Journal of Petroleum Technology — January 2013
Change Language:
CO2-EOR Mobility And Conformance Control: 40 Years Of Research And Pilot Tests

Carbon dioxide (CO2) has been used commercially for enhanced oil recovery (EOR) for more than 40 years. The CO2-EOR process could be improved if the high mobility of CO2 relative to reservoir oil and water could be reduced effectively and affordably. The water-alternating-withgas (WAG) technology is the preferred method to control CO2 mobility, along with the use of mechanical techniques (e.g., cement, packers, well control, infield drilling, and horizontal wells) to help control the CO2-flood conformance. If the next-generation CO2-EOR target of 67 billion bbl is to be realized, new solutions are needed that can recover significantly more oil than the 10–20% of the original oil in place (OOIP) associated with current flooding practices. This literature review concentrated on the history and development of CO2-mobility-control and profilemodification technologies.


Every day, more than 110 field projects in the US, 60% of which are in the Permian Basin in west Texas, inject 3.1 Bscf of CO2 into sandstone and carbonate formations to recover more than 280,000 bbl of oil. This corresponds to approximately 5% of the nearly 5 million BOPD of US oil production. The number of CO2-flood projects has increased steadily in recent years, in contrast to other EOR methods. The limiting factor for expanding CO2 flooding is the availability of large volumes of high-pressure CO2 because most CO2 suppliers are operating at full capacity. For example, new projects in the Permian Basin are constrained by the need for an additional 0.3–0.4 Bscf/D of CO2 in the region. Conversely, the rapid growth of CO2 EOR in Mississippi reflects the growth of CO2 supplies in that region.

Approximately 74.4% of the CO2 used for EOR comes from gas-treating and -processing facilities associated with the production of CO2-rich natural gas, while 19.4% originates from natural-gas plants, 4.8% from coal-synfuel plants, and the remainder from various chemical and petroleum facilities. The dehydrated, high-pressure CO2 from various sources is transported to oil fields where it is compressed to the desired injection pressure, combined with recycled CO2 and injected into the reservoir.

Regardless of whether a reservoir has been waterflooded, typical primary recovery is approximately 5–15% of OOIP, while secondary recovery is responsible for an additional 20–40% of OOIP. Although CO2 is capable of displacing most of the oil from the portion of the porous media through which it flows, miscible-CO2 floods typically recover 10–20% of the OOIP by injecting a volume of dense CO2 that is equivalent to approximately 80% of the hydrocarbon (oil) pore volume (HCPV). Immiscible- CO2 floods recover only 5–10% of OOIP because of the interfacial tension between the CO2 and viscous oil. As a result, 35–65% of the OOIP remains unrecovered after CO2 flooding. Further, on the basis of the domestic totals of 3 Bscf/D of CO2 consumed for the production of 280,000 BOPD, the average CO2-use ratio is 10,700 scf of CO2/bbl of oil, a much larger solvent/oil ratio than desired.

The fundamental causes of this low oil recovery can be traced to the density and viscosity of dense CO2. First, the low density of high-pressure CO2 relative to oil promotes gravity override of the CO2, reducing oil recovery in the lower portions of the formation. Second, the viscosity of dense liquid or supercritical CO2 at typical CO2-flooding conditions is approximately 0.05–0.10 cp, a value so much lower than typical oil- and brine-viscosity values that it results in an unfavorable mobility ratio. The result is viscous fingering, which in turn leads to early CO2 breakthrough, high CO2-use ratios, delayed CO2 production, depressed oil-production rates, and a low percentage of recovered OOIP. These problems can be made worse when the injection well is completed in two or more zones. The low viscosity of CO2 promotes its flow into the more-permeable layers that have been waterflooded effectively, while small amounts of CO2 enter the lower-permeability zones that contain more recoverable oil.

The objective of this paper was to review chemical means for addressing CO2-flooding mobility-control and conformance-control problems that stem from the low density and viscosity of CO2 at reservoir conditions. The density of the CO2 entering the formation cannot be altered significantly by the addition of a CO2-soluble additive or implementation of an injection strategy. The density of CO2 is, essentially, a function of only temperature and pressure, and the mitigation of gravity override must be accomplished by mobility control or conformance control. It is possible, however, to alter the mobility of dense CO2 by reducing its relative permeability through WAG-injection strategies, increasing its viscosity with the addition of direct CO2 thickeners, or decreasing its mobility by generating CO2-in-brine foams. It also is possible to alter the distribution of the injected CO2 into a layered formation favorably, especially if the injected fluids are diverted from high-permeability watered-out thief zones into lower-permeability oilrich zones. Although conformance control is achieved to some extent by each of these mobility-control strategies, there are techniques such as gels, foam gels, or dispersions of swellable preformed particle gels designed specifically to block high-permeability zones.


Mobility control has been accomplished most readily with injection of water and CO2 into the formation, usually in an alternating sequence that promotes near-wellbore injectivity and diminishes mobility away from the wellbore. The alternating injection of brine and CO2 does not make the CO2 more viscous; rather, it increases the water saturation and thereby decreases the CO2 saturation within the pores. Many studies have demonstrated that reducing the CO2 saturation reduces the relative permeability of CO2. This, in turn, lowers the mobility ratio and inhibits the formation of viscous fingers.

The WAG process has been used for several decades. However, WAG injection is not suitable for tight reservoirs or water- sensitive reservoirs—continuous CO2 injection is more appropriate in these cases. CO2 floods conducted in the 1980s used a constant WAG ratio throughout the entire CO2 flood, with CO2-slug sizes on the order of 1% of HCPV. In 1989, Amoco implemented a tapered WAG in the Slaughter and Wasson fields of west Texas. This variable-WAG process has since been implemented by most operators. In the variable-WAG process, a relatively large slug of dry CO2 is introduced initially, followed by a tapered WAG ratio that is increased incrementally from values less than unity to values greater than unity, a process referred to as wetting the WAG.

Direct CO2 Thickeners

Polymeric Thickeners. Thickening CO2 by use of polymers has proven extremely difficult. CO2 is a feeble solvent for extremely-high-molecular-weight polymers (M>1,000,000). Although a few polymers with lower molecular weights (M=1,000 to 1,000,000) have been designed or identified that can dissolve in liquid or supercritical CO2, the pressure required for dissolution of dilute concentrations typically is in the 10,000- to 40,000-psia range. These pressures are significantly higher than the typical minimum miscibility pressure associated with CO2 floods (1,200–4,000 psia). There also has been little success in the use of small, associating molecules to thicken CO2, primarily because CO2 is an extremely poor solvent for the polar and ionic associating groups that are commonly incorporated into small-molecule thickeners.

Small-Molecule Thickeners. The second strategy explored for CO2 thickeners has been the design of small molecules that can associate and form viscosity-enhancing macromolecular structures. In each case, the molecule contains a segment that is sufficiently CO2-philic to promote dissolution of the compound in CO2. The compound also contains one or more CO2-phobic moieties that are intended to be attracted to or associated with the CO2-phobic moieties of neighboring molecules, thereby establishing a viscosity-enhancing, associating, noncovalently bound, macromolecular network. Trialkyltin fluorides have the ability to induce large viscosity increases in light alkanes at dilute concentration (e.g., approximately three orders of magnitude at 1 wt%). The tin atom is slightly electropositive, and the fluorine atom is slightly electronegative, while the three butyl arms extending from the tin atom enhance the solubility of the molecule in alkanes. These molecules apparently form linear, transient, associating polymers in solution as the tin atom of one molecule is attracted to the fluorine atom of the neighboring molecule. The butyl arms do not interfere with these associations and help to stabilize the linear macromolecule. Unfortunately, neither tributyltin fluoride nor other synthesized trialkyltin fluorides were CO2-soluble enough to serve as a CO2 thickener, although some success was realized in thickening propane and butane. After establishing that fluorination of alkyl groups could enhance CO2 solubility, tri(2-perfluorobutyl ethyl) tin fluoride was synthesized. Although this compound was soluble in CO2 without the need for a cosolvent, the viscosity increase was far less than expected.

Attempts to thicken CO2 by heating a mixture of CO2 and metallic-stearate powders also were unsuccessful. When this mixture is heated in hydrocarbon oils, the attractive forces between these compounds are weakened, enabling the compound to dissolve in the oil and form viscosity-enhancing metallic stearates as the solution cools. However, even at high-temperature and -pressure conditions, the metallic stearates could not dissolve in CO2.


There have been many successful laboratory-scale tests involving water-soluble surfactants capable of stabilizing CO2-in-brine foams. Thirteen reports of pilot tests conducted between 1984 and 1994 have been published; most of these pilot tests were aimed at attaining conformance control. Five of these projects were considered successful technical efforts, and favorable economic assessments were associated with most of them. It appears that the emergence of robust gel-based conformance techniques (including monomer solutions that polymerize and crosslink the gel in situ, polymer solutions that crosslink in situ, foams that gel in situ, and preformed particle-gel dispersions) coupled with WAG for mobility control may have led to a decline in the use of foams as a conformance-control technique, especially in extremely-high-permeability flow paths where foams generally are ineffective. These gel methods appear to be more effective and robust than CO2 foams, as demonstrated in a series of generally successful field tests beginning in the late 1970s and continuing to today. Many laboratory-scale studies were directed at the design of mobility-control foams. This led to two pilot tests solely aimed at CO2-mobility control, one of which indicated that a 60% in crease in the apparent viscosity of CO2 occurred where the foam formed. Other pilot tests were designed to increase the apparent viscosity of CO2 and block a high-permeability zone; one test demonstrated that CO2 foams could enhance conformance control and mobility control simultaneously. Recently, several ideas for CO2 foams have been proposed. CO2 foams generated with CO2-soluble nonionic surfactants have been tested successfully in the laboratory and through an ongoing pilot test. Further, laboratory-scale testing of foam stabilization with water-dispersible nanoparticles has been initiated in an attempt to circumvent problems often associated with surfactant solutions flowing for extending periods of time through a porous medium, such as adsorption losses and chemical instability of the surfactant.

The results of 40 years of research and field tests indicate that mobility and conformance control for CO2 EOR by use of thickeners, foams, and gels can be attainable technically and economically for some fields. However, significantly more research must be conducted to improve the technology and the economics.

An affordable CO2 thickener has been recognized as an important technology for more than 25 years but has not yet been developed. The design of such a thickener is a much more challenging problem than was envisioned several decades ago, primarily because of the low CO2 solubility (or complete CO2 insolubility) of compounds that contain the chemical groups responsible for viscosity-enhancing intermolecular associations. Despite several technical advances and the successful design and synthesis of an expensive fluorinated CO2 thickener, an inexpensive analog has yet to be identified. Development of an affordable CO2 thickener that could change the CO2 viscosity to that of the oil being displaced would have profound effects on oil recovery and is worth pursuing.

The improved performance of the gel technologies in blocking flow paths, enhanced gel robustness, the lack of a pressing need to reverse foam-conformance-control treatments, and operator ability to apply gel treatments in formations with fractures or highly permeable open flow paths have made gels popular tools for conformance control during a CO2 flood. It appears that unless significant advances are made, persuading operators to consider CO2- conformance-control foams as an alternative technology may be difficult given the mixed successes and disappointments during a decade of CO2- conformance-control-foam field tests and the refinement of robust, alternative technologies that use chemical gels. However, there seems to be consensus that the CO2-conformance-control foams are less expensive and more readily reversible (by water injection, if desired) than any of these gel treatments. Given the ability of foams to be designed for conformance control or mobility control, it may be prudent to combine the two technologies such that gels are used for conformance control and CO2-in-brine foam (rather than WAG) is used for mobility control.

Research has demonstrated that surfactant-induced CO2 foams are an effective method for mobility control in CO2-foam flooding, but they have potential weaknesses. Because the foam is by nature ultimately unstable, its longterm stability during a field application is difficult to maintain. However, a convincing argument can be made that the potential of CO2-mobility-control foams from water-soluble surfactants has not been explored fully in pilot tests, especially given the immense body of promising laboratory-scale technical knowledge that has been reported. CO2-soluble surfactants, which are being tested in the field at the Scurry Area Canyon Reef Operators Committee project in west Texas, ensure that the surfactant appears (and the foam forms) only where the CO2 flows. These surfactants may be capable of providing a modest degree of conformance and mobility control, are easy to implement even for operators who use only continuous CO2 injection, and may reduce the need for alternating slugs of brine greatly. New nanoscience technologies may provide an alternative for the generation of stable CO2 foam. The use of nanoparticles to stabilize CO2-mobility-control foam may overcome the long-term instability and surfactant-adsorption-loss issues that affect surfactant-based CO2-EOR processes.

This article, written by Senior Technology Editor Dennis Denney, contains highlights of paper SPE 154122, “Mobility and Conformance Control for CO2 EOR—Thickeners, Foams, and Gels: A Literature Review of 40 Years of Research and Pilot Tests,” by R.M. Enick, SPE, University of Pittsburgh; D. Olsen, SPE, IBM Global Business Services; J. Ammer, US Department of Energy, National Energy Technology Laboratory; and W. Schuller, URS Corporation, prepared for the 2012 SPE Improved Oil Recovery Symposium, Tulsa, 14–18 April. The paper has not been peer reviewed.

For a limited time, the complete paper is free to SPE members at