The goal of this project is to conduct a carbon dioxide (CO₂) injection test at the Citronelle oilfield in Mobile County, AL. The project will introduce CO₂-enhanced oil recovery (CO₂-EOR) for tertiary recovery from Alabama’s uniquely structured sandstone reservoirs, providing oilfield operators and CO₂ producers with improved estimates of oil yields from EOR, and the capacity of the depleted reservoirs to sequester CO₂. The research will improve the reliability of computer simulations of oil yield from CO₂-EOR, calculations of sequestration capacity, and the rate at which CO₂ can be introduced into underground formations. The simulations of the Citronelle field will be integrated with computer visualizations of the migration of oil, water, and CO₂ with the results made accessible to reservoir engineers, geologists, utility planners, and researchers studying carbon sequestration.

University of Alabama at Birmingham, Birmingham, AL
Alabama Agricultural and Mechanical University, Normal, AL
Denbury Resources Inc., Plano, TX
Geological Survey of Alabama, Tuscaloosa, AL
Southern Company, Birmingham, AL
University of Alabama, Tuscaloosa, AL
University of North Carolina at Charlotte, Charlotte, NC

Carbon dioxide-enhanced oil recovery is a well-established method for increasing oil recovery from the Permian Basin oilfields of Texas and New Mexico and from the Williston Basin in North Dakota and Montana. Denbury Resources has also been successful in applying the technique in Mississippi oilfields. Typically, ten percent of the original oil present in a reservoir at the start of production can be recovered using CO₂-EOR. A recent study by Advanced Resources International of Arlington, VA, estimated that 64 million additional barrels of oil could be recovered from the Citronelle field using this technology. When oil production is complete, the reservoir and adjacent formations can provide sites for storage of CO₂ produced from the combustion of fossil fuels in power plants and other industrial processes that generate large quantities of CO₂. Southern Company is evaluating the capacity of such reservoirs as possible sites for permanent sequestration of CO₂ that has been separated from coal and natural gas combustion products produced by its power plants.

The Citronelle field is an ideal site for CO₂-EOR and sequestration from both reservoir engineering and geological perspectives. The field is mature and waterflooded, with existing infrastructure, including deep wells, and consists of fluvial-deltaic sandstone reservoirs in a simple structural dome. Presence of the regionally extensive Ferry Lake Anhydrite seal, four-way structural closure, and lack of faulting make the field naturally stable with respect to CO₂ storage. However, the geology of the heterogeneous siliciclastic rocks in this field is very different from fields where CO₂-EOR has been applied commercially, such as in the carbonate strata of the Permian and Williston basins. The proposed demonstration will introduce CO₂-EOR for tertiary recovery from Alabama’s uniquely structured energy resources and thus realize benefits to the nation from additional petroleum production.

A successful demonstration of the technology in this project could lead to new commercial CO₂-EOR and sequestration efforts across the nation (including reservoir types where CO₂-EOR has thus far not been applied) offering a potential two-for-one solution to the United States' energy security and environmental concerns.

The initial principal focus (January 1, 2007 to August 31, 2008) was on selection and preparation of the test site and thorough studies of its geology, environmental conditions (air, soil, and vegetation) and produced fluids, as well as seismic imaging and establishing the background conditions prior to CO₂ injection.  Researchers determined and documented the petrology, sedimentology, and stratigraphy of the Rodessa Formation in the vicinity of the test site at an unprecedented level of detail and performed realistic and informative reservoir simulations and visualizations. A study thoroughly documented the environmental and ecological background conditions surrounding the site.  Seismic signals were recorded under baseline water flood conditions.  The minimum miscibility pressure and absence of precipitation from oil in the presence of CO₂ were established.  An economic analysis identified the optimum CO₂ slug size for water-alternating-gas oil recovery under specified CO₂ cost and oil price constraints. All indications were that the pilot test would provide an unequivocal demonstration of CO₂-enhanced oil recovery and essential data and simulations on which to base a commercial CO₂ flood in the Citronelle Field. The CO₂ storage capacity of depleted oil reservoirs and saline formations in Citronelle Dome was estimated by static calculations to be between 0.5 and 2 billion short tons of CO₂, sufficient to sequester the CO₂ from coal-fired generation at a nearby 1500 MW (electric) power plant for 40 years. 

A parametric study of WAG recovery using the MASTER 3.0 reservoir simulator showed that a properly designed WAG recovers as much oil as continuous CO₂ injection. Using the simulation results, three-dimensional animations were programmed showing the evolution of fluid saturations in Donovan Sands 14-1 and 16-2 during two CO₂ injections of 7500 tons each, followed by water. The animations nicely capture the mobilization of oil by CO₂, development of the oil bank, the role of water in driving the bank, and the residual oil left unrecovered.

Carbon dioxide injection began in December 2009.  After resolution of some initial difficulties in pumping and replacement of tubing, CO₂ injection continued without significant interruption at an average rate of 31 tons/day from January 27 to September 25, 2010, in good agreement with the average rate of 35 tons/day anticipated by reservoir simulations.  The 8000 tons of CO₂ allocated for the test were injected.  Oil produced from three wells in the inverted five-spot test pattern (B-19-7, B-19-8, and B-19-9) was gathered, along with production from five other wells to the north and east, at Tank Battery B-19-8.  Oil produced from Well B-19-11 in the test pattern went to Tank Battery B-19-11, along with production from three other wells to the west and south.  Oil production at B-19-8 Tank Battery responded immediately to CO₂ injection and increased steadily (from 45 to 59 bbl./day) throughout the injection.  Early breakthrough of CO₂ at Well B-19-11 in the southwest corner of the test pattern resulted in the loss of oil production from that well.  Breakthrough also occurred soon after at two other wells to the northeast and southwest of the injector. 

Shear-wave velocities were measured to depths of 12,500 feet using the Refraction Microtremor technique.  Wireless geophones were placed along two straight paths spanning 30,100 and 25,600 feet to the south and southwest, respectively, of the injection well.  Shear-wave velocities recorded before and during CO₂ injection suggested a 10 percent increase in stress associated with CO₂ injection in the geologic layers above the injection zone. An interesting systematic increase in the dependence of shear-wave velocity on depth was observed during injection of CO₂ and water at depths below ~4500 feet.  

A pressure-transient test in the injector, conducted in November and December 2011, provided strong evidence for the presence of a 600- to 1000-foot-long vertical fracture intersecting the injection well. The presence of a fracture was completely unexpected because the injection well had never been intentionally hydraulically fractured and the hydrocarbon-bearing sands at Citronelle are free of natural fractures. Although the direction of the fracture cannot be determined from the results of the pressure-transient test, a fracture in the direction of maximum horizontal compressive stress in the southeastern U.S. would explain the early excessive breakthrough of CO₂ and loss of oil production from Well B-19-11 in the southwest corner of the test pattern and the early breakthrough of CO₂ at other wells located in that approximate direction relative to the injector. The avoidance or management of hydraulic fracturing will be an important component of the reservoir management plan during a commercial CO₂ flood. 

Two problems were experienced upon returning to water injection following completion of the CO₂ injection in September 2010:  (1) excessive erosion-corrosion of the downhole power-oil pumps by particles and scale mobilized by the CO₂ and water and (2) loss of injectivity to water.  These changes were accompanied by a dramatic reduction in oil production (from 59 to 21 bbl/day) at B-19-8 Tank Battery.  The erosion-corrosion problem was solved by replacing the power-oil pumps, as they failed, with pumps having longer stroke and fabricated using harder materials.  As the pumps were replaced, oil production at B-19-8 Tank Battery increased to between 33 and 44 bbl./day.  Attention was then focused on understanding the loss in injectivity to water following the injection of CO₂.  

Injection profile tests run in the injector before, during, and after CO₂ injection showed that the loss in injectivity to water following the CO₂ was due to reduced injectivity in only one of the two target sands.  Injectivity to water decreased from approximately 200 to only 30 bbl water/day in the upper sand (Sand 14-1) and remained the same, at approximately 40 bbl water/day, in the lower sand (Sand 16-2).  Sand 14-1 is the finer of the two sands. 

Two tests were conducted to investigate the cause of the loss in injectivity to water:  (1) treatment of the injector with surfactant to relieve capillary blocking, and (2) treatment of the injector with hydrocarbon solvent and asphaltene dispersant to remove heavy ends that might have been left near the well bore when CO₂ extracted lighter components.  A leak developed in the tubing during the first treatment with surfactant.  The team discovered while repairing the leak that the internal plastic coating in the tubing had been flaking off and replaced it with bare steel tubing before continuing with the surfactant treatment.  A transient increase in injectivity to water was observed during the surfactant treatment, but it had no permanent effect on the water flow rate.  Likewise, the solvent and dispersant treatments had no lasting effect.  Following these tests, the injector was flowed back in an effort to flush out the flakes of internal plastic coating.  Flow-back resulted in a dramatic increase in the water injection rate, but the effect was short-lived.  Within 2½ months, water injection had returned to its typical post-CO₂-injection rate.  However, the response of the injection rate to flow-back indicates that plugging of perforations and the formation by fragments of the internal plastic coating was the most likely explanation for low injectivity to water.  The failure of the internal plastic coating was determined to have resulted from a lapse in quality control on the part of the tubing supplier. 

Recent oil production at B-19-8 Tank Battery stands at 33 to 42 bbl/day, compared with 51 to 62 bbl/day before CO₂ injection.  The principal reason for the loss in oil production is the loss in water injectivity due to deterioration of the internal plastic coating in the tubing.  Recent oil production at B-19-11 Tank Battery stands at 21 to 30 bbl/day, compared with 48 to 54 bbl/day before CO₂ injection.  The principal reasons for the loss in production are the reduced injectivity to water and loss of production from Well B-19-11 due to the fracture intersecting the injector, and most likely running in the direction of maximum horizontal compressive stress, roughly northeast to southwest. 

Lessons Learned 
The performance of triplex positive displacement pumps can be problematic when injecting liquid CO₂.  This is a potential problem for small-scale tests using liquid but not for commercial-scale projects using pipeline CO₂. 

Power oil pumps of the type currently in use in the Citronelle Field need to be modified using harder materials and adjusted for longer stroke in order to minimize erosion-corrosion by particles and scale mobilized by CO₂-water. 

Some of the problems experienced during this pilot test were associated with the transition from CO₂ back to water injection.  Problems such as erosion-corrosion of the power oil pumps, failure of the internal plastic coating in the tubing, and loss of injectivity to water might be avoided by continuous CO₂ injection, rather than WAG. 

Detailed understanding of geology, including connectivity of sands and propensity for hydraulic fracturing, are important considerations in well placement. 

Thorough environmental evaluation before and during CO₂ flooding will go a long way to reassuring stakeholders that CO₂-EOR and storage are safe and secure. 

Large-scale, high-resolution simulations are a useful tool for optimizing well placement and anticipating the profitability of a commercial CO₂-EOR project. 

A pilot test, such as the one conducted here, can provide valuable experience and reveal unexpected potential problems. 

All the approved project research work has been completed as of January 31, 2014. The final report is available below under "Additional Information".

$3,464,549

$4,325,543 (55 percent of total)

NETL – Chandra Nautiyal (chandra.nautiyal@netl.doe.gov or 281-494-2488)
University of Alabama at Birmingham – Peter Walsh (pwalsh@uab.edu or 205-934-1826)

Final Project Report [PDF-17.6MB] June, 2014

Publications 
J.C. Pashin and R.A. Esposito, "Citronelle Dome: A Giant Opportunity for Multi-Zone Carbon Storage and Enhanced Oil Recovery in the Mississippi Interior Salt Basin of Alabama," Annual Convention and Exhibition of the American Association of Petroleum Geologists, Long Beach, CA, April 1-4, 2007.

R.A. Esposito, J.C. Pashin, and P.M. Walsh, "Citronelle Dome: A Giant Opportunity for Multi-Zone Carbon Storage and Enhanced Oil Recovery in the Mississippi Interior Salt Basin of Alabama," 2007 Annual Convention of the Gulf Coast Association of Geological Societies and the Gulf Coast Section of the Society for Sedimentary Geology, Corpus Christi, TX, October 21-23, 2007.

R. A. Esposito, J. C. Pashin, and P. M. Walsh, "Citronelle Dome: A Giant Opportunity for Multi-Zone Carbon Storage and Enhanced Oil Recovery in the Mississippi Interior Salt Basin of Alabama," accepted for publication in Environmental Geosciences, June 2008.