You are on page 1of 17

ELSEVIER Journal of Petroleum Science and Engineering 18 (1997) 113 - 129

S t e a m- CO 2 dri ve experi ment s usi ng hori zontal and verti cal wel l s
F. Giimrah *, S. Ba~cl
Petroleum and Natural Gas Engineering Department, Middle East Techmcal University, 06531 Ankara, Turkey
Recewed 4 July 1996; accepted 17 January 1997
Abs t ract
Research into the application of a simultaneous s t e a m- CO 2 dri ve process and the exami nat i on of vertical and horizontal
i nj ect i on- pr oduct i on wel l confi gurat i ons was conduct ed in a physical model of 1/ 12t h of an i nvert ed regul ar seven-spot
pattern to det ermi ne the r ecover y performance of 12.4API heavy oil. Three groups of wel l confi gurat i ons were mai nl y
investigated: a vert i cal i nj ect i on and production wel l s scheme (group 1), a vert i cal i nj ect i on and horizontal product i on wel l s
scheme (groups 2A and 2B), and a horizontal injection and product i on wel l s scheme (groups 3C and 3D). A total of 17
experi ment s of whi ch havi ng fi ve steam-alone and t wel ve s t e a m- CO 2 processes were conduct ed for the above wel l
configurations.
In st eam-al one tests, the vert i cal i nj ect or and hori zont al producer scheme (group 2B) supplied a hi gher r ecover y than that
of the others. The oil r ecover y was 33.6% of original oil in pl ace (OOIP) in group 2B compared to 7.8% of OOI P for the
vert i cal i nj ect i on and product i on wells scheme (group 1) at 1 pore vol ume (PV) of steam injected. The l owest ultimate
r ecover y was obtained from the horizontal i nj ect or - hor i zont al producer wel l confi gurat i on (group 3C).
For s t eam- CO~ tests, oil recoveri es were 58.3% and 25. 3% of OOI P for a CO2/ s t e a m ratio of 14.2 dm3/ 1 in group 2A
and the horizontal i nj ect or and producer (group 3C) with a CO2/ s t e a m ratio of 13.4 dm3/ 1, respect i vel y. The co-injectaon of
CO 2 with steam i ncreased the ultimate oil r ecover y and the product i on rate over steam alone. The r ecover y effi ci ency of
horizontal i nj ect or - hor i zont al producer (group 3C) was also the l owest one, but vertical i nj ect or - hor i zont al producer (group
2A) gave the best performance when compared to other tests.
When steam-alone and s t e a m- CO 2 tests were compared, the oil recovery i ncreased with i ncreasi ng CO2/ s t e a m ratio till
an opt i mum value was reached, after whi ch a di mi ni shi ng effect was observed. The opt i mum CO2/ s t e a m ratio for
maxi mi si ng oil r ecover y was ~ 14 dm3/ 1 for all wel l configurations. Therefore the val ue of CO2/ s t e a m ratio was one of
the important factors whi ch affected the performance of the process. The ot her factor whi ch i nfl uenced the oil r ecover y was
the wel l type of i nj ect or a n d / o r producer whet her it is horizontal or vertical. The distance bet ween the wells also affect ed
the effi ci ency of the process. The primary mechani sms for the mobi l i sat i on of oil were vi scosi t y reduction, steam distillation,
stripping and gas drive effect of CO 2.
Keywords: steam-CO 2 drive: physical model; vertical-horizontal wells, heavy-oil recovery
1. I n t r o d u c t i o n
* Corresponding author. Fax:
Fevzi @ rorqual.cc.metu.edu.tr
90-312-2101271; e-maih
He a v y - o i l r e s e r voi r s pr e s e nt p r o d u c t i o n p r o b l e ms
b e c a u s e t he hi gh oi l v i s c o s i t y and l o w r e s e r v o i r
e n e r g y r es ul t i n l o w r e c o v e r y r at es a nd p o o r r e c o v -
0920-4105/97/$17.00 Copyright 1997 Elsevier Science B.V. All rights reserved.
PII S0 9 2 0 - 4 1 0 5 ( 9 7 ) 0 0 0 0 3 - X
114 F. Gfimrah, S. Ba~ct / Journal of Petroleum Science and Engmeermg 18 11997) 113-129
ery efficiency when compared to those of conven-
tional oil reservoirs. The most widely used thermal
technique to extract residual oil from heavy-oil fields
is steam injection (Chu, 1985). There is an interest in
using CO 2 gas as immiscible phase for recovering
heavy oil (Roadifer, 1986). In heavy-oil reservoirs, a
combination of steam and non-condensable gases
can be used to increase heavy-oil production. In
Turkey 80% of the oil reservoirs are heavy-oil reser-
voirs (Kantar and Topkaya, 1983). These heavy oils
are usually mobile at reservoir conditions and most
can be produced with primary production but recov-
ery is very low. The Bat1 Kozluca Field which has
98 106 stb crude oil with 12.4API is located in
the southeastern region of Turkey. The presence of
the Dodan natural gas field having 93% of CO~
close to the oil field may favour the use of the CO 2
injection process with steam. In the following para-
graphs, the results of field tests, simulation studies
and laboratory experiments for the combined use of
steam and gases are reviewed.
Steam-air stimulation field tests were reported
by Rintoul (1979) and Meldau et al. (1981). The
improvement in oil recovery over conventional steam
stimulation was recorded in these applications. Sperry
(1981) reported the results of the Vapour Therm TM
system in three different fields in the U.S. mid-conti-
nent region. An improvement in oil/steam ratio was
achieved. The steam-generating systems such as
downhole steam generators (Fox et al., 1981), the
Vapour Therm TM process (Sperry, 1981). and the
wet-air oxidation technique (Wilhelmi and Knopp,
1979) consider the injection of produced non-con-
densable gases with high-quality steam into the
reservoirs. The injection of gases together with steam
is believed to benefit the recovery process by the
presence of non-condensing gas phase. Schirmer and
Eson (1985) reported the concept of using a direct
fired downhole steam generator in thermal oil recov-
ery projects in the Kern River Field, California. The
oil/steam ratio and peak production rate obtained
from downhole the steam generator were higher
compared to previous responses with conventional
steam injection.
Numerical reserL:oir simulators have been used to
predict the effect of various conditions on the use of
non-condensable gases with steam for oil recovery
(Weinstein, 1974; Fox et al., 1981; Meldau et al.,
1981; Balog et al., 1982; Leung, 1982; Claridge and
Dietrich, 1983; Stone and Malcolm, 1985a,b). Co-in-
jection of a non-condensable gas with steam acceler-
ated the oil production compared to the steam-only
case. The ultimate recovery in a certain period was
about the same for both cases, or higher for the
co-injection case, depending on operating and reser-
voir conditions. This indicates that an interpretation
of simulation studies should be made with awareness
of the conditions studied.
The results of laborato~ experiments which have
been conducted to study the expected beneficial ef-
fect of co-injecting a non-condensable gas together
with steam are summarised in the following para-
graphs. Ozen (1967) conducted a laboratory experi-
ment to test the effect of co-injection of nitrogen gas
on steam flood residual oil saturation. It was found
that N2-steam flooding increased the oil recovery by
~ 4- 5% more compared to the steam-only case.
This result was attributed to the presence of gas in
the core which would aid in the distillation of crude
leading to increased oil recovery. Slobod and Mer-
riam (1969) investigated the contribution of factors
as gas drive and vaporisation to improve the effi-
ciency of displacement of oil by steam flooding. But,
hot water was used to introduce heat into the system,
and nitrogen was used to provide a vapour space in
the porous system to isolate the actions produced by
heat alone and vaporisation alone. Pursley (1975)
carried out experiments to investigate the effect of
injecting air, methane, or CO 2 on steam stimulation.
A dramatic improvement in the oil/steam ratio was
observed as a result of injecting methane or air. The
addition of CO~ was somewhat less effective be-
cause of its high solubility in water. Fox et al. (1981)
conducted laboratory experiments to examine recov-
ery with soluble gas-steam drive using core sam-
ples. It was found that soluble gas-steam drive
recovered more rapidly than the steam-only case.
Redford (1982) reported the effects of adding CO 2
or ethane to steam in a 3-D physical model. Adding
of CO 2 or ethane to steam greatly improved the
recovery of Athabasca tar sand over that recovered
with other additives. This was attributed to a solution
gas drive effect which produced the fluid from the
cooler portion of the reservoir. Briggs et al. (1982)
presented the results of a 1-D physical simulator of
cyclic steam injection with CO~ and naphtha addi-
F. Giimrah, S. Ba~ct / Journal of Petroleum Sctence and Engineering 18 (1997) 113-129 115
rives. The experiments were done with Athabasca tar
sand. The use of CO 2 with steam improved recovery
primarily by providing additional drive energy on the
depletion portion of the cyclic process. Harding et al.
(1983) reported results of a physical model study of
steam flooding with nitrogen and CO 2 additives
which were injected into a linear porous medium
saturated with a moderately viscous refined oil and
water. It was observed that the simultaneous injec-
tion of the gases with steam resulted in a significant
improvement in the ultimate recovery of the crude
oil. Paracha (1985) studied the effects of CO~ addi-
tion to steam in a 1-D laboratory model on heavy
oils. The results indicated that although CO: with
steam increased the rate of recovery significantly, the
oil viscosity and hence the API gravity were en-
larged. Stone and Malcolm (1985b) conducted high-
pressure steam-CO, co-injection experiments in a
1-D physical model with Athabasca tar sand. The
results from the experiments were compared with
results from a numerical model study. Both models
gave results that co-injection of CO z and steam
increased ultimate recovery. Stone and Nasr (1985)
analysed the use of steam-CO 2 and steam-N 2 mix-
tures in a set of continuous injection experiments in
a test bed. The addition of CO 2 to the steam resulted
in a significant change in the displacement mecha-
nism. An enhanced bitumen stripping and the forma-
tion of a gas zone around the injection well resulted
in increased conformance in the test bed and in-
creased bitumen production, as well as in bitumen
viscosity reduction due to dissolved CO 2 and an
increased pressure gradient between the injection and
production wells. Stone and Ivory (1987) carried out
steam-CO 2 experiments in a large pressure oil sand
vessel. CO 2 pre-soak followed by steam injection,
steam-CO 2 co-injection and steam, CO, and solvent
co-injection experiments were employed to under-
stand the mechanisms behind recovery processes. In
all cases investigated, the addition of CO~_ to steam
resulted in improved utilisation of injected energy
and improved the oil recovery over that from steam
alone. Nasr et al. (1987) studied the effects of steam,
steam-CO 2, steam-N 2 and steam-COR-N 2 mix-
tures on bitumen recovery from oil sands by using a
3-D physical model. The test results showed that the
addition of flue gas to steam substantially improved
both rate and ultimate recovery of bitumen as com-
pared to that obtained by steam alone. The steam-
CO 2 mixture was superior to either the steam-N 2 or
steam-flue gas combinations. Giimrah and Okandan
(1987) studied simultaneous st eam-CO, injection
processes in a 1-D physical limestone pack model
with heavy oils. The results indicated that an opti-
mum COJ s t eam injection ratio was present. The
steam-CO 2 process accelerated production of heavy
oil above that of the steam-only case. Doscher et al.
(1988) conducted scaled physical model experiments
to investigate the advantage of using high-velocity
gas injection for recovering reservoir fluids, the in-
jection of gas and steam after steam breakthrough for
increasing the profitability of some steam drives and
the use of specially fractured horizontal wells to aid
in the profitable production of viscous hydrocarbons.
Frauenfeld et al. (1988) conducted physical model
experiments to study the effects of a steam injection
process. For oils without an initial gas content, co-in-
jection of CO 2 with steam was capable of improving
oil recovery over that obtained with steam alone.
When an initial dissolved gas was present, co-injec-
tion of CO 2 was not beneficial. Injection of CO 2 or
CH 4 slugs just before steam injection was beneficial
in increasing oil recovery for experiments where an
initial dissolved gas was present. Metwally (1990)
conducted a laboratory program for the Lindbergh
Field, Alberta, to investigate the effect of CO~ and
methane on the performance of steam processes. The
results indicated that the presence of a non-con-
densable gas improved steam injectivity. Injectivity
improvement was most pronounced when a gas slug
was injected prior to steam injection, but the pres-
ence of a non-condensable gas with steam did not
improve recovery and resulted in much higher resid-
ual oil saturation compared to steam injection alone.
Hornbrook et al. (1991) carried out a high-pressure
1-D laboratory displacement study to evaluate the
effects of adding CO 2 to steam on the recovery of
West Sak crude oil. It was found that adding CO 2 to
steam improved the recovery and recovery rate of
the crude over conventional steam flooding. Giimrah
and Okandan (1992) conducted steam-CO 2 experi-
ments in 1-D and 3-D laboratory models to evaluate
the benefits of CO 2 addition to steam on the recov-
ery of heavy oils. The linear tests indicated that the
oil recovery increased with increasing CO2/steam
ratios until an optimum value was reached. Light-oil
116 F. Gi i mrah, S. Ba~ct / J o u r n a l o f Pe t r o l e u m Sc i e nc e a n d Engi ne e r i ng 18 ( 1997) 1 1 3 - 1 2 9
recovery was relatively less i mproved by the addi-
tion of CO 2 to the injected steam; however, the oil
production rate was increased considerably for all
oils. The production of lighter-oil fractions increased
with increasing CO 2 concentration and API gravity.
Nasr and Pierce (1995) used a high-pressure and
high-temperature scaled model to evaluate oil recov-
ery processes by a series of experiments on s t e a m-
CO 2 injection strategies for bot t om wat er reservoirs.
The co-injection of CO 2 with steam accelerated and
i mproved oil recovery rates as compared to steam-
only injection. The s t e a m- CO 2 continuous injection
resulted in a better performance than that from
steam-only or s t e a m- CO 2 sequential injection.
St eam-onl y injection resulted in a dramatic i mprove-
ment in oil recovery as compared to hot wa t e r - CO 2
injection.
No experimental data known to us have been
published on steam flooding in the presence of CO 2
for Batl Kozl uca crude oil. Also, there has no com-
parative study been made on the performance of
s t e a m- CO 2 processes by the use of horizontal and
vertical i nj ect i on-product i on wells. Because of the
amount of heavy oil in reserve and the presence of a
natural CO 2 source, it was decided to study the
effects of adding CO 2 to steam on the recovery of
Bat1 Kozl uca heavy oil at laboratory conditions.
2. Experimental apparatus and procedure
2.1. Laboratory model
In recovery processes, the injection and producing
wells can be arranged in some type of pattern. From
. d .
@
p r o d u c l n g w i l l s
l l n ( m o f
_ _ l~,llarn b o u n d ~ l ~
d v ~ l l s p a e l n g
Fig. 1. Invert ed regul ar seven- spot well array.
this pattern a small el ement can be derived by sym-
met ry to represent the flood in a model study. The
flood performance for the entire pattern can be de-
vel oped by reproducing the performance of the pro-
cess in this small element. The smallest element
from the pattern that can be used as a model is
derived by constructing lines through all planes of
symmet ry. These lines of symmet ry represent invari-
ant lines across which there is no flow. In this work,
the inverted regular seven-spot pattern is studied.
This pattern consists of one injection well sur-
rounded by six producing wells. Fig. I shows the
arrangement of wells in a hexagonal array. The use
of this symmet ry to divide the pattern into smaller
elements is also shown in Fig. 1. The shaded area is
the smallest modelling el ement obtained from this
pattern and represents the model shape used. The
injection well of the model actually represents 1/ 12
of an injection well in the pattern, where as the
Table 1
The properties of physi cal and prototype model s
Pattern Scaling paramet er Prototype Physi cal model
1 / 1 2 t h of i nvert ed regul ar 7-spot 1 / 1 2 t h of i nvert ed regul ar 7-spot
Di st ance bet ween injection L p / L m = a = 250 202.5 m
and produci ng wel l s
Thi ckness Hp / H m = a 25. 0 m
Permeability K p / K m = l / a 40. 0 mD
Porosity 1 25.4%
Temperat ure - 50.0C
Oil vi scosi t y 1 607. 0 mPa s
Ti me t p / t m = a z 43. 4 days
Pressure drop A p p / A Pm = a 3447.5 kPa
Injection rate Qp / Qm = a 14.4 m3 / d a y
81.0 cm
10.0 cm
10.0 D
38.0%
50. 0C
607.0 mPa s
1 rnin
20.7 kPa
40. 0 c m3 / mi n
F. Giimrah, S. Ba~cl / Journal of Petroleum Science and Engineering 18 (1997) 113-129 117
WET TEST HEATER
METER HEATER CONTROLLER TEMPERATURE SCANNER CONTROLLER
~ i i - ~ ; 1 ~ 1 I ] " 1 , ' t ' [ I ~ -ROTAMETER (~,J2)~
',I .-'" ,',,/i k ",:,:,,','.~..... " : co
r - - T , - , T , , , , , _ l . . . . . . . . . o o L _ J
b : -. : : :-'/- . . . . . . . . . . . . . . . . . . . . . J STEAM
SEPARATORS 3-D MODEL GENERATOR
Fig. 2. A s c he ma t i c r e pr e s e nt a t i on of t he e xpe r i me nt a l set -up.
Group 1 :vertical injection-vertical production
Injection well (vertical)
. ~ T l e n g t h = 9.7 cm
a = 40.5 cm ~ perforation= 4 x 5 mm
P r o d u c t i o n w e l l (vertical) ~l ~_ ~, , / ~'~
length = 9,7 om
perforation= 4 x 5 mm h = 10ore
produotlon c = 81 m Injection
Well well
Group 2 : vertical injection-horizontal production
Group 2 A, / " P~
length =40,5 m / ~ ~ T
perforation= 10 x 2.5 mm f / -
r.5 ~AI ~ lil
~ r o d u o t l o n InJeotlon
w e l l / . ~ w e l l
Group2B. , / " ~ ~
7.5
~ " ~ o d u ~ o n Inj~=tlon
Well well
G r o u p 3 : horizontal injection-horizontal production
/ - Injection well (horizontal)
j " " " ~ length =39,2 cm
Group 3C, / / ~ p e r f o r e t l o n = 10 x 2.5 mm
inj~"tion
/
Group 3D. / / ~ o~,,~
7.5 ~ L ~
produo~on inJeo~on
well w~l
Fi g 3. Wel l c onf i gur a t i ons .
[ 18 F. Giimrah, S. Ba~,cl/Journal of Petroleum Science and Engineering 18 (1997) 113-129
pr oduct i on wel l of the model represent s 1 / 6 of a
pr oduct i on well. The model was t ri angul ar in shape.
A scal i ng f act or of 250 was ascert ai ned to corre-
spond to the pr ot ot ypes in the field. Puj ol and
Bobe r g' s (1972) t hermal scaling appr oach was used
in the desi gn of the scaled model . The propert i es of
the scal ed model and its cor r espondi ng pr ot ot ype are
gi ven in Tabl e 1.
A schemat i c represent at i on of the experi ment al
set-up is illustrated in Fig. 2. The set-up is compr i sed
of f our mai n parts: a st eam and CO 2 i nj ect i on sys-
tem, a physi cal model , a dat a r ecor di ng syst em and a
pr oduct i on system. The well confi gurat i ons are de-
l i neat ed in Fig. 3. The model has a di mensi on of 81
c m 70.1 c m x 40. 5 cm wi t h a t hi ckness of l 0
cm. The t op of the model was r emovabl e and acts as
a fl ange so that water, oil and crushed l i mest one
mi xt ure can be packed easily. To measure the three-
di mensi onal t emperat ure distribution inside the
model , 62 t her mocoupl es were installed at the top,
centre and the bot t om planes of the model . Insul at i ng
materials and heaters were also used in the model .
2.2. Experimental procedure
The pr emi xi ng met hod was used in prepari ng the
unconsol i dat ed l i mest one pack mi xt ures f or the ex-
peri ment s. The wat er and cl ean crushed l i mest one
were mi xed initially to meet the condi t i ons of a
wat er - wet system. Then the oil was mi xed homoge-
neousl y wi t h t hem to yi el d the desi red fl ui d satura-
tions and careful l y packed i nt o the model . The oil
and wat er saturations were chosen as 75% and 25%,
respect i vel y, and kept the same f or each experi ment .
The l i mest one pack wi t h pore vol ume of 5373 cm 3
gives 38% por osi t y and absol ut e liquid permeabi l i t y
of 10 darcies. Af t er packi ng, the model was moved
hori zont al l y inside t he i nsul at i on j acket , the set-up
was t hen prepared f or the test and the model was
heat ed appr oxi mat el y to 50C whi ch was t he desi red
reservoi r t emperat ure. To initiate an experi ment , the
st eam generat or was br ought to its maxi mum t em-
perat ure and the mass fl ow cont rol l er was set to the
desi red CO 2 fl ow rate. The si mul t aneous i nj ect i on of
st eam and CO 2 was started. Temperat ure distribution
inside t he model was cont i nuousl y registered. The
ot her paramet ers t hat were r ecor ded t hr oughout t he
tests were fluid i nj ect i on and pr oduct i on pressures
O_
g
O3
o
o
>
. J
o
1 0 0 0 0
1 ooo
1 O0
0
I
BATI KOZLUCA CRUDE OIL
\
1'0 2'0 3'0 4'0 g o 6'0 i o 80
TEMPERATURE (C)
Fig. 4. Vlscosiues of Batl Kozluca crude oil (12.4API).
and oil, wat er and gas pr oduct i on data. The effl uent s
f r om the model were col l ect ed in a t wo-st age separa-
t i on syst em. Bot h of t hem are operat i ng at at mo-
spheric pressure. The t op of the second separat or was
connect ed t o a wet test met er t o measure the amount
of gas produced. To cont rol the pressure of the
pr oduct i on well, back pressure regul at ors were lo-
cat ed at the fluid st ream end of the first separat or
and at t he upst ream end of the first separator. The
back pressure regul at ors were adj ust ed to a pressure
whi ch was ~ 13.8 kPa ( Ap = 2 psi) l ower t han t he
value of i nj ect i on pressure. The lighter-oil fractions,
whi ch were t er med condensat e duri ng this work,
were pr oduced f r om the second separator. Bat1 Ko-
zl uca crude oil (12. 4API) f r om sout heast ern Tur key
was used. The vi scosi t y of the oil sampl e is illus-
trated in Fig. 4.
3. Re s ul t s a n d d i s c u s s i o n
A total of 17 experi ment s were conduct ed wi t h a
physi cal model represent i ng 1 / 1 2 t h of an i nvert ed
regul ar seven-spot pattern. The ai m of the present
st udy was to i nvest i gat e the per f or mance of st eam
f l oodi ng in the presence of CO 2 f or heavy- oi l r ecov-
ery by usi ng vertical and hori zont al wel l confi gura-
tions. The experi ment al condi t i ons are present ed in
Tabl e 2.
The experi ment s were conduct ed under three mai n
F. Gi~mrah, S. Ba~ct / Journal of Petroleum Science and Engineering 18 (1997) 113-129 119
Table 2
Experimental conditions
STEAM INJECTION STEAM - C02 INJECTION
Well c o n f i g u r a t i o n co=/ , . ~ . ~ co= c o= . . ) = o . ~ co= ~ , . ~=~. = . . . . co= co=/ ,.~=~o. = . . . co=
P r m l J m Ri ce R i t e :Stwrn P r l l l l J r e R l a R J I ~ p r u i J r l R i l l ~ ~ P r l l i J r l R a l R i m
(1112th of inverted regular 7-spot pattern) , a-,*o R-,,o ~=o Rmo
[dm3,4.] [ k P I ] [c.tmin][c/mtn]l [dm3/~] [ k P i ] [/~ln][c,'mk'l] fdm3/I.] [ k P I ] [c/mt~][cc#nln].[~] [I<P] [/mln][c#nln]
. _ ,, - - . . . . .
t
Group.l vertical tnjectlo n-verti~,al~productlon 0.0
= L / ' ~ ~ 0.0 351.7 38,3 84,3 365,4 33.8 2850, 22. 3 368.9 71.9 1602
produdlonwell/ InJectlonwell 3145 14.2
Group.2 vettioal InJec, tlor',.horizontal procluctlon
GroopaA. 0.0 372.3 54.1 0.0 38.4 377.2 s e. 3 377.2 72. 1058 29.8 53.81
p r o d u c l o n w e l l Inj~--.~on w e l l
Group 2B. . / . ' ~
p r o d u c t i o n w e l l I n j e c t i o n w e l
Group.3 horizontal Injection-horizontal production
Group 3(3. / ~
p r o d g c l o n w e l l I n j e c t i o n "t~dl
Group 3D. / '/ ~
p r o d u c t i o n w e l l Inject!on we l l
0.0 379.2 38,2 0,0
0,0 379,2 44,5 0,0
0.00 379.2 35.5 0,0
140 385.4 17.0 2380
49,7 401.3 63.3 3145
38.1 365.4 73.7 2808
24.3 337.9 65.8 1602
13,4 411,6 78,9 1056
11.7 377. 1 90,4 1066
14.1 372.3 66,9 945
groups of well configurations. These are vertical
injection-production wells, vertical injection-hori-
zontal production wells, and horizontal injection-
production wells. The experiments were chosen so
that the comparisons could be made between steam
flood performance with the addition of CO 2 and
steam flood performance without CO 2. Performance
comparisons were made on the basis of overall oil
recovery.
3.1. General observations
3.1.1. Group 1, vertical injection-vertical produc-
tion scheme
In this group, one steam-alone and two steam-
CO, tests at a CO2flsteam ratio of 22.3 and 84.3
l oo-
9 0 -
E 80-
5
O 70-
o~
v 60-
>-
Lu 50-
>
O 40-
0
I L l
n- 30
._1
5 2o 20-
10-
0
C0 2 t S T E A M RATIO = 1 ~
0.5 1 1.5 2
STEAM INJECTED (PV of cwe)
2.5
Fig. 5. Oil recoveries for vertical i nj ect i on-vert i cal production
scheme (group l).
120 F. Giimrah, S. Ba~cl / Journal of Petroleum Sctence and Engmeermg 18 (1997) 113-129
dm3/1 were carried out. Fig. 5 compares the experi-
mental recoveries as a function of pore volumes of
injected steam as cold water equivalent (cwe). The
highest recovery was obtained with the CO2/steam
ratio of 22.3 din3/1. The oil recovery was the same
for the CO2/steam ratio of 84.3 dm3/l when com-
pared to the steam-only case. This result was at-
tributed to the highest amount of injected gas with
steam. As a result it might prevent the movement of
oil through the production well because of higher
mobility of gas. For the steam-alone test, the recov-
ery was increased after 1 PV of steam injection. The
22.3-dm3/1 CO~-steam mixture recovered 34.4% of
OOIP and the steam-only case recovered 7.8% of
OOIP when 1 PV of steam was injected into the
model. The recovery of the steam-only case was
reached to 33.6% of OOIP when the injected amount
of steam was 1.68 PV. The steam/oil ratios were
6.83 cm3/ cm 3 for the steam-only case, 18.83
cm3/ cm 3 for the CO2/steam ratio of 84.3 dm3/1,
and 4.31 cm3/ cm 3 for the CO2/steam ratio of 22.3
dm3/l at the end of the tests. Therefore, a lower
amount of steam was required to produce the same
amount of crude at the CO2/steam ratio of 22.3
dm3/1.
3.1.2. Group 2; t,erticaI injection-horizontal produc-
tion scheme
In this group of tests, a horizontal production well
was placed at two different locations of the model.
Therefore the experiments were done under two
subgroups, 2A and 2B. Scheme A includes a hori-
zontal production well along the shorter side of the
model and scheme B has a horizontal well located at
the longer side of the model (Fig. 3).
3.1.2.1. Group 2A. A total of four experiments were
conducted. The CO2/steam ratios were 14.2, 29.8
and 36.4 din3/1, and a steam-alone test was done to
compare the performance of the steam-CO 2 experi-
ments. Fig. 6 shows the results of these experiments.
The highest recovery was obtained for the
CO2/steam ratio of 14.2 dm3/I. The steam-only
case supplied the lowest recovery. The recoveries of
other tests were ranged between them. Oil recoveries
were 58.3% of OOIP for the CO2/steam ratio of
14.2 dm3/1, 43.2% of OOIP for 29.8 dm3/1, 41.9%
of OOIP for 36.4 dm3/l, and 23.8% of OOIP for the
100
90 t GROUP 2A
~" 80
oO 70 [ CO.STEAM F~T[ O =
60
[u 50-
0 4 0
0 _ , ~
w
r r 30-
O 20-
10-
0
0 0.5 1 1.5 2 2 5 3 5
STEAM INJECTED (PV of cwe)
Fi g 6. Oi l r e c ove r i e s f or ver t i cal i n J e c t i o n - h o r i z o n t a l p r o d u c t i o n
s c h e me ( g r o u p 2 A)
steam-only case. The steam/oil ratios at 1 PV of
steam injection were 5.6 cm3/ cm 3 for the steam-
alone test and 2.32 cm3/ cm 3 for the 14.2-dm3/1
CO2/steam ratio test. In all CO2-steam experi-
ments, the oil production rates were high and tapered
off rapidly after 1 PV of steam injection. The pres-
ence of a horizontal producer prevented the early
production of CO 2 gas which was accumulated at
the top of the model. Then, it exerted an extra force
to drive the heated oil through the production well.
The contact area of the horizontal production well is
four times larger than that of the vertical well. This
effect is also an important factor for higher oil
production rate.
3.1.2.2. Group 2B. In this group, the production well
was placed along the longer side of the model, it has
the same length of the production well of group 2A.
A total of four experiments were done. One of them
is steam alone and the others are CO2-steam experi-
ments in which the CO2/steam ratios were 14.1,
24.3 and 140 din3/1. Fig. 7 shows the oil recoveries
of these tests. For the 14.1-dm3/1 test, the highest oil
recovery was obtained as 52.4% of OOIP. The re-
coveries were 36.0% of OOIP for 24.3 dm3/ k 28.8%
of OOIP for 140 dm3/l, and 33.6% of OOIP for the
steam-only case. The highest COJ s t eam ratio case
recovered lower oil than that of the steam-only case.
The presence of a larger amount of CO 2 supplied
worst performance than the other cases. For higher
CO2/steam ratio, a larger amount of CO 2 was pro-
F. Giimrah, S. Ba~ct / Journal of Petroleum Science and Engineering 18 (1997) 113-129 121
1 0 0 -
9o-
E " 8 o -
O 70-
uJ 50-
O 40- / ~
O _ - - - - - I ~ [ o I
L _ J
r r 30-
---_1
O 20"
10 i ~ . + . ~ , , ,
0
0 0. 5 1 1.5 2 2. 5
STEAM I NJECTED (PV of cwe)
Fig. 7. Oll recoveries for vertical i nj ecUon-hori zont al production
scheme (group 2B)
100
GROUP 3C
9 0 -
~" 8 0 -
0 70-
6 0 -
> , .
r r
50-
>
0 40- ~
W
~ 3 0 -
- 2 0 - ~ 0
10-
0 :,,.~-~-- ~ , , , ~
0 0. 5 1 1.5 2 2
STEAM I NJECTED (PV of cwe)
Fig. 8. Oll recoveries for horizontal i nj ect i on-hori zont al produc-
tion scheme (group 3C).
duced, and as a result, steam followed the path
already swept by the CO 2. This result has also
pointed out the importance of using optimum mix-
ture of steam and COt in the tests. The steam/oil
ratios at 1 PV of steam injection were 2.60 and 3.97
cm3/ cm 3 for the steam-CO 2 (14.2 dm3/1) and
steam-alone tests, respectively. The co-injection of
CO 2 with steam supplied better performance till a
certain CO2/steam ratio was reached.
3.1.3. Group 3; horizontal injection-horizontal pro-
duction scheme
3.1.3.1. Group 3C. The horizontal production well
was placed along the shorter side of the model. The
location is 2.5 cm higher than the bottom of the
model. In this group, two steam-CO 2 tests and one
steam-alone test were carried out to investigate the
performance of CO2-steam mixture in horizontal
well combinations (Fig. 8). Oil recoveries were
30.3% of OOIP for the CO2/steam ratio of 49.7
din3/1, 25.3% of OOIP for 13.4 dm3/1, and 8.8% of
OO1P for the steam-alone test. If the recovery of the
test of 49.7 dm3/1 CO2/steam ratio is compared
with the 13.4-dm3/1 case, a 3.7 times higher amount
of CO 2 had to be injected to attain only 20% incre-
mental oil recovery over the 13.4-dm3/1 case. There-
fore, the cost effect of injecting larger amounts of
CO 2 gas should be considered. Both steam-CO,
tests recovered more oil than the steam-only case.
These results were also observed for the values of
steam/oil ratios. The steam/oil ratio was 16.6
cm3/ cm 3 for the steam-only case, 5.8 cm3/ cm 3 for
the CO2/steam ratio of 49.7 dm3/l, and 5.2
cm3/ cm 3 for the CO2/steam ratio of 13.4 din3/1.
3.1.3.2. Group 3D. The horizontal production well
was placed along the longer side of the model. The
location of horizontal injection well is the same as in
group 3C and its position is 2.5 cm higher than the
bottom of the model. Fig. 9 compares the recoveries
of two steam-CO 2 tests and one steam-alone test.
Recoveries were 51.1% of OOIP and 8.9% of OOIP
for the CO2/steam ratios of 11.7 and 38.1 din3/1,
respectively. For the steam-only case, it was 18.4%
1 0 0
O
>-
t"r
IaJ
>
o
o
w
rc
0
90-
80-
70-
60-
50-
40-
30-
20"
10-
0-.4-
0
GROUP3D]
[ ' : ' ~ - ' ~ - ~ ' ~ = I ~ [ ]
[ ]
[ ]
0. 5 1 1.5 2 2. 5
STEAM I NJECTED (PV of cwe)
Fig. 9. Oil recoveries for horizontal i nj ect i on-hori zont al produc-
tion scheme (group 3D).
122 F. Giimrah, S. Ba~ct / Journal of Petroleum Science and Engineering 18 (1997) 113-129
STEAM ONLY GROUP 3D
A INJECTED STEAM=O. 045 PV A'
INJECTION
A INJECTED STEAM=l . 230 PV A"
I NJECTI ON
A INJECTED STEAM=1. 680 PV A'
INJECTION T e mp e r a t u r e s i n "C
PRODUCTION INJECTION
Fig. 10. Temperature distribution of steam-alone test for horizontal injection-horizontal production (group 3D).
STEAM-CO 2 GROUP3D
C0 2 / STEAM RATI O=3 8 . 1 d m3 1 L
A INJECTED STEAM =0 . 2 9 0 PV A'
. .
/ / / /f "-d
INJECTION
A INJECTED STEAM = 1, 290 PV A"
INJECTION
A INJECTED STEAM = 2 . 0 5 0 PV A'
INJECTION
T e mp e r a t u r e s in "C
PRODUCTION I NJECTI ON
Fi g. | 1. Temperature di st ri but i on of steam-CO~ test f or hori zontal i nj ect i on- hor i zont al producti on (group 3D),
F. Gi~mrah, S. Ba~ct / Journal of Petroleum Science and Engineering 18 (1997) 113-129 123
of OOIP. The s t eam/ oi l ratios were 8.5 c m3 / c m 3
for the st eam-onl y case, 19.8 c m3 / c m 3 for the 38.1-
din3/1 case, and 3.8 c m3 / c m 3 for the l l . 7- dm3/ 1
case. In this scheme, the distance between the hori-
zontal producer and the horizontal injector was the
other factor influencing the recovery of oil from the
model. Two representative experiments, steam only
and s t e a m- CO z (38.1 dm3/ 1) are selected to show
the typical behavi our of the temperature distribution
in the model tests (Figs. 10 and 11). Temperat ure
data were taken from three levels: top, centre, and
bot t om of the model. St eam tends to reach the
producing end by flowing through the upper parts of
the model. This was demonstrated by the higher
temperatures being recorded at the top rather than at
the centre and bot t om of the model. Effective heating
occurred after steam breakthrough in the vertically
downward direction along the model, indicating
steam-zone enlargement. A larger portion of the
model was then heated by steam. The temperatures
of the bot t om of the model were lower in the steam-
alone test as compared to the s t e a m- CO 2 test.
3.1.4. Comparison between the runs
3.1.4.1. Steam-alone tests. A compari son of the re-
covery efficiencies of different well configurations is
shown in Fig. 12. From the comparison of oil recov-
eries for steam-alone tests of five well configura-
tions, a vertical injector and a horizontal producer
well scheme (group 2B) gave a better performance
than the other tests. The recoveries were the same for
the vertical i nj ect i on-product i on wells scheme (group
1) and the horizontal i nj ect i on-product i on wells
scheme (group 3C) till 1.1 PV of st eam injection is
reached. Then more oil was produced for group 1.
Although the ultimate recoveries of the vertical in-
j ect i on-hori zont al production wells scheme (group
2A) and group 1 were the same up to 2 PV of st eam
injection, the recovery was much higher up to 1.2
PV of steam injection in group 2A. This was at-
tributed to the beneficial effect of the horizontal
producer which accelerated the oil production. The
lowest ultimate recovery was obtained for group 3C,
the distance between the horizontal injector and hori-
100-
90-
8 0
0_
O 7 0
O
6 0
>-
n"
LU 50-
>
O
( j 4 0 -
LLI
r r
3 0 -
O
20-
10-
0:
I STEAM ONLY
i v e r t J c a l l r t j - h o r i z o r r t a l p r o d ( g r o u p ? A ) I
i i . _ _ . . .
I
I I I l I [
0 0. 5 1 1.5 2 2. 5
STEAM I NJECTED (PV of cwe)
Fig. 12. Comparison of oil recoveries for steam-alone tests.
3 3. 5
124 F. Giimrah, S. Ba~ct / Journal of Petroleum Science and Engineering 18 (1997) 113-129
1 0 0
13..
0
0
>-
n,-
IJJ
>
0
0
HJ
r
0
9 0 - [ S T E A M - C O 2
I(C02/steam ratio) I
8 0 - [ v o ~ . , o , ~ o . = ~ r o d . o r o o ~ . 4 ~ , ]
~ 0 - # I ~ . . . . . ~,,o,-~ . . . . . O, or*.o~ooo~D,,,~
6 0 , / ~ . ~ - E ~ ....... ~ ~ " . x
, " .> ,~ ,..X '" [ v e ~ c a l m I -horizontal prod ,group 2B 14.1
4 0 - ..~ >~-c " - - ' j m
~ u ..':~ r - - - -
. ~ ~ , l l I l - - . ~ ~ prod.,group I (22..3)]
2 0 4 j ~ ~ ~ ; -
, , ' ,
0 . 5 1 1. 5 2
S T E A M I N J E C T E D ( P V o f c w e )
Fig. 13. Comparison of oil recoveries for steam-CO~ tests.
2. 5
zontal producer was another important factor for the
displacement of oil. At 1 PV of steam injection, the
recovery of group 2B was 33.6% of OOIP compared
to 7.8% of OOIP for group 3C.
3.1.5. St eam- CO 2 tests
3.1.5.1. Oil recoveries. Fig. 13 shows the compari-
son of the results of tests which were conducted at
13_
O
O
>-
r r
uJ
>
O
O
UJ
r r
O
1 0 0
9 0 -
8 0 -
7 0 -
S T E A M - C O 2
~,ertical inj.-horizontal p r o d ,group 2A]
60- ~ . ~ "
5 0 " ~ , " ........... ~ertical inj-horizontal prod., g r o u p 2B ]
~ : ~ '-. "~ -.. ~ p , . ~ven cal inj.-vertical p r o d , group I [
2 0 ' - " . ~ - - -
~ ' ~ Z ' " ' . ~ [ horizontal ,nj -horizontal prod, group 3D
l o ~ ( ~ , " ~ - ~ - - " ~ . . . .
I I horizontal ,nj.-h . . . . . ~ prod, group 30 I
0 l " , T , ; " , , ,
0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0 1 6 0
C O 2 / S T E A M R A T I O ( d m 3 / L )
Ftg. 14. Oil recoveries as a fund]on of CO2/steam ratio.
F. Giimrah. S Ba~,ct / Journal of Petroleum Sctence and Engmeering 18 (1997) 113-129 1 2 5
about t he same CO2 / s t e a m ratios and di fferent well
confi gurat i ons. The results of tests havi ng the hi ghest
recoveri es were pl ot t ed in this figure. The r ecover y
effi ci ency of gr oup 3C was al so the l owest one
among the s t e a m- CO 2 tests, but gr oup 2A gave the
best per f or mance when compar ed t o ot her tests. Oil
recoveri es were 58. 3% of OOI P and 25. 3% of OOI P
f or t he CO2 / s t e a m ratio of 14.2 dm3/ 1 (group 2A)
and 13.4 dm3/ 1 ( gr oup 3C), respect i vel y. Fig. 14
compar es the experi ment al recoveri es as a funct i on
of CO2 / s t e a m ratio at 1 PV of st eam injected. The
CO2 / s t e a m ratio was r anged f r om zero t o 140 dm3/ 1.
The r ecover y was i ncreased up t o a cert ai n poi nt
wi t h i ncreasi ng CO2 / s t e a m ratio, after whi ch a di-
mi ni shi ng effect was observed. The use of t oo much
CO 2 can have undesi rabl e effects. Firstly, the gas
occupi es vol ume and reduces the amount of steam
injected. Secondl y, i f t oo much gas is injected, it
i ncreases the gas saturation f l owi ng bet ween the
i nj ect i on and pr oduct i on wells and causes chan-
nelling of the st eam to the pr oduct i on point. Thi s
effect is not ed in the experi ment al results present ed
here by the r educed r ecover y and l ower st eam injec-
tivity at hi gher amount s of CO 2. Hence by this
analysis, it became apparent that an opt i mum
CO2 / s t e a m ratio existed, whi ch is ~ 14 dm3/ 1 f or
Bau Kozl uca crude oil (12. 4API) whi ch caused the
best per f or mance in t erms of hi ghest oil recovery.
The CO2 / s t e a m ratio that resulted in the maxi mum
oil r ecover y was the same value f or all well conf i gu-
rations except in gr oup 3C. The i ncrement al oil
r ecover y over the st eam- onl y case decr eased wi t h
i ncreasi ng CO~/ s t e a m ratio. Ther ef or e the value of
CO2 / s t e a m ratio was one of the i mport ant fact ors
whi ch affect ed the per f or mance of the process. The
ot her fact or was the t ype of i nj ect or a n d / o r pro-
ducer, whet her the well was pl aced in hori zont al or
vertical position. The di st ance bet ween the wells also
affect ed the ef f i ci ency of process. I f t hey were cl ose
to each other, because of hi gher fluid mobi l i t y, earl y
br eakt hr ough occurred. As a result, the maj ori t y of
subsequent l y i nj ect ed fluids f ol l owed this est abl i shed
pat h of least resi st ance and process ef f i ci ency was
impaired.
2 0 0
1 8 0 -
2 " 1 6 0 -
E
o 1 4 0 -
co"
(3_
O 1 2 0 -
u J"
r r l O 0 -
z
O
r-- 8 0 -
C.)
- 3
( 3
O 6 0 -
CE
0.
. J
5
steam-CO2, (OPR)
4 0 - " -i ~/ ,- "
20- ...i -~
O'
0
v e r t i c a l i n j e c t i o n - h o r i z o n t a l p r o d u c t i o n , g r o u p 2 B
CO2/ste~'n ratio = 141 dm3/L ]
~..tlW_:~,.-stea m . , I , - - - only, (SOR)
,,~ A ' k " ~ . , ~ - - - - - - - I I D ~ steam-CO2, (SOR)
L
I steam only, (OPR)
~] i - "
... ~ ..... F i ........
'5 0 . 5 1 . 2
S T E A M I N J E C T E D ( P V o f c w e )
Fig. 15. Production data for vertical rejection-horizontal production (group 2B).
1 0
- 9
- 8
- 7
- 6
- 5
- 4
- 3
- 2
-1
0
2 . 5
%-
o
o,9_,
O
O
V--
<
r
._1
<
W
I - -
1.1.1
>
- 3
- 3
O
126 F. Giimrah, S. Ba~cz / Journal of Petroleum Science and Engineering 18 (1997) 113-129
250
vertical i n j e c t i o n - h o r i z o n t a l p r o d u c t i o n , g r o u p 2 A ]
0 150
1 o o ..:* T i
i , ! i
~ 1 i [ I r ~ . . ~'" steam -CO2,(SOR)i'
~ 5
"" "[
[ C 0 2 / s t e a m r a * J O = 1 4 2 d m 3 / L ] ......... -,,,,,,,
steam-C02, (OPR) ,="
= ~ .~ x." ",~"' "=~" T steam only, (SOR) . / "
steam only, (OPR)
3
0 . 5 1 1.5 2 2~5 3
S T E A M I N J E C T E D (PV o f c w e )
Fi g. 16. Pr oduc t i on dat a f or v e mc a l i n j e c t i o n - h o r i z o n t a l pr oduc t mn ( gr oup 2A).
- 1 0
- 9
"G
0
- 8
o
c~
- 7 O
o_
. - I
- 5 5
<
W
- 4 I---
O3
W
>
"-3 ~
5
o
-1
- 0
3 . 5
8 0 - - 2 5
7 0 -
C
E
6 0 -
o
c6
a_ 5 0 -
0
w"
I - -
4 0 -
z
0
I--
o 3 0 -
Q
0
a . 2 0 -
. . . I
5
1 0 -
l a m - C 0 2 , (01~i)
" i /
, , ' "
~ / ' ! . * - . i
t ' ~,', Z
h o r i z o n t a l i n j e c t i o n - h o r i z o n t a l p r o d u c t i o n , g r o u p 3 C
C O 2 / s t e a m r a t i o = 1 3 4 d m 3 / L ]
~K" ~ " " ~ "~,,
#" ~ k .......... k - . ' J.
' ' ) = ' " ~ l ~ e a m only, (SOR)
' L _
steam-C02, (SOR)
_ _ J
. . ~ l . . . ~ , - l r 1 " ~ l
. . . . . . i . . . . . . . .
s t e a m only, (OPR)
. . . . . . . . . . . . . . . .
' ~ ' 2 ' . 4 0 1 2 0 . 4 0 . 6 018 1 1 . 1 116 118
S T E A M I N J E C T E D (PV o f c w e )
Fi g. 17. Pr oduc t i on da t a f or hor i z ont a l r e j e c t i o n - h o r i z o n t a l p r o d u c n o n ( gr oup 3C).
%-
0
- 2 0 "~
o
-
0
0
- 1 5 ~
. . J
<
U . I
- 1 0 f -
w
>__
i---
5
- 5 ~
0
2
F. Giimruh. S. Ba~ct / Journal of Petroleum Science and Engmeermg 18 (1997) 113-129 127
3.1.6. Oil production rate and st eam~oi l ratio
The oil production rate and steam/oil ratio as a
function of steam injected are given for steam-CO 2
and steam-only experiments of groups 2B, 2A and
3C in Figs. 15-17, respectively. An increased rate of
oil production was observed and the oil production
rate reached a peak value. Then a gradual decline in
oil production rate was observed during the early
production phase of all steam-CO 2 tests. The rate of
oil production in the steam-CO~ tests was relatively
constant after the production of oil bank. By compar-
ison, the rate of oil production was fluctuated in
steam-alone tests. As can be seen in production data
more oil was mobilised during the steam-CO 2 dis-
placement. So the size of the oil bank was larger.
This may have caused the observance of higher
pressures at the inlet. After the oil bank reached the
producing end, pressures declined and stayed almost
constant as expected. The cumulative steam/oil ratio
as a function of steam injected is also shown in Figs.
15-17. After the production of oil bank, the
steam/oil ratios slowly increased. Since steam/oil
ratio is an indication of the economics of the process,
the effect of CO 2 should also be considered in
addition to steam for st eam-CO, tests. The
steam/oil ratio was also lower for the case of the
steam-CO 2 process when compared to the steam-
only case. Because of the high viscosity of unheated
oil, the response to steam injection in this type of
reservoir usually was delayed and did not change
significantly until heat arrived at the producing well.
In steam-alone tests, the recovery of heavy oil after
steam breakthrough was more significant, indicating
fluctuated oil production rates after breakthrough,
and the mechanism of mobilisation of in-situ oil for
this case was explained by Clossmann and Seba
(1982) in the following paragraph as:
. . . while tendency for steam fingering to be sup-
pressed by condensation, steam penetration neverthe-
less proceeded at a rate greater than the rate of
movement of viscous oil left behind. In this connec-
tion, it is worth noting that the effect of scaling and
heat losses would probably be increased somewhat
the total heat volume of steam condensate that passed
&
0
0
>-
r
uJ
0
0
w
r
w
o9
z
u J
a
z
0
0
12
vert~,aJ m I -horizontal prod.group 2A (14,2) I
I
4 -
l STEAM- C02
(CO2/steam ratio)
~ 1 v e ~ c a l inj - h . . . . . t a l p r o d ,group 2 B (141) ]
horizontal m I -horizontal p r o d ,group 3D (11 7) ]
J ~ ~ I honzorrtal Inl -honzontal p r o d ,group 3C (13 4)
0 ' . s 1 . ' 5
STEAM I NJECTED (PV of cwe)
F~g. 18. Condensate recovery of steam-CO~ tests.
'1 vertical Inl -vertical p r o d , g r o u p 1 (22 3) I
2. 5
128 F. Gi~mrah, S. Ba~, ct / Journal of Petroleum Science and Engineering 18 (1997) 113-129
through the system, for the same rate of advance of
the steam front and possibly some before its passage.
3.1.7. Condensate recoveries
Fig. 18 shows the condensate recoveries as a
function of steam injected for the tests at about the
same CO2/steam ratios and different well configura-
tions. For steam-alone tests, condensate production
was not observed. CO~ might be primarily responsi-
ble for stripping the lighter ends. The vertical injec-
tion-horizontal production well configuration of both
groups 2A and 2B recovered condensates at ~ 11%
of OOIP. The condensate recovery was ~ 3% of
OOIP for the other well configurations. The densities
of condensate fractions were measured at 20C. The
condensate fractions of the produced fluids had colour
variations from pale straw to dark brown. The con-
densate densities were ranged from 0.77 to 0.88. The
densities were found to be higher at the initial period
of condensate production. Then lighter ends were
produced, which indicates the effect of CO 2 strip-
ping and transport of the lighter compounds of the
in-situ oil. The amount of condensate production
increased with the addition of CO 2 in the flowing
stream. The presence of CO 2 assisted the transport
of large volumes of lighter ends for a long time
when compared to steam-alone tests.
It is evident from these results that non-con-
densable gas was of benefit in accelerating the rate
of oil production and enhancing the recovery of
more viscous oil. The flooding experiments showed
that the simultaneous injection of CO 2 with steam
will definitely increase the ultimate recovery over
steam alone and that there is an optimum CO2/steam
ratio that results in the highest recovery. The results
indicated that the optimum COJ s t eam ratio for
maximising the recovery was ~ 14 dm3/1. Steam-
CO, mixtures recovered 73.5% more crude for
steam-CO 2 test (group 2A) than the steam-only
case. The other experiments at various CO2/steam
ratios yielded recoveries between the results of steam
alone and steam-CO 2 at optimum conditions. The
gas injected with steam was non condensable, hence,
it created a permanent gas phase across the top of the
model. This assisted steam propagation along the top
of the model. Consequently, heat arrived at the pro-
ducing well much sooner than when steam alone was
injected, and oil production was accelerated.
4. Conclusions
The following conclusions can be drawn from:
(1) Steam-only drive experiments. Oil recovery
for the vertical injector-horizontal producer (group
2B) well configuration was higher than that of the
other well configurations. The lowest ultimate recov-
ery was obtained from the horizontal injector-hori-
zontal producer well configuration (group 3C).
(2) St eam- CO, drive experiments. The co-injec-
tion of CO 2 with steam increased the oil production
rate and ultimate oil recovery over the steam-only
case. The recovery efficiency of horizontal
injector-horizontal producer (group 3C) was also the
lowest one, but vertical injector-horizontal producer
(group 2A) gave the best performance when com-
pared to other tests. The optimum COz/ st eam ratio
for maximising oil recovery was ~ 14 dm3/ l for all
well configurations.
(3) All experiments. The other factor which influ-
enced the oil recovery was the type of the injector
and/ or producer whether it is placed in horizontal or
vertical position. The distance between the wells also
affected the efficiency of the process. The effective
mechanisms for the mobilisation of oil were viscos-
ity reduction, steam distillation, stripping effect, and
additional gas drive effect of CO 2.
Acknowledgements
The authors would like to acknowledge Turkish
Petroleum Corporation for providing the crude oil
and reservoir data from the Batt Kozluca oil field.
The authors wish to thank Tu~rul Aybak for con-
tributing to the experimental program. We also grate-
fully acknowledge the support from METU Research
Fund No. AFP 92-03-06-01.
References
Balog, S.E., Kerr, R.K. and Pradt, L.A.. 1982. The wet-air
oxidation boiler for enhanced oil recovery J. Can. Pet Tech-
nol. (Sept.-Oct.), pp. 73-79.
Briggs, P., Redford, D.A. and Harris, P., 1982. Core-sized physi-
cal simulator experiments to study the effect of naphtha and
CO~ addition to steam m the recovery of bitumen from od
sands. Presented at 1982 Annual Tech. Meet., Pet. Soc., Can.
F. Giimrah, S. Ba~ct / Journal of Petroleum Sctence and Engineering 18 (1997) 113-129 129
Inst. Min. Metall. 6t h Symp. on Engi neeri ng Appl i cat i ons of
Mechani cs, Calgary, Alta., June 6- 9, 1982, Pap. CIM 82-33-
84.
Chu. C., 1985. State of t he art review of st eam flood field
projects. J. Pet. Technol . (Oct.), pp. 1187- 1902.
Claridge, E.L. and Dietrich, LK. , 1983. Vi scous fi ngers and
gravi t y t ongues in CO 2 - s t e a m stimulation. Present ed at Soc.
Pet. Eng. , Calif. Reg. Meet. , Ventura, CA, Mar. 23- 25, 1983.
Pap. SPE 11707.
Cl ossmann, P.J. and Seba, R.D., 1982. Laborat ory tests on heavy
oil recovery by st eam injection. Present ed at Soc. Pet. Eng. ,
Calif. Reg. Meet. , San Francisco, CA, Mar. 24- 26, 1982, Pap.
SPE 10778.
Doscher, T.M., Ying, A. and Kost ura. J.A., 1988. Improvi ng t he
st eam drive: Physi cal model st u&es of inert gas injection and
horizontal wells. Present ed at Soc. Pet. En g , Calif. Reg.
Meet. , Mar. 23- 25, 1988, Pap. SPE 17448, pp. 443- 448.
Fox, R.L., Donal dson, A. B. and Mul ac, A.J., 1981. Devel opment
of t echnol ogy for downhol e st eam production. Present ed at t he
Soc. Pet. E n g / U. S . Dep. Energy 2nd Joint Syrup. on En-
hanced Oil Recovery, Tul sa, OK, Apr. 5 - 8 , 1981, Pap.
S P E/ DOE 9776.
Frauenfeld, T. W. J. , Ridley, R. K. and Nguyen, D.M., 1988. Effect
of an initaal gas cont ent on t hermal EOR as applied to oil
sands. J. Pet. Technol . (Mar.). pp. 333- 338.
Giirrlrah, F. and Okandan, E., 1987. The use of CO 2 wi t h st eam
for heavy oil recovery Proc. 4t h Eur. Symp. on Enhanced Oil
Recovery, Hambur g, 2 7 - 2 9 Oct., 1987, pp. 535- 546.
Giimrah, F. and Okandan, E., 1992. St e a m- CO 2 flooding: An
experi ment al study. In Situ. 16(2): 89- 108.
Hardi ng, T.G., Farouq All, S.M. and Flock, D.L., 1983. St eam
flood performance m the presence of carbon di oxi de and
nitrogen. J. Can. Pet. Technol. ( Se pt . - Oc t ) , pp. 3 0 - 3 7
Hombr ook, M. W. , Denghani , K., Qadeer, S., Ost ermann, R.D.
and Ogbe, D.O., 1991. Effect s of CO 2 addition to st eam on
recovery of West Sak crude oil. Soc. Pet. Eng. Reservoi r Eng.
(Aug. ), pp. 278- 285.
Kantar, K. and Topkaya, I., 1983. Devel opment of oil and natural
gas resources and prospect s of enhanced oil recovery in
Turkey. Turki sh Pet rol eum Corp., Ankara, Rep. 21.
Leung, L.C., 1982. Numeri cal eval uat i on of t he effect of si mul t a-
neous st eam and CO 2 injection on t he recovery of heavy oil.
Present ed at Soc. Pet. Eng. , Calif. Reg. Meet. , San Franclsco,
CA, Mar. 24- 26. 1982, Pap. SPE 10776.
Mel dau, R.F., Shlpley, R.G. and Coats, K.H., 1981. Cycl i c
g a s / s t e a m snmul at i on of heavy oil wells. J. Pet. TechnoL
(Oct.), pp. 1990- 1998.
Met wal l y, M. 1990. Effect of gaseous additives on st eam pro-
cesses for Li ndbergh Field, Alberta. J. Can Pet. Technol
( Nov. - Dec. ) , pp. 26- 30.
Nasr, T.N. and Pierce, G.E., 1995. St e a m- CO 2 recovery pro-
cesses for bot t om wat er oil reservoirs. J. Can. Pet. Technol .
(Sept.), 34(7): 42- 49.
Nasr, T. N. , Prowse, D.R. and Frauenfeld, T. W. J. , 1987. The us e
of fl ue gas wi t h st eam in bi t umen recovery f r om oil sands. J
Can. Pet. Technol . ( May- J un. ) , pp. 62- 69.
0z e n, A. B. , 1967. An i nvest i gat i on of the recovery of oil by
st eam flooding. M, Sc. Thesi s. Pet rol eum and Natural Gas
Engi neeri ng Depart ment , Pennsyl vani a State Uni versi t y, Uni -
versi t y Park, PA (Mar.).
Paracha, O. U. R. , 1985. An experi ment al i nvest i gat i on of t he effect
of carbon dioxide on st eam recovery. Ph.D. Dissertation,
Uni versi t y of Okl ahoma, Norman, OK.
Pujol, L. and Boberg, T. C. , 1972. Scal i ng accuracy of laboratory
st eam fl oodi ng model s. Present ed at Soc. Pet. Eng. 1972 43rd
Annu. Calif. Reg. Meet., Bakersfield, CA, Pap, SPE 4191.
Pursl ey, S.A., 1975. Experi ment al st udi es of t hermal recovery
processes. Present ed at Maracai bo Heavy Oil Symp. , Mara-
caibo, Jun. 4, 1975
Redford, D. A. , 1982. The use of sol vent s and gases wi t h st eam in
the recovery of bi t umen f r om oil sands. J. Can Pet. Technol .
( Jan. - Feb. ) , pp. 45- 53.
Rintoul, B., 1979. GATEOR project at Edison. Pac. Oil Worl d
(Sept.), pp. 40- 41.
Roadifer, R.E., 1986. How heavy oll occurs world wide. Oil Gas
J. (Mar. 3), pp. 111- 115.
Schirmer, R. M. and Eson, R.L., 1985. A dtrect fired downhol e
st eam generat or - - f r om desi gn to field test. J. Pet. Technol .
(Oct.), pp. 1903- 1908.
Slobod, R.L. and Merri am, L.G., 1969. Us e of a per manent gas
phase to augment t he benefits of st eam injection. Prod. Mon.
(Jan.), pp. 6- 15.
Sperry, J.S., 1981. Heavy oil recovery syst em compl et es three
field t est s in Mi d-cont i nent region. Oil Gas J. (Jul. 27), pp.
225- 237.
Stone, T. and Ivory, J., 1987. An exammat i on of s t e a m- CO 2
processes. J. Can. Pet. Technol . ( May- J un. ) , pp. 54- 61.
Stone, T. and Mal col m, J,D., 1985a. Si mul at i on of a large s t e a m-
CO 2 comj ect i on experi ment . J. Can. Pet Technol . ( Nov. -
Dec.), pp. 51- 59.
Stone, T. and Mal col m, J.D., 1985b. Si mul at i on of a large
s t eam- car bon &oxi de co-mj ect i on experi ment . J. Can, Pet.
Technol . ( Nov. - Dec. ) , pp. 51- 59.
Stone, T. and Nasr, T.N., 1985. Si mul at i on of a series of st eam
additive expenment s . Present ed at 1st Annu. Tech. Meet. ,
Sout h Sask, Sect., Pet. Soc., Can. Inst. Min. Metall., Regi na,
Sa s k, Sept. 15- 17, 1985.
Wei nst ei n, H.G., 1974. Mat hemat i cal model s for t hermal recovery
processes. Present ed at Symp. on Heavy Crude Oil, Maracat bo
(Jul.).
Wilhelm1, A. R. and Knopp, P.V., 1979. Wet air oxidation an
alternative to incineration. Chem. Eng. Progr. (Aug. ), pp
46- 52.

You might also like