SPE 88641

CO2 Recovery and Utilization for EOR

Kamal Morsi, ADNOC (E&P); John Leslie, SNC-Lavalin International Inc.; and Doug Macdonald, SNC-Lavalin
International Inc.

Copyright 2004, Society of Petroleum Engineers Inc.

• A desire not to radically modify the existing fired
th
This paper was prepared for presentation at the 11 Abu Dhabi international Petroleum equipment configuration.
Exhibition and Conference held in Abu Dhabi, U.A.E., 10–13 October 2004.
It was agreed that the one essential CO2 recovery process
This paper was selected for presentation by an SPE Program Committee following review of
information contained in a proposal submitted by the author(s). Contents of the paper, as selection criteria is commercial viability within a schedule that
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to permits a 2009 plant startup. In effect, at the very least a
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at process would have to have one semi-commercial application
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper about to start construction at the present time.
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300
Pre- and post-combustion schemes for CO2 capture were
words; illustrations may not be copied. The proposal must contain conspicuous evaluated including:
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Oxygen + CO2 as Boiler / Reformer Combustion Gas. This
process (“oxy-fuel”) uses combustion in an oxygen-enriched
Abstract
stream to obtain flue gas with a higher concentration of CO2
In order to replace valuable natural gas presently used for
and a smaller volume. It would require major modifications to
pressure maintenance ADNOC is investigating capture of
the existing fired equipment, plus installation of an air
carbon dioxide (CO2) from several flue gas sources and its
separation unit. It has been tested at bench scale.
transport to a developed field for Enhanced Oil Recovery
(EOR). A study was commissioned to evaluate available
Hydrocarbon Gasification. Gasification is well understood
technologies and determine the optimum technical and
and commonly used for many refinery and petrochemical
economic configuration for capture and delivery of high
applications. It involves the partial oxidation of a
quality CO2 for injection.
hydrocarbon stream using either air or oxygen. The gas
Three flue gas sources for CO2, were considered,
stream is at high pressure and is rich in CO2. As with oxy-
appropriate capture technologies were chosen and a formal
fuel, a key constraint for this project is the desire to minimize
licensor selection process was performed. Key
changes to the existing fired equipment.
considerations in selection of technology and licensors were
existence of successful commercial applications and utilities
Chemical (Amine) Absorption. Existing commercial
requirements (natural gas, power and cooling water).
installations recovering CO2 from flue gas are based on
The process configuration chosen involves placing the
chemical absorption using some form of alkanolamine solvent.
absorber equipment near the flue gas sources, with a
The CO2 in the gas reacts with the amine in an absorber, and
common stripping operation located at a single site.
then is recovered from the amine through addition of heat in a
The study also included design of compression and
second column. For flue gas applications the incoming gas
pipeline facilities to deliver dense-phase CO2 to the
nominated field, and booster pumps to raise the CO2 to must be cooled to about 50°C. Impurities such as SOX and
required injection pressure. NOX cause the amine to deteriorate and their removal may be
The paper describes the process employed to select required to maintain a reasonable amine cost.
technologies and discusses the considerations involved to
arrive at the final configuration. Gas Absorption Membranes. These are a subset of chemical
absorption schemes, employing a membrane to promote
Technology and Process Selection absorption of the CO2 by the amine. They appear to be a
The main characteristics of flue gas that drive process promising alternative to conventional packed towers, offering
decisions in CO2 recovery for ADNOC are: the possibility of significantly reduced equipment sizes. The
technology is in use in the North Sea to strip CO2 from
• Low pressure of the flue gas; produced natural gas for re-injection and storage in a deep
• Relatively low CO2 content and thus low CO2 partial saline aquifer.
pressure and large raw flue gas volume;
• High temperature of the raw flue gas; Gas Separation Membranes. These are used commercially
• Presence of trace contaminants such as SOx, and NOx; for gas processing both onshore and offshore for streams at
high pressure and CO2 content. The driving force for these
2 SPE 88641

membranes is difference in partial pressure across the Design Basis

membrane, which should be about 20bar. A visit to the site confirmed the overall design basis for all
parts of the project and identified issues specific to the
Cryogenic Separation. CO2 can be separated from other project’s location and configuration.
gases through cooling and condensation. This well-
understood commercial process is widely used to purify CO2 Required CO2 Quantity and Quality. Approximately
streams for food or industrial use when the feedstock already 100MMscf/d (5 million tonnes per year), 99.0+% pure after
has a CO2 content in excess of 90%. dehydration; water dew point 0 deg. C or lower.

Physical Absorption. This is a well-established method of Potential CO2 Sources. Flue gas from (1) Steam reformer
removing CO2 from gas streams. The CO2 is physically furnace at a refinery: (2) Steam reformer furnace at a fertilizer
absorbed by the solvent. CO2 solubility is proportional to plant: (3) General Utility Plant boilers (GUP) serving both
pressure and inversely proportional to temperature. The plants. The three sources are located within 5 km of each
solvent is regenerated by pressure reduction or by heating, other.
depending on the process design and solvent being used.
Physical absorption would likely be the technology of choice CO2 Destinations. (1) Injection into an oil reservoir for
for CO2 capture in a hydrocarbon partial oxidation/gasification enhanced oil recovery, nominally 100 km from the CO2
system, where the syngas is typically at about 20bar (300 recovery sources; (2) return to the fertilizer plant for their
psia). internal use.

Adsorption. This is a commercial process for removing small Configuration. Absorbers located adjacent to each flue gas
amounts of CO2 from produced hydrocarbon gas streams, source, with a common regeneration, drying and compression
generally used at higher pressures and lower CO2 facility located at the GUP.
concentrations (400 ppmv to 1.5% by volume). The gas
stream flows through a bed of a solid material with a very high CO2 Transportation. In dense-phase form, via pipeline.
surface area that preferentially adsorbs the CO2, which is then Compressor discharge pressure was set at approximately
recovered by depressuring or heating the bed. 144bar; a function of injection pressure requirements and
pipeline design.
CO2 from PSA Purge Gas. During initial process reviews it
was noted that part of the fuel for the refinery reformer CO2 Injection. The existing natural gas injection system for
consists of a purge gas stream from a Pressure Swing the field will be modified as necessary for CO2 Injection.
Adsorption (PSA) Unit. The composition of the stream is
about 47% CO2, offering potential for a less expensive Key Issues:
approach for capture of some of the project’s CO2 • Availability and consumption of water for flue gas
requirements. A hot carbonate chemical absorption process quench. UAE desalinates all of its fresh water and this
was considered. makes it more expensive than in some other places.
• Cooling water temperature is expected to be high. A
Licensor Selection potential trade-off is to use seawater cooling with
Although there are proprietary differences among various associated extra heat exchanger metallurgy costs.
proponents of amine-based CO2 recovery the base remains the
• Space is tight at two of the CO2 source sites for
same:
equipment to process large volumes of flue gas at low
• Cool the flue gas; pressure.
• Remove or otherwise cope with impurities such as SOx • Amine CO2 recovery and transportation processes
and NOx; typically require large amounts of rotating horsepower for
• Absorb the CO2 in a lean amine solution; feed blowers and CO2 compression, as well as large
• Use low-grade heat to regenerate the amine and release a amounts of low-grade heat for amine regeneration. It was
pure CO2 stream; decided to produce steam at 61bar and use back pressure
• Dry and compress the CO2 for transport; steam turbines to generate shaft horsepower required for
• Cope with potential corrosion issues associated with wet blowers and compressors along with the required
CO2. quantities of 3.4bar steam to regenerate the amine and
From several possibilities, two licensors of amine-based release the CO2.
CO2 recovery were selected who had commercial applications Figure 1 shows a block flow diagram of the overall
in similar conditions. Each licensor was provided with a project components.
design basis and asked each to prepare a technical and
commercial proposal. CO2 Recovery Process Licensor Selection Criteria
Within the confines of the design basis, each licensor was free
to decide on optimum process conditions, percentage
recoveries from each source, and equipment configurations.
Each licensor was asked for a budget estimate, which was then
SPE 88641 3

checked against in-house knowledge and normalized for nominally about 100 km long and follow an existing main
differences in scope of the packages. Part of the study crude oil line between the oil field and the terminal.
mandate was also to produce conceptual capital and operating Two cases were evaluated: (1) 164bar compressor
cost estimates for the entire project. discharge at GUP and booster pumps at the oil field; (2) No
Licensors were evaluated on the following criteria, using booster pumps, compressor discharge pressure at the GUP
a simple weighting and decision analysis procedure: sufficient to meet the injection pressure at each well.
• Similar commercial plant experience;
• Capital cost; Reservoir Performance
At this stage of study, the reservoir work done has been
• Operating costs and utility consumption;
theoretical. It is based on knowledge of how other oil
• On-line availability; reservoirs have responded to CO2 flooding plus existing data
• Simplicity of design and operation; for the oil field and its behaviour with natural gas injection.
• Flexibility to handle conditions outside of design limits; The reservoir analysis was carried out by a well-known
• Start-up, shutdown and safety requirements; Canadian reservoir consultant with extensive experience in
• Willingness to offer process guarantee; CO2 EOR.
• Licensor fee and royalty terms; About 1.45 MMscf/d (76 TPD) of CO2 is required to
replace 1 MMscf/d of natural gas injected, based on a
• Potential for process improvements and ADNOC access
volumetric equivalent.
to them;
Due to CO2 dissolving in both oil and water and potential
• Experience with Front End Engineering Design package swelling of the reservoir oil, there is reason to believe that the
preparation. actual amount of CO2 required could be higher.
The density of dense-phase CO2 is such that successful
Compression and Dehydration injection at reservoir pressure is likely to require a minimum
Six compressor vendors were asked to provide a preliminary Well Head Injection Pressure (WHIP) that is 45-60% of that
design and budget price for this service. All six responded. required for natural gas injection.
Centrifugal compressors are recommended, driven by Pure CO2 will likely be miscible in the oil at reservoir
back pressure turbines using medium pressure steam. conditions.
Total power demand is approximately 30 MW (40,000 While initial CO2 breakthrough is expected in a few years,
HP). Regeneration steam required is approximately 250 t/h volumes high enough to impact solution gas treating
(based on 3.4bar steam). operations may not occur until much later.
Compressors of this type and pressure range are Three processing technologies can be considered for
commonly used in urea production, although generally at separation and recycle of CO2 produced as part of the
lower throughput. Except for the challenge presented by the associated gas from the oil reservoir:
high flow associated with one 100% unit, design and supply of
• Selective absorption using amines;
the compressors should be technically feasible and straight
forward. • Cryogenic separation by cooling the gas and condensing
Four vendors provided technically acceptable packages the CO2;
for this service. • Membrane separation of CO2.
All the acceptable vendors have units in similar service in All three processing routes for separation of CO2 from
terms of pressure range, but at generally lower throughput. produced gas at the field exist in commercial applications
Most of these units were supplied for urea production. elsewhere. Cryogenic separation and membranes become
All the vendors proposed one 100% capacity unit, and relatively more attractive than they were for flue gas CO2
stated they did not see any significant problems with the removal owing to higher gas stream pressures and increased
design and supply of this large a unit. Cost savings appear to CO2 concentrations. The eventual choice can be based on
be significant for one unit as compared to two 50% units. relative economics, which will be influenced by the actual
However, it appears that the 100% unit would be much larger stream composition and by the configuration of existing
than compressors previously supplied for CO2 service and solution gas processing operations.
therefore it is recommended that this issue should be more Historically CO2-EOR floods have increased ultimate oil
thoroughly discussed with each vendor during final tender. recovery by 10 – 15% of original oil in place.
Middle Eastern reservoirs tend to have higher pressures
Pipeline and Injection System and temperatures than most existing CO2 floods. To our
CO2 transport pipelines are well established in the US, with knowledge, this will be the first case where CO2 has followed
about 2400 km of CO2 pipelines operating. A dense-phase natural gas injection.
pipeline to carry 95 MMscf/d of 98% pure CO2 was recently Based on expected reservoir performance, modifications
built for an EOR project in Weyburn, Canada. to existing production equipment and gathering lines are
To simulate the pipeline and injection system, it was anticipated to be minimal, at least in the short run.
assumed that either the existing natural gas injection network
would be used, or the new network would parallel existing Project Capital and Operating Costs
lines. It was also assumed that the new CO2 line would be Estimated capital and operating costs are stated in USD third
quarter 2004. The estimates are conceptual, based on partial
4 SPE 88641

equipment lists and are expected to have a +/-30% confidence Booster pumps are not required based on a minimum
interval. wellhead injection pressure (WHIP) of 137bar (2000 psig). A
Total Installed Capital Cost (TIC) is expected to be compressor discharge between 150-170bar (2200-2465 psig)
approximately USD 300 million. is expected, depending on line size. If reservoir studies and
It is expected that the facilities could be designed and field work indicate that the WHIP is significantly higher, say
constructed within a 36- month period. 170bar or more (2500 psig), then booster pumps would likely
Operating costs, including capital charges assessed at a be required.
nominal rate of 10% of TIC per year, are expected to be in the Provided the water dew point of the gas stream can be
order of USD28-31/tonne or 1.47-1.63/mcf of CO2 injected, maintained at a very low level to avoid corrosion, the pipeline
using a nominal natural gas value at the field of USD 1.10 per can be standard carbon steel with modifications and provisions
million BTU. for CO2 transport, nominally 12 inches diameter.
Total installed cost for a 12-inch diameter line is
Conclusions estimated to be in the order of USD50-60 million. Cost for a
Technology and Process Selection. Pre-combustion schemes 16-inch line is estimated to be in the order of USD60-70
such as oxy-fuel and gasification do not fit the mandate of this million, including contingency.
study, as they would involve major changes to fired The above costs could potentially be reduced through
equipment. value engineering. Examples are increasing the pipeline yield
Chemical (amine-based) CO2 absorption schemes offer strength, or eliminating the internal coating allowance.
the only commercial operating experience. There are multiple Although booster pumps would reduce the cost of the
licensors of the chemical absorption process who have pipeline, the booster station cost substantially outweighs any
commercial applications. savings.
Gas absorption membranes would result in smaller Subject to confirmation of integrity, the existing natural
equipment sizes and offer some promise but no commercial gas distribution system can be used for CO2 injection. Some
experience in this application. components may have to be replaced but these costs are
Gas separation membranes offer the most promise for an expected to be minor.
eventual alternate to chemical absorption but have no Health and safety issues associated with transport of CO2
commercial operations. The chief impediment to successful are manageable, and in general, significantly less than those
implementation of gas separation membranes is the high associated with the transport of gases containing H2S. CO2 is
differential pressure required to make them work. Combined not toxic or flammable, but it is heavier than air and could
with the low pressure of the flue gas, this yields very high pose a threat of asphyxiation under certain circumstances.
operating costs for compression of the entire flue gas stream. Isolation valves and a leak detection system should be
Other schemes (Cryogenic Separation, Physical included.
Absorption, Adsorption) are not suited to flue gas
applications, once again because of the higher pressures Reservoir Performance. Replacement of natural gas
required to make them effective. injection with CO2 is a technically viable option for pressure
Chemical absorption of CO2 from the PSA purge gas maintenance and could lead to an increase in ultimate oil
stream was investigated but did not appear to be attractive due recovery.
to the requirement to compress the stream to about 20bar for A significant amount of reservoir engineering, laboratory
treatment and to the resulting reduction in overall CO2 analysis and field testing remains to be carried out to confirm
recovered from the refinery reformer. the reservoir parameters developed for the conceptual design.
Several schemes (Membranes, Chemical Absorption,
Physical Absorption) may be amenable for use in CO2 Project Capital and Operating Costs. The calculated unit
removal from higher-pressure solution gas streams produced operating cost of approximately USD 28-31 per tonne or 1.47-
from the reservoir post-breakthrough of the injected CO2. 1.63 / mcf of CO2 injected is within the range of expectations
for CO2 sourced from flue gas.
Process Licensor Selection. As of the date of publication of The actual economics of substituting CO2 for natural gas
this paper, specifics of actual licensor selection were not in this situation will depend on the nature and size of the
publicly available. enhanced oil recovery effects in the reservoir.

Compression and Dehydration. There are no technical Next Steps

issues associated with compression and dehydration. In the ADNOC expects to decide early in 2005 if it intends to
next phase a decision will need to be made if sufficient proceed with the project. Initial activities will involve
confidence exists to save capital cost by installing a single reservoir engineering studies and pilot injection of CO2 to
very large unit. confirm reservoir parameters.

Pipeline and Injection System. Transport of CO2 from the Nomenclature

sources to the oil field is technically feasible, using a ADNOC = Abu Dhabi National Oil Company
combination of new pipeline and the existing natural gas BTU = Higher Heating Value, British Thermal Unit
distribution system.
SPE 88641 5

EOR = Enhanced Oil recovery. In this paper the term refers

exclusively to EOR using CO2 miscible flooding
techniques.
GUP = General Utility Plant, one of the CO2 sources; a steam
and other utility generation operation serving the
plants containing the other two CO2 sources
HP = Nominal rated horsepower
MMscf/d = million standard cubic feet per day
MW = Megawatts
ppmv = parts per million by volume
PSA = Pressure Swing Adsorption; a process commonly used
in hydrogen manufacture to purify hydrogen and reject
CO2 and other impurities
T, or tonne = metric tonne; 1000 kg
TIC = Total Installed Cost
USD = USA dollars, basis third quarter 2004
WHIP = Wellhead Injection Pressure; in this case the
minimum pressure required to successfully inject a
fluid

Figure 1

C O 2 R E C O V E R Y A N D IN JE C T IO N S C H E M E

PRODUCED LEAN GAS

RECYCLED CO2
FLUE GAS CO2
RECOVERY NGL

B O IL E R S
AND CO2 RECOVERY
C O 2 P IP E L IN E
REFO RM ERS PLANT & IN J E C T IO N
100 K M
C O M P R E S S IO N O IL /G A S C R U D E O IL
CO2 S E P A R A T IO N
PLANT
F L O W L IN E S O IL /G A S
R E F IN E R Y

S U B -S U R F A C E
C O 2 R E C O V E R Y S IT E CO2 IN J E C T IO N S IT E