. .. - .

SPE 9928
SPE
CO INJECTION
AS AN IMMISCIBLE
APPLICATION
FOR
ii
E~./U!CED
RECOVERYINHEAVYOILRESERVOIRS

by Ala K. Khatib and R.C. Earlougher, Godsey-Earlougher Div.

of Williams Brothers Engine’wing Company;
Kutluyil Kantar, Turkish Petroleum Corporation

Copyright 1981, Society of Petroleum Engineers

This paper was presented at the 1981 California Regional Meeting in Bakersfield, California, March 25.26, 1961. The materisl ia subject to
correction by the author. Permission to COPYia restricted to an abstract of not more than 300 words. Write: 6200 North Central Expressway,
Dallas, Texas 75206.

ABSTRACT

Most field applications and pilot tests of C02 several ongoing or planned field or pilot projects
injectic% are for miscible drive, and are there- in which the reservoir oil is too heavy for misci-
fore confined to light oil reservoirs. However, it bility with CG2 to take place. !!owever,these
has been known for decades that as C02 dissolves in still appear under the heading of “miscible C02
heavy oil, it swells the oil and reduces its injection.” The objective of this paper is to
viscosity. Core displacement tests in the fifties review results of early, as well as recent, labor-
indicated that substantial additional recovery of atory work and field tests of C02 injection as an
heavy oil, and in some cases light oil, was possible EOR technique in heavy and light ofl reservoirs,
with carbonated water. Until recently, unfavorable where the process is immiscible. The term “heavy
economics discouraged field application, and yet in oil” herein encompasses crudes in the approximate
1976 a field project was commenced in southern range of 10” to 25° API.
Arkansas.

Two techniques have been field-tested. These

are “huff-and-puff” C02 injection for stimulation It has been known for many years that as C02
of the producing wells and either alternate or dissolves in oil, it swells the oil and reduces its
simultaneous injection of c02 and water. viscosity. In 1945 Poettmann and Katz discussed the
phase behavior c C02 and paraffin systems2. Typical
Both process mechanisms are discussed and are figures for a heavy crude are 10 to 22 percent in-
illustrated, where possible, by the results of crease in oil volume and viscosity reduction to
field tests in the United States. These have been less than 0.1 of original value at 120”F (49”C)
in sandstone reservoirs containing light oils and and 800 to 1200 psi (55 to 83 bars)3. At lower
heavy oils. Several other pilot or field projects temperature, increase in volume is greater. In
are in the planning stage. A planned field demon- 1951 and 1952 there were at least eight articles
stration project of both processes, in different or papers published relating to the use of C02 and
test areas of a fractured limestone reservoir carbonated water for increased recovery of lighter
(oil 9° - 15°) in Turkey, is summarized. oils4-11 of 30° to 50° API gravity. In March 1959,
four U. S. Patents were Issued to J. W. Martin,
The possible range of candidate reservoirs ia et a112, relating to the immiscible use of C02
discussed, and general economic comparison with (carbonated water) for oil recovery.
steam injection is given. The obvious candidates
are heavy oil reservoirs that are too deep for cur- Other immiscible mechanisms that enhance oil
rent steam injection technology and economics. recovery by C02 injection have been reported s:.
decompression (dissolved gas drive) and reduc:ion
—INTRODUCTION of interracial tension13. Because of these prop-
erties, immiscible displacement with c02 should be
Reviews of enhanced oil recovery methods and more efficient than that with natural gas. Two
their applications generally refer to miscible dis- papers14*15 that have appeared in the literature
placement with carbon dioxide. In the survey of give good reviews of the properties of C02, C02/oil
EOR projects published in the Oil & Gas Journal systems, and the possible oil recovery mechanisms.
every two years (last waa in March 1980)1 there are The recent report by Gruy Federal, Inc. for the
U. S. Department of Energy16 providea a compre-
hensive compilation of published data. The field
References and illustrations at end of paper. tests studied in the reprzt, however, relate to
I
.

C02 INJECTION AS AN IMifISCIBLEAPPLICATION FOR ENHANCED RECOVERY IN HEAVY OIL RESERVOIRS SPE 9928

miscible dit;placementof oil by C02 except for two Pc.croleumCompany also :an many tests in its labor-
heavy oil projects which are immiscible appli- atory in Oklahoma City under an agreemrnt with Oil
cations. Recovery Corporation. Tests by EE were conducted
in high-pressure flood pots at 800 to 1000 psi (55
It is appropriate here to distinguish between to 69 bars) and those by Sohio in a specially
miscibility and volubility. Two miscible fluids designed pressure cell which permitted ‘“-ect
will combine in any proportion to for~ one phase, measurement of C02 volumes. Gravity of oil in the
e.g., light hydrocarbons or LPG with reservoir oil core samples ranged from 28” to 50° API. Displace-
of 35+0 API. Otherwise, one fluid will dissolve in ment was immiscible under the test conditions.
another only up to a certain proportion, termed
volubility, governed by the nature of the fluids, Fresh cores from waterflooded fields in the
temperature, and pressure, e.g., the volubility of Mid-Continent, Illinois, and Pennsylvania were used
C02 in 20° API oil at 2f,00psi (138 bars) and 100”F to compare recoveries by displacement with water,
(38”C) is 700 scflbbl (125 m3/m3). An intermediate and with carbonated water. In practically all
phenomenon is dynamic miscibility where misci- samples, cores flooded with carbonated water yielded
bility is gradual, multi-contact, and involves higher oil recoveries, compared to samples flooded
extraction or vaporization of light ends, e.g., C02 only with water. The residual oil saturation
and reservoir oil of 35° AP113. after carbonated water usually was 2 to 26% PV less
than for water alone. When the reduction in Sor
Core displacement tests, beginning in 1948, averaged 8% PV or more for a group of samples from
indicated that a substantial recovery of light oil a reservoir, response to C02 was considered signif-
in several reservoirs was possible with C02 or icant. In reviewing average values obtained by EE
carbonated water. Until recently, unfavorable for 39 wells (Fig. 1) L. F. Elkins, et a120 of
economics discouraged field application, although Sohio noted that for most OF the Mid-Continent cores
a few pilot tests and one commercial-size field the increased oil recovery could be accounted for
application of the Oreo Process were carried cut. by increased FVF of the oil, but in some the supple-
The first pilotll was started in 1949 in western mental recovery was more than could be accounted
New York State, and the full-scale project17 was for by the swelling of the oil. Cores from five
commenced early in 1958 in northeastern Oklahoma. wells in western New York State (north end of the
Bradford Field) yielded substantially more oil
A U. S. Patent was issued May 6, 1969, to P. C. recovery (+10% PV) than accounted for by swelling
Keith (an associate of U. S. Oil and Refining Co.);it for 50” API gravity oil, and 10% of the Mid-
contained ten claims including the obtainment of Continent well averages indicated that swelling
C02 from varioua sources for injection into heavy alone would not account for the increase. In
oil reservoirs for increased oil recovery, and addition, Elkins concluded that the EE laboratory
also in conjunction with huff-and-puff steam flood pot procedure resulted in the residual oil
injection. This patent describes an immiscible in the cores being fully saturated with c02 even
recovery process suggesting a net injection c: C02 though only carbonated water was used.. This iS
sufficient to fully saturate the oil and w;ter in probably true.
the reservoir prior to blowdown.
A supplement reported on laboratory tests run
In early 1969, U. S. Oil and Refining Company in an effort to determine whether it was the nature
started a C02 project in the Ritchie Field18 in of a given rock or of a given oil that resulted in
southern Arkansas but lost its C02 supply after 264 the high incremental oil recovery. The results
MMcf (7.48 x 106 m3) had been injected into three indicated that it probably was a combination of the
wells in a period of 77 days. In February 1976, a character of the oil and the rock together, i.e.,
full-scale C02 project was started in the nearby the character of the system. Also it was deter-
Lick Creek Field19 by Phillips Petroleum Company and mined that reduction in interracial tension was too
U. S. Oil and Refining Company with Phillips as the little to provide a satisfactory explanation.
operator.
EARLY FIELD TESTS
A planned field demonstration project for a
large heavy oil carbonate reservoir in Turkey is For the application of the Oreo Process (car-
summarized. In addition, recent laboratory dis- bonated waterflooding), Oil Recovery Corp. developed
placement test data on core samples from the the theory that the C02 requirement in the field
WilmiligtonField are included. was only that volume of the gas required to saturate
approximately 25% of the oil and water in the
FAI.Y LABORATORY DISPLACEMENT TESTS reservoir.

The initial work in the late fortiee was done First Pilot
by Montclair Research, New Jersey on sa13 packs,
for Oil Recovery Corporation. Results indicated The Bradley Producing Corp. of Wellsville, New
that residual oil saturation (Sor) could be reduced York carried out the initial field testll starting
by as much as 15% PV by displacement with carbon- in September 1949. A total of 370 tons of dry ice
ated water after waterflooding. was the source of c02 for four injection wells,
forming an enclosed 0.47-acre (1900 - m2) five-spot.
In 1948-1952 several hundred laboratory flood Results of the testwere disappointing and indicated
tests of fresh core samplee using carbonated water an oil recovery of only about 25% of that suggested
(Oreo Process) were run by Earlougher Engineering by laboratory flood pot tests. The amount of C02
(EE), Tulsa, for Oil Recovery Corporation. Sohio used in the field was equivalent to about 25% of

462
“.W ..-” n. .. ● *W-A x.” . A.* “ . -U.l”””,,u.. , -.” L.* .-. 4.-.

that used in the laboratory. However, the following The field application of cyclic well stimulation
were learned during the first 14 months of oper- with C02 was first conducted in the Ritchie Field in
ation21: southern Arkansas in 1969. C02 was injected in
three wells. After the C02 source was cut off the
1.. Additional oil had been recovered. wells were put on production. One of them flowed
125 BOPD22.
2. Application of Oreo Process had produced no
serious corrosion problems to date. In the injection cycle of this process, as the
well is pressured up, C02 dissolves in the oil and
3* Injection capacity of the four five-spot inputs conditions it, through swelling, viscosity reduction,
had been increased approximately 50%, with such etc. for production by gas drive. Following
increase being maintained for at least 6 months. injection, the well is shut in to allow soaking to
take place. In the production cycle, as the well
4. A deposit of barium sulfate had occurred in is allowed to flow, gradual pressure depletion causes
several producers, including the test five-spot. C02 to come out of solution al~dform a gas phase
Such scale deposit had not occurred during which expels the oil from the matrix into more open
normal waterflooding operations. channels, and thence to the well bor:. When oil
production declines and the C02-oil xatio becomes
5. Production of natural gas, which was practically excessive, the well is again put on injection. The
nil before the test, increased to 160 scflbbl cycles can be repeated until incremental oil recovery
of oil in 2 months time and then averaged is no longer economic. In successive cycles the
approximately 300 scffbbl the last 5 months of injected C02 penetrates further and further into
1950. Gas analyses indicated as much as 20% the reservoir until a large volume of the oil be-
unsaturates, such as ethylene. comes saturated. In field tests carried out so far,
two cycles appear to be optimum.
6. The presence of the barium sulfate scale and
the unsaturated hydrocarbons in the natural gas In an operational mode field-wide, produced
suggested some type of chemical reaction within C02 is collected for compression and reinfection.
the reservoir system triggered by the C02. The cyclic operations may also be followed by
waterflooding or alternating water/C02 flooding,
Second Field Test for additional oil recovery;

In 1951 a single enclosed five-spot test of the At least two field pilot projects are planned
Oreo Process was made on a flooded-out lease on the in California -- one in the Tar zone of Fault Block
Reno property south and west of Nowata, Oklahoma. 11 (14°, 300 cp) of the Wilmington Fieldl, and the
Results of the test were negative but congruous with other on the North Bolsa Strip Lease of the
the laboratory carbonated flooding tests which were Huntington Beach Field (16°, 175 cp)23. A large
much less favorable than those for the Knight Farm. demonstration project is planned in the Bati Raman
Field ~9-157 in Southeast Turkey.
Full-Scale Oreo Flood
Alternate Injection of COP and Water
Starting in early 1958$ Oil Recovery Corporation
instituted a commercial scale Oreo flood develop- The alternate injection of slugs of C02 and
ment on a 240-acre (96 hectare) tract it had water into a viscous oil reservoir, in a frontal driv<
acquired northwest of Bartlesville, 0klahoma17. mode, is an immiscible displacement process. The
Carbon dioxide was generated on the lease by burn- possible advantage over displacement with C02 alone
ing natural gas. The carbonated waterflood was is improved mobility control, resulting in more
successful and would have been economic if this efficient displacement and lower C02 injection (or
single project had not been burdened with prac- recycling) requirements. But there is a potential
tically the total cost of the C02 generation plant. disadvantage of oil trapping by water, which would
The C02 was injected down l-inch macaroni string prevent C02 from contacting the trapped oil. This
inside cement-lined 2-1/2 inch tubing, may be partially mitigated by the fact that C02 is
simultaneously with water so that carbonation com- several times more soluble in oil than in water.
menced in the well bore. The water injectivity
factor for most of the 35 injectors following C02 After waterflooding, reservoir pressure in many
injection increased two to fourfold. fields can be allowed to decline to recover addi-
tional oil by solution gas drive. However, as the
PROCESS DESCRIPTION blowdown proceeds the oil produced becomes heavier
and more viscous, as its CO~ content declines..
The following is a description of the mech-
anism and mode of application of both recovery The encouraging laboratory and field results of
techniques of C02 injection in heavy oil reservoirs. reducing C02 mobility with aqueous surfactant instead
of water, reported in a recent paper24 on miscible
Cyclic Well Stimulation drive, should be equally applicable to immiscible
displacement with C02. The intimate mixing of C02
This is a localized stimulation process, similar and the surfactant in the reservoir produces a foam
tc the “huff-and-puff” steam injection. Zt first or an emulsion, thereby reducing the mobility of C02.
.ippearedin a patent by P. C. Keith in 1969. This
actually described the cyclic application of a At present, there are two field projects employ-
mixture of C02 and steam. ing alternate C02jwater injection in Arkansas, and

463
 --
CO, TNJECTION
—..—— AS AN IMMISCIBLE APPLICATION FOR ENHANCED RECOVERY IN HEAVY OIL RESERVOIRS
--—-. .----— .—-
SPE 9928

at least two in the planning stage -- one in The impetus for undertaking C02 injection was
California, and another in Turkey. the favorable response obtained in the field test
in the neighboring Ritchie Field in 1969. A C02
FIELD APPLICATIONS compression and dehydration plant and a 6“ x 65-mile
pipeline were built to transport C02 at about 1350
Data on the Arkanaas fields, both now operated psi (93 bars) as a super-critical fluid.
by Phillips Petroleum, were obtained from Ref. 15
and a recent report prepared b Gruy Federal, Inc. T. B. Reid, J. E. Chrisman, and H. J. Robinson
for the U. S. Dept. of Energyl8 $19. of Phillips Petroleum Company are presenting a
paper on Lick Creek at the SPE/DOE Improved Oil
Ritchie Field Recovery Symposium in Tulsa in early April. The
project is reportedly a technical and an economic
The Ritchie Field in southern Arkansas produces success.
from the Baker sand at a depth of about 2600 feet
(790 m). Average reservtiirrock parameters are: BATI MMAN FIELD AND C02 INJECTION
net pay 9 feet, porosity 31%, permeability
2750 md, and initial water saturation 20%. The The Bati Raman Field, operated by the Turkish
stock-tank oil has 16° API gravity and a viscosity Petroleum Corporation (TPAO), was discovered in
of 195 cp at the reservoir temperature of 126°F 1961 in southeastern Turkey. It is the largest oil
(52°C). The reservoir covers an area of some 220 field in Turkey, having an estimated original oil
acres (87 hectares) containing 16 wells. Pro- in place (OOIP) of about 1.75 billion STB (278
duction started in 1964 and the field was unitized million m3). Water drive appears insignificant,
in 1968. Estimated primary recovery was 690,000 STB and all primary recovery, estimated at 1.5% of
(77,900m3) or 14% of the OOIP. Figure 2 shows OOIP, is due to fluid and rock expansion, there
the reservoir performance history. being practically no solution gas drive below the
bubble point26.
In early 1969, C02 was injected at 3 to 4 MMcfd
(85,000 to 113,000 m3/d) for 78 days into three The producing formation of the Bati Raman Field
wells. The C02 source was then lost and the pro- is the Garzan Limestone, an elongated east-west
duced C02 was recycled for stimulating producers. anticline, about 17 by 4 km.
Reservoir response was swift. Oil production
doubled by comparison with extrapolated primary The Garzan Limestone exhibits pronounced reser-
performance, from about 65 to 130 BOPD (10.3 to voir hetsrogeneitiea, both laterally and vertically.
20.7 m3/d) in the second half of 1969. This rate The reservoir rock is a fractured vuggy limestone
was maintained for about one year before water in the western and central parts of the field, but
injection started in May 1970. The increase in is chalky and tighter to the east. Porosity is
production was therefore solely due to C02 injection. generally 14 to 20%, and the matrix permeability
is typically 10 to 100 md, as given by core
Water was injected peripherally at rates to analysis. However, well tests show effective
balance withdrawals; the cumulative injection to permeabilities in the 200 to 2000 md range, con-
July 1978 was 3.8 million bbls (604,000 m3). firming the contribution of secondary porosity
There was immediate reaction to water injection, (fractures, vugs, and connecting cracks).
the oil rate increasing to about 400 BOPD (64 m3/d).
It was maintained at this rate for a year and then The oil ranges from 15° in the west to 9° API
started declining steadily to 50 BOPD (8 m3/d) in in the east and the viscosity from 300 to 1000 cp.
July 1978, when the reservoir was shut in. Incre- The solution gas-oil ratio is only 18 scf/bbl,
mental recovery over primary was 415,000 STB resulting in a low bubble-point pressure of 160 psi.
(66,000m3) or 12.5 percent of the OOIP. Anew The reservoir temperature is 150°F (65.5”c) and
program of alternate injection of C02 and water the original reservoir pressure was about 1800 psi.
commenced in July 1978 and is continuing. The average reservoir pressure has now declined to
about 400 psi, after a cumulative production of
The supplemental recovery that can be attri- about 30 million STB. Fig.3is used to illustrate
buted to C02 injection is all the additional reservoir performance history.
recover before water injection of 22,000 STB
(3500m ?3’
), plus some unknorn portion of increased The field was initially developed on 500 m spac-
production during water injection. For illustra- ting(25 hectares or 62 acres per well), and infill
tion purposes, if this is also about 22,000 STB, the drilling started in 1974 in the western region.
C02 utilization factor is 6.0 Mcf/bbl (1070 m3/m3). Active producers number about 120. All are on pump
This falls within the range of 3 to 7 Mcf/bbl for and produce a total of about 3300 BOPD, compared
optionally sized slugs in miscible displacement if to a peak rate of about 9000 BOPD in 1969. Initial
produced C02 is reinjected25. well production rates ranged up to 400 BOPD, but
are now averaging about 30 BOPD per well.
Lick Creek Field
Various enhanced recovery tests have been tried
This is the largest C02 injection operation in in the field, viz. water injection, steam cycling,
a heavy oil reservoir. The oil has 17” API gravity steam drive, and air injection (for in situ com-
and a viscosity of 160 cp at the original reservoir bustion). About 3.2 million barrels of water were
conditions of 118°F (48”C) and 1200 psi (83 bars) injected in the central area of the field between
C02/water injection started in February 1976 and is 1971 and 1978. This process resulted in a marked
continuing. increase in oil production, over the decline trend,

464
SPE 9928 A. K. KHATIB. R. C. EARLOUGHBR, AND K. KANTAR

from wells over the whole central area, as illus- Field Demonstration Proflect
trated in Fig. 3. Response to water injection in
this central area was good. The water oil ratio- Extensive laboratory investigationa, theoret-
curaulativeoil performance was analogous to Tar V ical and numerical simula~ion studies, and a com-
in the Wilmington Field. parative engineering and economic study of all EOR
techniques, led to a decision in 1980 to construct
Results of the steam injection were encour- facilities necessary to test the application of
aging, although the test was terminated prematurely. C02 injection in two demonstration areas of the Bati
Air injection did not ignite the Gil. Although a Raman Field. In each of the test areas the demon-
sizable C02 gas reservoir is available 85 km away stration project is expected to laat two to three
at Dodan Field, it has not been possible to cor.duct years.
a C02 injection test in the field, short of making
large capital investments in pipeline and compres- The C02 injection project will entail producing
sion. a C02-rich gas from the Dodan gaa f~.eld,processing
it through H2S-removal and dehydration plant, and
Laboratory Displacement Tests compressing it to 1730 psi (119 bars). After
cooling, the gas will be delivered at about 1700 psi
The Institut Fran~ais du P6trole laboratories, (118 bars) and 150”F (66”c), as a aupercritical
outaide Paris, conducted various laboratory tests fluid, to a 10-inch 85-km pipeline to the western
on Bati Rsman cores using Bati Raman oil and Dodan area of the oil field.
gas (88% C02). In one series of tests the incre-
mental recovery by C02 flooding, over that by Construction of the Dodan/Bati Raman project
waterflooding, was investigated. The experimental is scheduled to start in the spring of 1982, and
data are summarized in Table 1 and shown in Figs. operations are scheduled to commence at the end of
4 and 5. In the experiment, virgin oil with the 1982. The gas conditioning plant and pipeline have
following characteristics was displaced by syn- been designed to deliver 55 MMcfd (1.56 x 106 m3/d)
thetic formation water at 28 bars and 65”c: of gas (94% L?12)to the oil field. About half of
the gas is eanc~rked for the cyclic injection test
GOR = 5 m3/m3 B = 1.035 in the western area. The other half will be used
P= 290 Cp P = 0.967 g/cm3 for alternate C02-water injection in the central
area of Bati Raman, as will be described later.
Flooding was continued to a WOR of 100 and a When the demonstration project is proven to be
residual oil saturation of 53% PV (Fig. 4). technically successful and clearly indicates
economic recovery from either or both techniques, a
In the second experiment Dodan gas under pres- C02 recycling plant will be installed at Bati Raman
sure of 28 bars was injected into the bottom of the and C02 injection will be extended to the rest of
waterflooded sample until gas broke through at the the field.
outlet of the cell. Pressure in the core was then
increased to 150 bars. At the end of 24 hours the The western test area covers about 1200 acres
sample was flooded with carbonated brine at 150 (4.75 km2) and will contain 32 wells. At
bars. This resulted in an Sor of 39% PV at WOR = lQO any time about half of the wells will be on injection
and an additional oil recovery of 14% PV. At a and the rest on production. The wells will prob-
WOR = 50 the incremental recovery was 17.5% PV. The ably be operated out-of-phase so that if a
reeults are shown by Curve 2, Fig. 4. particular well is on injecti.n, adjacent wells will
be on production, and vice versa. Injection, at
In a third experiment, the same core sample a wellhead injection pressure of about 1400 psi
was cleaned and dried. It was saturated with water, (97 bars), will be continued until gas breaks
followed by oil. Swi was again 14%. The oil was through in adjacent producers, or until a pre-
displaced with Bati Raman oil which had been satur- determined BHP is reached in the injection well.
ated with Dodan gas at 150 bars. The oil had the When the injection cycle is completed, the well will
following properties: be switched to production, and the adjacent wells to
injection. The cycle durations suitable for
GOR = 70 m3/m3 B = 1.125 different wells may prove to be different, and a
P = 42 Cp P = 0.781 g/cm3 practical compromise will have to be arrived at
after some field experience. A complete injection-
The oil was then flooded with water saturated production cycle is expected to last 4 to 6 months.
with C02, also at 150 bars. Flooding was con- In case of early gas breakthrough in producing
tinued to a WOR of 62, at which point Sor was 38% PV. wells, one or more of the following techniques will
be employed:
The flood performance data, shown in Fig. 5,
are plotted as WOR versus oil recovery. Recovery at 1. Selective injection in reservoir layers
WOR = 50 was 46.5% PV, the same as for experiment 2.
2. Use of plugging agents
Comparison of the three curves indicates that
there was appreciable tertiary oil recovery, and 3. Alternate injection of C02 and water
that the ultimate recoverywas comparable with that
obtained by carbonated waterflooding in one stage, Since the cyclic procedure is expected to show
after saturating the oil with C02. best results in parts of the reservoir containing

465
 .

COO INJECTION AS AN IMMISCIBLE APPLICATION FOR ENHANCED REcoVERy IN HEAVY OIL RESERVOIRS
&
SPE 9928

no (or minimal) mobile water saturation, the and then flooded ~~ithhigh viscosity oil from the
western area of Bati Raman, where no water injection Tar zone until all mobile water had been removed.
had taken place, was selected for its application. The external preosure on the sample and the pore
pressure then were simultaneously increased by
The central test area will encompass 850 psj.(59 bars) to maintain the 1600 pai stress.
nine fully enclosed 62-acre (25 hectare) 5-spot The sample was then flooded with brine to a water-
patterna with a total a~fected area of 990 acrea oil ratio in excess of 100/1 to eatablieh the
(4.00 klrl). Alternate volumes of C02 and water will WOR-cum oil recovery curve, Carbon dioxide waa
be injected into 16 pattern wells serving 25 pro- then introduced into the sample to saturate the oil
ducers. The C02 and water injected will build up and left pressurized at 850 psi for 24 hours.
pressure around the injectors. Production from the Thereafter the sample was flooded with carbonated
producers will be regulated to keep pressure from brine. At a WOR = 50, the incremental oil recovered
building up above about 2000 psi around the varied from 12% PV to 20% PV and averaged 17Z PV
injection wells. The C02 will diffuse into both &or the three samples. It is believed that this
water and oil. The oil, if fully saturated, will increased recovery is significantly more than can
swell by about one-fifth and viscoeity will be be explained by swelling of the oil and reduced
significantly reduced. During the term of the -.?iscosity
caused by the C02. The magnitude of the
demonstration project, C02 will enhance the water increase in recovery is similar to results obtained
displacement mechanism principally by swelling of in cores from wells tested in western New York
oil and viscosity reduction with not much benefit State and tests for Bati Raman, Turkey, but higher
from C02 coming out of solution. In field-uide than those obtained for most of the Mid-Continent
operations, advantage can be taken of this mech- area wells. Three additional samples are being
anism by cycling a group ~f patterns between about tested.
1000 and 2000 psi (67 and 133 bars).
On January 29, 1931, Long Beach Oil Development
WILMINGTON FIELD AND C02 INJECTION Company and the City of Long Beach committed to purc-
hase 7.5 billion cubic feet of C02 (84% pure) in a
The Department of Oil Properties (DOP) of the two-year period with option to renew for three years
City of Long Beach in 1963 considered a test at the same rate (12 million scflday). TF.egas will
injection of C02 into the Tar or Ranger zones in be used in the south 45% of the Tar V zone.
the Tidelands area. The work by Welker and Dunlop27
relating to the large reduction in viscosity of IUOIGEOF APPLICATION
heavy oil by C02 and resulting enhancement of the
brco Process prompted the investigation. Decision The first requirement is the availability of
was made, however, to forego a test at the time a reliable c02 source at a coat low enough to
because of the low oil price and no ready source of justify the project. The target reservoirs for the
low-cost C02. In 1973 the use of C02 was briefly application of immiscible C02 injection are those
considered when results of U. S. Oil and Refining with oil too heavy to achieve miscibility with C02
Compan~’s initial COz injection test in the Ritchie and too deep for practical and economic applicatio;l
Field2 in southern Arkansas were presented to DOP. of steam injection. The characteristics of such
In mid-1975 possible use of C02 waa again considered reservoirs would generally be within the following
when it was learned through THUMS Long Beach ranges:
Company that stack gas (84% C02) could be obtained
from a nearby refinery. Following a preliminary Oil gravity 10 - 25° API
economic evaluation, consideration for use of C02 oil viscosity up to 1000 Cp
was again deferred because of low oil price and the Depth 73000 ft (910 m)
capital cost required to get the gas to the field.
The process should of course be equally appli-
In early 1980, when it was learned that stack cable to lighter oil reservoirs, but unless the
gas from the refinery could be made available by reservoir is too shallow co withstand miscibility
a third party and at a delivery point in the field pressure, miscible displacement should recover more
at a pressure of 1500 psi (103 bars), consideration oil, and is therefore preferred. There may, how-
aqain was given to the use of C02 injection in the ever, be borderline cases where the additional
Tidelands area. recovery may not justify the higher compression cost.

Subsequently, and fortified with some knowledge The obvious alternative processes are steam
of field results of the two C02/heavy oil projects cycling or steam flooding. Current technology con-
in southern Arkansas and laboratory C02 displace- finea their application to shallow reservoirs, no
ment data in the heavy oil reservoir in Turkey, deeper than about 3000 feet (910 m) for two reasons:
displacement tests were run on two preserved core
samples from the Ranger zone, Long Beach Unit. 1. Heat losses in the well bore would render the
Laboratory results were so encouraging that addi- process into hot water injection.
tional core tests were authorized. Figurea 6 - 8
and Tables 2 and 3 show the results of the first 2. The integrity of the well tubulars cannot be
three tests. Each of the samples waa approximately maintained.
2 x 5 inches (5 x 12.7 cm). General procedure was
to place the frozen sample into pliable plastic The current development of the downhole steam
tubing and then into a high-pressure flood cell. generator may make steam injection both practical
The sample was stressed to 1600 psi (110 bara), and ~conomic in deeper reservoirs. If this provea
saturated with brine to replace free gas saturation, to ~e the case, a thorough comparison between steam

466
, .

SPE 9928 A. K. KHATIB, R. C. EARLOUGHER, AND K. KANTAR

and C02 injection would have to be made for each 4* Weber, G.: ItNewRecovery Technique”> .=
prospective reservoir. Gas Jour. (Jan. 25, 1951} 49, No. 38, 171.

CONCLUSIONS 5. Schiffman, L. and Breston, J, N.: “Effect of

Low Pressure Gas Saturation on the Physical
1. The use of carbon dioxide aa an invniscibledis- Properties of Bradford Crude”$ Prod. Monthiy
placement agent should have wide application in (March 1951] 15, No. 5, 36.
the field of enhanced oil recovery.
6. Ralph, H. D.: “New Recovery Tool:$ Oil and Ga~
2. Use of C02 is especially attractive in heavy Jour. (May 3, 1951) 49, No. 52, 53.
oil reservoirs, but it way also be applicable
in a number of light oil reservoirs at shallow 7. Weber, G.: “New-Method Report”, Oil and Gas
depths where pressure required for miscibility Jour. (June 21, 1951) 50, Nc, 7, 154.
cannot be used.
8. Martin, J. W.: ~!Add~t~JnalOil Production
3. Results of field applications to date have demon- Through Flcoding with Carbonated Water”, Prod.
strated that substantial additional volumes of Monthly (July, 1951) 15, No. 9, 18.,
oil can be obtained by the “huff-and-puff”
method as well as by the alternate andlor 9. Saxon, J., Jr., Breston, J. N. and Macfarlane$
simultaneous injection of C02 and water. How- R. M.: “Laboratory Tests with Carbon Dioxide
ever, more than half of the incremental recovery and Carbonated Water as Flooding Mediums”,
probably will come from frontal displacement Prod. Monthly (Nov., 1951) 16, No. 1, 8.
with some combination of alternate or
simultaneous injection of C02 and water. 10. Johnson, W. E. Macfarlane, R. M., Breston, J.N.
and Neil, D. C.: “Laboratory Experiments with
4. For maximum oil recovery by immiscible displace- Carbonated Water and Liquid Carbon Dioxide as
ment it appears that the net quantity of C02 Oil Recovery Agents”, Prod. Monthly (Nov., 1952)
injected should be sufficient to saturate the 17, No. 1, 15.
residual oil and water at reservoir pressure.
11. Martin, J. W.: “Field Data on Operation of Oreo
5. Laboratory displacement tests of p~eserved core Process Near Ri.chburg,N.Y.”, Oil and Gas Jour.
sampies are very desirable in the evaluation (Dec. 13, 1951) 50, No. 32, 82.
of the susceptibility of a reservoir to the
use of C02 for enhanced oil recovery. 12. Martin, J. W. et al: U. S. Patents 2,875,830,
2,875,831, 2,875,832, 2,875,833, Issued in
6. The principal constraint today on the immiscible March, 1959.
use of C02 for enhanced oil recovery is an
adequate source of C02 at low enough cost. 13. Helm, L. W.: “Carbon Dioxide Solvent Flooding
Until recently the low price of oil also was for Increased Oil Recovery,” Trans., AIME (1959).
a major constraint.
14. Helm, L. W. and Josendal, V. A.: “Mechanism of
7. The failure of carbonated waterflooding (Oreo Oil Displacement by Carbon Dioxide”, J. Pet.
Process) 20 to 30 years ago resulted directly Tech. (Dec., 1974) 1427 - 38.
from lack of availability and high cost of
● COO and the belief that the amount of CO~
. 15. McKee, B. C.: “C02: How it Works, Where it
re~uired was only that needed to saturate 25% Works”, Petr Eng (Nov., 1977) 52 - 63.
of the reservoir oil.
16. Gruy Federal, Inc: “Target Reservoirs for C02
ACKNOWLEDGEMENTS Miscible Flooding”, DOE/MC/08341-17 (Oct., 1980)
Vol I, Chapter II.
We wish to thank the management of Turkish
Petroleum Corporation for its permission to publish 17. Christenson, R. J., Hickock, C. W., Ramey,
information on the Bati Raman EOR Project, and H. J. Jr: “Progress Review of the K&S Carbon-
express our thanks to L. W. Brock and John D, ated Waterflood Project”, J. Pet. Tech. (Dec.,
Carmichael, Department of Oil Properties, City of 1960) 20.
Long Beach; E. H. Mayer, THUMS Long Beach Company;
and C. H. Jones, Long Beach Oil Development Co. 18. Durham, E. N, and Watson, J. A.; “C02 Field
for permission to use data relating to the Injection Project Conducted by Phillips
Wilmington Field. Petroleum Company - Ritchie Field. Unfon County
Arkansaa” Tar~et-Reservoirs for c09 Ntscible .
REFERENCES Floodin~$ DOE/MC/08341-17 (Oct., 1980) Vol’~1, 92

1. The Oil & Gas Jour. (March 31, 1980) 82, 93. 19. Durham, E. N. and Watson, J, A.: “C02 Field
Injection Project Conducted by Phillips Petroleum
2. Poettmann, F. H. and Katz, D. L.: “Phase Be- Company - Lick Creek Fj.eld,Bradley and Union
havior of Binary Carbon Dioxide-Paraffin Counties, Arkansas”, op tic, 21.
Systems”, Ind. Eng. Chem. (1945) 37, 847,
20. Elkins, L, F., Pyle, F. M., Maclvor, R. K.:
3. Beeson, D. M. and Orthloff, G. O.: “A Laboratory “Development of the Oreo Process”, October 2,
Investigation of the Water-Driven Carbon Dioxide 1950, unpublished report.
Process for Oil Recovery”, Trans. AIME (1959) 216,

467
 C07 INJECTION AS AN IMMISCIBLE APPLICATION FOR ENHANCED RECOVERY IN HEAVY OIL RESERVOIRS SPE 9928

21. Bradley Producing Corporation unpublished 25. Stalkup, F. I.: “Carbon Dioxide Miscible
reports, Oct. 1, 1950 - Feb. 6, 1951. Flooding: Past, Present, and Outlook for the
Future”; J. Pet. Tech. (Aug., 1978) 1107.
22. Personal communication with P. C. Keith and
D. B. Seifert in November 1973. 26. Genca, M., et al: “Prospects for Enhanced
Recoverv from Bati Raman Field, Turkey”, ~
23. Patton, J. T. et al: “Design of a C02 Future ~f Heavy Crude and Tar Sands, McGraw
Stimulation Process for Heavy Oil Reservoirs” Hill, New York (1981) 388-395. Proceedings
Paper SPE 8897 presented at 50th California of Uiitar Conf., Edmonton, Alberta, June 4-12,
Regional Meeting of SPE, Pasadena, Apr. 9-11, 1979.
1980.
27. Welker, J. R. and Dunlop, D. D.: “Physical
24. Bernard, G. G., Helm, L. W., and Harvey, G. P,: Properties of Carbonated Oils”, 228, 1-873;
“Use of Surfactant to Reduce C09 Mobility in J. ~et. Tech. (Aug. 1963) 873.
Oil Displacement”, Sot. of Pet.-Eng. Jour.
(Aug. 1980) 281-292.

TABLE 1

RESULTS OF WATERFLOOD AND C02 FLOODS

ON CORE SAMPLE FROM BATI RAMAN WELL 211

% Pore Volume
Viscosity
Displacing Pressure Initial Sor @
Oi1 Cp (?65°c
Test Fluid bar so ~ WOR = 100 Recovery ~ Water
1 Water 28 86.1 13.9 52.9 33.2 290 0.65
2 Carb Water 150 52.9 47.1 39.0 13.9 290 0.65
47.1
3 Carb Water 150 86.1 13.9 38.0~ 48.1* i2 0.65

Properties of Core Sample

Dimensions 5.0 x 19.1 cm AAt Wf)R= 62
Pore Volume 83.7 cm3
Porosity 17,1 %
Air Permeability 24.0 md
Swi 13.9 %
Kro at Swi 0.671

TABLE 3

TABLE 2
LONG BEACHONITWELLNO. I1OIA
C02 USED IN COKSTESTS
LONGBSACH UNIT WILL NO. 11OIA
P~SRVSOCO~~ SAFVLS?.F S&WO C02
C02 c02Inj~ct~d inSolution

CP
f?
Cnrb.anated ILsterflmd Tests, % Pore Volume
fter After oil viscosity, A Oil as Mcf/Bbl @ Blowdown
Sple Initial ——N9 C02 Recovery 125°F Sple z Pv of OilProd CFtBbl
~bOsOsrwsO
—.— . ——
so m
——
C02 Tot Oil Water & @ 50WOR @ 50 WOR Total Oil Water
b3 157 29.4 72.5 27.5 57.8 //2.1 14.7 15.7 30.6 472 ,556 6.0 2k0 40
64 138 20.8 7B.5 21.5 41.3 20.0 37.2 21.3 58.5 451 .556
63 11.8 2.7
6S 73.5 16,9 73.7 26.3 51.1 ~ 22.6 23.0 45.6 502 .577 64 19.2 1.5 6.1 450 74
Avg 123 22.4 7b.9 25.1 50.1 30.1 24.8 20.0 44.8 475 .5b3 6S 20.2 1.4 3.8 413 69
AVS m n X3335 G
1600 psi overburden pressure
850 psi carbonation pressure
@ 125”F
> 28
5 Wells, Western N.Y. State
4
0 39 Wells, Mi(l-Continent
-- Increased Oil Recovery &
From Swelling A
4

1
RESIDUAL OIL SATURATION AFTER WATERFLOOD, % PV

Fig. 1 - Carbonated waterflooding tests on fresh cores.

Average values for 8 or more samples per well.

mm

ri
W,lC1,O!I

,tom

Fig. 2 - Performance history. carbon dioxide iniection urolect.

Ba~er Sand Unit, Ritchie Fi~ld, Arkansas, after-Durham,’E.”N. ‘
and Watson, J. A., Ref. 16
 ,.6. ,08-

1 I I [ 1 I I I I ! 1 I I I 1 I I I I 1 I I I I I
1- I [OIL IPROLIJ- ] fi— — -! I

,f.J9.
1o7-

1 1 1 c1,teu t 1 I I I I
:.
. .U I

,& ,06-

11 I ,, _
I 11

,03- 10s-
TIME

m Benefit derived from

m pilot water injection

Fig. 3 - Production history, Garzan Limestone,

Bati RarnanField, Turkey

‘OOf------F--+--7
t b, ‘1
i d

I
I
10 .
I 10 -

5 - I !x

I ~
I u
R
I I 0
RI
1 -
I
.5 - I .5 -
I
I
. lokd~
I
A- .1
10 20 3(I 50 60 0 10 20 30 40 50
OIL RECOVERY, % PV OIL RECOVERY, % PV

Fig. 4 - Tests 1 and 2, waterflood Fig. 5 - Test 3, carbonated water-

followed by C02, Bati Raman Well 211 flood, Bati Raman Well 211
CUMUIATIVh OIL PRODUCED, % PV CUMULATIVE OIL PRODUCED, % PV

Fig. 6 - Sample 63 Fig. 7 - Sample 64

Waterflood followed by C02 Waterflood followed by C02

50 -

10 .
!x
~s
g
D lx
o
0
E o
l-l
1 . m

.5 . m

: m
.1 —
0 10—?0 30 40 50 60
CUMULATIVE OIL PRODUCED, % PV

Fig. 8 - Sample 65
Waterflood followed by C02