. At
ments drawn through the low and high recovery seg- 3000 psi, miscibility development was first observed
ments of the curves determines the simulated slim for a rich gas containing 50.5 mol% methane with an
tube ME requirement for each displacement pressure. intermediate distribution similar to the rich gases
From Figure 1, minimum enrichments p£ about 55 and used for other phase behavior studies. 7 No explana-
50 mol% C2+ are required for this rich gas system in tion was provided as to the criteria by which misci-
order to achieve an efficient displacement perform- bility determination was established. Nevertheless,
ance at 2500 and 3000 psi, respectively. comparing with the simulated performance at
3000 psi, it is noted that the experimental ME
These simulated slim tube ME's are now compared agrees favorably with the simulated slim tube ME,
with the thermodynamic ME's calculated from the rather than the thermodynamic ME, or that based on
algorithm of Turek et al. 19 Figure 2 shows the pre- the method of zero dispersion (Figures 1, 2 and 3,
dicted the.rmodynamic MMP versus solvent (C2+) mole respectively).
fraction. Two important observations are made.
37
AN INTERPRETATION OF THE DISPLACEMENT BEHAVIOR OF
4 RICH GAS DRIVES USING AN EQUATION-OF-STATE COMPOSITIONAL MODEL SPE 18061
38
SPE 18061 J. MANSOORI AND S. P. GUPTA 5
39
AN INTERPRETATION OF THE DISPLACEMENT BEHAVIOR OF
6 RICH GAS DRIVES USING AN EQUATION-OF-STATE COMPOSITIONAL MODEL SPE 18061
B - Comparison of the Displacement Performance at very efficient and the oil recovery was mainly con-
Low Dispersion trolled by a mass transfer process similar.to a
classical condensing mechanism; the effect of vapor-
The two mechanisms show similar behavior for ization was negligible. Displacements below the
displacements at low dispersion. The predicted oil slim tube ME were less efficient; however, both the
recovery performance and the oil saturation profiles condensing and the vaporizing mass transfer proc-
by the two descriptions under slim tube conditions esses influenced the recovery efficiency. Displace-
(Pe = 800) at an enrichment level of 58.5 mol% C2+ ments simulated at enrichment levels much above the
(above the slim tube ME, but below the thermodynamic slim tube ME had characteristics similar to a clas-
ME) are shown in Figure 14. For this low dispersion sical vaporizing mechanism.
simulation, the recoveries are comparable and the
only minor difference is in the residual oil satu- 2 - The slim tube ME predicted by the
ration left behind the front. The condensing/vaporizing mechanism was compared with
condensing/vaporizing process, because of the vapor- that obtained from the thermodynamic miscibility
ization effect, leaves a very small amount of resi- criteria of Hutchinson and Braun. 1 For this rich
dual oil behind, whereas the classical condensing gas system, a considerably higher enrichment was
process displaces 100% of the contacted oil. Over- required to achieve thermodynamic miscibility. The
all, the two mechanisms predict very similar dis- onset of transition from a condensing/vaporizing
placement behavior at low dispersion. mechanism to a classical vaporizing mechanism, as
predicted by the simulator, was in excellent agree-
C - Comparison of Displacement Performance at High ment with the thermodynamic ME.
Dispersion
3 - For the rich gas system studied, the ther-
At higher levels of dispersivity, the recovery modynamic ME agreed favorably with the minimum
efficiency predicted by the condensing/vaporizing enrichment needed to recover 100% of the oil at one
mechanism suffers more dramatically than that of the pore volume of solvent injection at the limit of
classical condensing mechanism. The simulated per- zero dispersion.
formance at 2500 psi and 58.5 mol% C2+ solvent
enrichment is compared in Figure 14 at a dispersion 4 - Despite requ1r1ng a much higher enrichment
level equivalent to Peclet number o£ 24. At this level to achieve thermodynamic miscibility, increas-
higher dispersivity, the oil recovery predicted by ing the solvent enrichment beyond the slim tube ME
the condensing/vaporizing process is somewhat more did not further increase oil displacement efficiency
affected than the classical condensing process. under slim tube conditions.
The differences in the predicted behavior can 5 - For a classical condensing mechanism, the
be explained by observing the shape of the in situ three methods of minimum enrichment determination
oil saturation profiles, as shown in Figure 15. (slim tube, thermodynamic criteria, and 1 PV recov-
Firstly, compared with the low dispersion simu- ery at zero dispersion) predicted the same enrich-
lation, the condensing/vaporizing process predicts a ment requirement.
higher level of ~esidual oil saturation behind the
gas-oil transition zone. In contrast, the classical 6 - Higher dispersion had a greater impact on
condensing mechanism achieves 100% recovery effi- reducing the displacement efficiency of a
ciency. The increase in the residual oil saturation condensing/vaporizing mechanism than that of a clas-
at higher dispersion predicted by the sical condensing mechanism.
condensing/vaporizing mechanism is consistent with
Stalkup observations. 2 A second factor affecting
the level of oil recovery at a given HCPVI for this
rich gas system is the difference in the distrib- REFERENCES
ution and magnitude of the oil saturation within the
transition zone. For the condensing/vaporizing pro- 1. Hutchinson, C. A., Jr., and Braun, P. H.,
cess, the saturation profile is more smeared and, "Phase Relations of Miscible Displacement in
hence, the lower oil phase mobility (due to smaller Oil Recovery," AIChE J. (1961), v. 7, 64.
oil saturation) within the 2-phase region results in
a lower cumulative oil recovery at 1.2 HCPVI. Based 2. Stalkup, F. I., "Displacement Behavior of the
on these observations, therefore, higher dispersiv- Condensing/Vaporizing Gas Drive Process,"
ity has a greater impact on reducing the displace- SPE 16715, presented at the 62nd Annual Techni-
ment efficiency of a condensing/vaporizing process cal Conference and Exhibition of the SPE held
compared with a classical condensing process. in Dallas, TX, September 27-30, 1987.
Whether or not these differences are significant
under actual field condition should be answered by 3. Benham, A. L., Dowden, w. E., and Kunzman,
conducting field scale compositional simulations. W. J., "Miscible Fluid Displacement- Predic-
tion of Miscibility," Pet. Trans. AIME,
SUMMARY AND CONCLUSIONS vol. 219, p. 229-1960.
40
SPE 18061 J. MANSOORI AND S. P. GUPTA 7
Total Oil Recovery," Pet. Eng. (Oct., 1956) 28, 18. Renner, T. A.: Amoco Production Company, Pri-
B-45 vate Communication.
6. Stalkup, F. I., Jr., "Miscible Displacement, 19. Turek, E. A., Luks, K. D., and Baker, L. E.,
Monograph Series, vol. 8, SPE of AIME, Dallas, "Calculation of Minimum Miscibility Pressure,"
Texas, 1983. SPE/DOE 14929 presented at the SPE/DOE Fifth
Symposium on Enhanced Oil Recovery held in
7. Stalkup, F. I., "Using Phase Surfaces to Tulsa, OK, April 20-23, 1986.
Describe Condensing - Gas - Drive Experiments,"
SPEJ, September 1965, 184-188. 20. Dumor~, J. M., Hagoort, J., Risseeum, A. S.
"An Ana 1 yt1cal
. Mod~l for One-Dimensional, '
8. Zick, A. A., "A Combined Condensing/Vaporizing Three-Component Condensing and Vaporizing Gas
Mechanism in the Displacement of Oil by Drives," SPEJ, Apr. 1984,
Enriched Gases," SPE 15493, presented at the
61st Annual Technical Conference and Exhibition 21. Turek, E. A.: Amoco Production Company, Private
of the SPE held in New Orleans, LA, Communication.
October S-8, 1986.
41
AN INTERPRETATION OF THE DISPLACEMENT BEHAVIOR OF
8 RICH GAS DRIVES USING AN EQUATION-OF-STATE COMPOSITIONAL MODEL SPE 18061
88158ART0144
42
SPE 18 0 6 1
TABLE A-1
Comparison of Calculated and Experimental Vapor-Liquid
Equilibrium Phase Compositlons7 ·
1.02 4000
a..
0
0 0.98
·a;
0 a.
co
a.t
0.94 a.: 3500
~
>
a..
0.90 ~
u legend 0
.EaJ 3ooo
:I: 0.86
• 2500 psi displacement c:::::
~ >-
....co 0.82
o 3000 psi displacement
"0
0
E
> 0.78 Q; 2500 - ------ - ----- - - -- - -,- --------- -
I
Q; .c I
0 .....
0 0.74
0
Q)
a: 0.70 2000,_----,-----.------.--~-.-----.---L--,
0.2 0.3 0.4 0.5 0.6 0.7 0.40 0.45 0.50 0.55 0.60 0.65 0.70
C2+ Mole Fraction C2+ Mole Fraction
Fig. 1-Siim·tube recovery as a function of solvent enrichment level. Fig. 2-Effect of enrichment level on thermodynamic MMP for a condens-
ing/vaporizing rich gas system.
43
SPE 18061
(0
CJ)
10
0.70 o~-.--.--.--.-~--~~--.--~~o
0.00 0.02 0.04 0.06 0.08 0.10 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
..J17Pe Fractional Distance
Fig. 3-EIIect of Inverse Peele! number (a/L) on oil recovery at1 PV Injection (2,500 psi) lor a condensing/vaporiz- Fig. 4-Displacement behavior at 2,500 psi and 63.5 mol% C 2 + enrichment (be-
Ing rich gas system. tween slim tube ME and thermodynamic ME) at 0.52 HCPVI.
0.86
c 1+N z+COz
0.60
c 0.66
0
·.;::; 0.60
u
co
0.46
..t
CD 0.40
0
~
0.36
c 0.30
0
.c 0.26 oil middle intermediates (C6-CJO)
:0
u 0.20 £i!....!jg!!.l...J!!termediat~2-C5L_ _
0
-o>- 0.16
9.C!.~ _'!!.isi..c!Lr::. }!l_t~!.'!:'.f!.c![g~t:~. {9_6_-:;_f; ;5_0)
:X: 0.10
gas lig_h_t_intermf!_qJgtes (C?-:;~§.2__
0.06
0.00
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Fractional Distance
Fig. 5-Compositlonal profiles ol various hydrocarbon fractions In a condensing/vaporizing system
at 2,500 psi and 63.5mol% C 2 + enrichment (between slim tube ME and thermodynamic ME) Fig. 6-Quaternary representation olthe displacement behavior at
.
at 0.52 HCPVI. 2,500 psi and 63.5 mol% C 2 + enrichment (between slim tube
ME and thermodynamic ME).
,__ ______
-'= Oil Saturation
'E> 0.9 ' Remaining OIP .c 70
I
+-'
~ 0.8
Ol
c
0.9
60
....J Q)
....J 0.8
Q) 0.7 +-'
c:: Q)
0.7 50 .......
:.:J 0.6 c ::J
I Vaporizing Condensing ::i 0.6 Oil Density
0
~ 0.5 Region <1---+-e..;.....r> Region I
Q) 40 ?i
0.4 i=
._
0.6
0.3
0.2
- - ,\,I
Tie Line / 0
g
0.4
0.3
·.;:; --····./,------lie Line
0.1
(0 0.2
~ -...._;····---·······-Gas Density 10
+-' 0.1
0~-.-----.----~--~~--.---~~ (0
CJ)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 04=~--.-----.-~--~~--.---~~o
Fractional Distance 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Fig. 7-Remalnlng ollln place lor a displacement at2,500 psi and 63.5 mol% Fractional Distance
C 2 + enrichment (between slim tube ME and thermodynamic ME) at Fig. a-Displacement behavior at2,500 psi and 43.5 mol% C 2 + enrichment (below
0.52 HCPVI. slim tube ME) at 0.39 HCPVI.
C1+N z+COz
,_._---I'
-'=
+-'
C) 0.9 Remaining OIP
c
Q) 0.8
....J
Q) 0.7
c::
:.:J 0.6
I
f Region
Condensing
Q)
0.5
i=
0.4
0
c:: 0.3 Tie Line
0
·;:; 0.2
0
(IJ
0.1
Lt
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Fractional Distance
Fig. 9-Quaternary representation olthe displacement behavior at2,500 Fig. 10-Remainlng oil saturation lor a displacement at2,500 psi and 43.5 mol%
psi and 43.5 mol% C 2 + enrichment (below slim tube ME). C 2 + enrichment (below slim tube ME) at 0.39 HCPVI.
44
SPE 18 0 6 1
Ct+N z+COz
a..
6
0 0.9
u
10
L.t 0.8
>
a.. 0.7
Dew Point u LEGEND
::I:
Curve • Condensing/Vaporizing Mechanism
~
0.6
Overall
Composition
Path -
~ -
10
>-
CD
0.6
o Classical Condensing Mechanism
Curve a: 0.3
0.2 0.3 0.4 0.5 0.6 0.7
Cz+ Mole Fraction
Fig. 11-Quaternary representation of the displacement behavior at 2,500 Fig. 12-Comparlson of oil recovery vs. enrichment level for a condensing/vaporizing and a classical condensing
psi and 68.5 mol% enrichment (above thermodynamic ME). mechanism at 2,500 psi.
a.
6
0
u 0.96 --- ---~ Condensing/Vaporizing Mechanism
10
L.t --------~ Classical Conden~ Mechanism_
0.90
>
a.
u
::I: 0.86
a..
3 MOL% C2+
i5
0
0.9 0.9
a; 0.80 0.8 c 0.8
> o 53.5 (Blow Thermodynamic ME) c 0.7 0 0.7
0 ·.;::;
CD
0 0.76 t, 55 (At Thermodyn6mic ME) ·g 0.6 co
:; 0.6
0 co 0.5 0.5
0 +-'
Q)
a: U: 0.4
co
(/) 0.4
0.70+---~--~---~----.----~
Condensing/Vaporizing Mechanism
6200
Classical Condensing_ Mechanism_
6000 Experimental
4800
P.r.t:~ !~ ~t:.<1.. r; ~ 0.c;i.~.r:.!?!r!9.!XP.I?. ~~!?. ~r:.9..¥~~~.C?.r:.i~ r:!
4800
a..
0.9
,--- 4400
Predicted Classical Condensi~echanism_
i5 0.9
0
c
0.8 c 0.8 I Ci5
4200
x Critical Point
0
·.;::;
0.7
0.6
0
·.;::;
co
0.7
0.6
I a..
u1
4000
3800
u a:
co
L.t
0.5 E
co
0.5 ( ::::>
(/)
3600
0
0.4
0.3
(/) 0.4
0.3 I (/)
UJ
a:
3400
3200
6 a..
(J)
a: 0.2
0.1
0.2
0.1
I 3000
6 0 0 --,
2800
2800
0 0.4 0.8 1.2 1.6 2 2.4 2.8 0 0.2 0.4 0.6 0.8 1
HCPV Rich Gas Injected Fractional Distance 2400
2200
Fig. 15-Comparlson of a classical condensing and a condensing/vaporizing mechanism at high dispersion, 2,500
psi, Pe=24, 58.5 mol% C 2 + enrichment (above slim tube ME for the condensing; between the slim tube
ME and thermodynamic ME for the condensing/vaporizing system). 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
SOLVENT FRACTION
Fig. A·1-Comparlson of the P·X diagrams.
45