Designing chemical floods for Kansas reservoirs

Reservoirs and leases that have been efficiently water-flooded have the highest performance potential for chemical flooding. This study identified efficient chemical systems for crude oils from nine reservoirs and leases in Kansas. The Pleasant Prairie, Trembley, Vinland, and Stewart oil fields were the most favorable of the studied reservoirs for a pilot chemical flood from geological considerations. Estimates of field applications indicated that chemical flooding is an economically viable technology for oil recovery.

Chemical flood studies

FIGURE 1. Oil production in Kansas is shown by stratigraphic unit between 1989 and 2002 based on information from the Kansas Geological Survey. (Images courtesy of NETL)

The stratigraphic intervals shown in Figure 1 may include many different units, some of which are fundamentally different reservoir types. Many subdivisions of the stratigraphic systems, stages, and groups are productive, and intervals are responsive to waterflood applications. Kansas reservoirs were classified using geological characteristics such as lithology, trap, and drive mechanisms. The Arbuckle Group reservoirs and the Mississippian reservoirs provide 52% of the oil production in Kansas. However, the project team found that neither of these is amenable to chemical flooding due to natural water drives from underlying aquifers.

Consequently, two approaches were used to identify and evaluate potential leases and reservoirs. The first was a broad approach where a database of pertinent information on Kansas oil fields was assembled from available public information. Geological and engineering analyses of the database would then provide a resource base for chemical flooding and target specific reservoirs for study. The second approach was to consult with technical employees of companies producing the largest amounts of oil in Kansas to determine specific leases that would be candidates for chemical flooding. Personal contacts with personnel from the oil producers proved the best approach. Sufficient information was lacking in public data for analysis.

On conducting interviews with several of the largest independent oil producers in Kansas, researchers discovered that the frequency of waterflooding in a specific rock unit was the best indicator of favorable characteristics for chemical flooding. Commonly, waterflooded formations in central and western Kansas include the Missourian Lansing and Kansas City groups; the Pennsylvanian Marmaton, Cherokee, and Morrow groups; the Mississippian Chester group; and the Ordovician Simpson group.

FIGURE 2. In a salinity scan, surfactants and cosolvents are prepared in a series of solutions where only the salinity is varied. These are scans for the formulation containing 0.625 wt% Petrostep S1, 0.375 wt% Petrostep S-2, 2 wt% SBA, 1 wt% sodium carbonate, and 2,000 ppm SNF 3330s with Trembley crude oil at 46.1?C (115?F).

Since chemical flooding is a slug-type process whereby residual oil is mobilized and displaced to production wells by the chemical slug moving through the reservoir, maintaining the integrity of the slug is crucial. Efficient sweep of the reservoir during a waterflood suggests that the fluid-flow characteristics of the reservoir are sufficient to maintain slug integrity and is therefore considered an indicator that a chemical flooding application will be successful. As detailed fluid-flow characteristics typically have not been determined for most mature Kansas reservoirs, other selection criteria must be employed to identify reservoirs that are the best candidates for chemical flooding: oil recovery response, waterflood age and well status, and evidence of flood containment.

During waterflooding, efficient volumetric sweep is indicated by a significant and sustained oil recovery response. Such sweep efficiency within a reservoir points to it being a strong candidate for chemical flooding. Under the second criterion, a reservoir that has a very mature water-flood, where the majority of wells have already been plugged, is not a strong candidate for chemical flooding.

To be successful a chemical flood must be contained within a specified pattern area. Boundaries must be clearly defined, and there should be no evidence of communication between multiple zones in a reservoir. Based on selection criteria, 10 leases were chosen for further study. The producing formations in the fields that were selected include the Lansing-Kansas City, Marmaton, Cherokee, Morrow, Chester, and Simpson.

Formulation, performance of chemical systems

Laboratory studies to formulate chemical systems and to test the performance of the systems in flow experiments through rock material represented a major component of this investigation. Results from this study were used to demonstrate the potential of chemical flooding to independent oil producers with the expressed purpose to generate commitment from producers to engage in a field demonstration of chemical flooding technology. Experimental procedures, analysis methods, and a summary of the results for all the selected leases and fields were conducted. However, in-depth studies were conducted only with the crude oils from the Trembley oil field and the Wahrman leases.

FIGURE 3. Oil cut and oil recovery in sandstone core are charted using oil from Trembley lease T-1 (core No. 2).

Chemical systems were prepared, mixed with the crude oil, and observed for many days in a procedure known as phase-behavior studies. Approximately 16,000 chemical formulations were prepared for these studies. Formulations displaying favorable phase-behavior characteristics were reported. Chemical systems that form middle-phase microemulsions with appropriate characteristics were sought during the phase-behavior studies. Desirable characteristics include microemulsion Type III phases that coalesce and equilibrate in less than seven days; values of the equilibrium solubilization parameters for both oil and brine that are greater than 10; the absence of viscous phases and macroemulsions; and a final mixture that is a one-phase, clear homogeneous mixture at both room temperature (simulated surface mixing facilities) and at reservoir temperature. Formulations exhibiting these criteria have been shown to correlate with efficient recovery of crude oil from rock material.

Core floods were conducted to test the performance of chemical systems identified in the phase-behavior studies. Cores were prepared at residual oil saturation by a water-flood. A chemical slug is injected and followed by the injection of a polymer drive to displace the chemical slug through the rock. The principal measure of performance is the percentage of residual oil that is recovered. Formulations that recover about 90% or better are considered efficient chemical systems.

Phase-behavior studies

Phase-behavior studies identified chemical formulations that met the criteria for efficient performance for eight of the 10 crude oils. One formulation for each crude oil is given in Table 1.

TABLE 1. Phase behavior results are summarized for efficient formulations for several crude oils.

During the studies, formulations were prepared and tested with small changes in the concentration of one component, and frequently several formulations met the criteria. The listed formulation in Table 1 was considered the best formulation because it had the highest optimum solubilization ratio with a sufficient difference between the optimum salinity and the aqueous stability values. Performance of the listed formulations was tested in core flooding experiments. Surfactants in the formulations listed in Table 1 are a combination of an alcohol propoxy sulfate (APS) and an internal olefin sulfonate (IOS). Petrostep S1 and Petrostep S13D are APSs, and Petrostep S2 is an IOS. The APS-IOS surfactant combination has been identified as a preferred system in other studies as well. The IOS is sometimes referred to as glycol butyl ether (EGBE) or diethylene glycol butyl ether (DGBE). Solvents were used to keep the surfactants dissolved in an aqueous solution. Increased solvent concentration increases the aqueous stability and, to a lesser extent, the optimal salinity. Relatively high concentrations of solvent were required to increase the aqueous stability value to a value higher than the optimal salinity value.

However, increased solvent concentration reduces the solubilization parameters. The solvent concentrations were judged high, and reduction in solvent concentration is a target for additional optimization of the formulations. Seven of the formulations contained sodium carbonate, an alkali component. Sodium hydroxide also was tested as an alkali component, but sodium carbonate was selected for most of the formulations due to reported field scaling issues with the use of hydroxide. Alkali was not used to produce soaps since the crude oils have low acid numbers and are not reactive. Alkali was used for the crude oils from carbonate formations because surfactant adsorption in carbonates is reduced when the pH of the surfactant fluid is above 9.5. It was found that alkali also improved the phase-behavior characteristics, in that formulations with alkali were more fluid and less likely to produce viscous phases. Being relatively inexpensive, alkali also was used in formulations for the sandstone reservoirs. Formulations and core floods for Muddy Creek Southwest were prepared with and without alkali.

Phase-behavior studies are mostly conducted by salinity scans. An example of a salinity scan is shown in Figure 2 for a surfactant system and crude oil from one of the selected leases.

Three liquid phases are observed in each tube in Figure 2: an aqueous lower phase, a middle-phase microemulsion, and an upper oil phase. The black line on the tube marks the level of the aqueous chemical formulation in a tube before the addition of crude oil. The middle-phase microemulsion is composed of surfactant chemicals and solubilized oil and water. The volumes of solubilized oil and water in the microemulsion are the volumes of microemulsion above and below the black line, respectively. The “optimal salinity” was determined from phase behavior as the salinity at which the microemulsion solubilizes equal amounts of oil and water.

As seen in Figure 2, these volumes are equal in the tube containing 4.24% sodium chloride. Interfacial tensions between the liquid phases are very low at this condition. If the ratio of the solubilized volumes of oil and water to the amount of surfactant in the system, termed a solubilization parameter, is large, the aqueous system gives excellent performance as an oil recovery fluid in flow tests through rock material.

Oil recovery performance

Oil recovery performance of the identified chemical system is measured in flow experiments in Berea sandstone cores. Foot-long cores were saturated with brine and oil-flooded. The cores were then waterflooded to establish residual oil saturations that are about 35%. During the chemical flood a slug of the chemical systems is injected, followed by a polymer drive. Polymer drives in these tests contained the same or slightly increased concentrations of polymer compared to the chemical slug. Performance was judged on the percentage of residual oil recovered by the chemical flood.

Performance of the chemical systems for crude oils from the Trembley, Wahrman, and Muddy Creek Southwest leases was determined in initial flow tests. The size of the chemical slug, in terms of pore volumes of the Berea cores, and the amount of oil recovered are given in Table 2.

TABLE 2. Oil recovery performance is shown. A recovery of 90% or more is considered successful.

Seventy percent or more of the waterflooded residual oil was mobilized and produced for the three crude oils. Although acceptable, oil recoveries in the 90% to 100% range were anticipated. Laboratory work should continue in an effort to increase oil recovery by testing additional chemical formulations and by improving the chemical flood design.

Core flood results

Core floods were performed to determine oil recovery of the optimized formulations. These floods also were essential to validate the theory of the fluid displacement mechanism and to optimize the surfactant and polymer slug injection design, which includes surfactant and polymer slug sizes, salinity, and polymer concentrations. Core floods were conducted in quarried rocks prepared from Berea sandstone and Indiana limestone.

Twenty-seven floods were conducted in Berea sandstone, and six floods were conducted in Indiana limestone. Core floods for Trembley crude oil featured in this work were performed in Berea sandstone cores. In the example core flood (Figure 3, core No. 2), the core was flooded at 0.15 mL/min (2.1 ft/day) with 0.3 pore volume (PV) of surfactant slug and followed by 1.7 PV polymer drive.

Oil recovery was calculated from the oil displaced during the flood. The oil bank arrived at 0.2 PV, and surfactant breakthrough occurred at 0.67 PV. Oil cut dropped below 1% after 90% residual oil recovery. Oil cut and cumulative oil recovery are plotted in Figure 3. With 90% recovery, the flood should be termed successful and demonstrate that the 0.3 PV surfactant slug proved sufficient for this formulation. Nine core floods with various formulations were performed with Trembley crude at the reservoir temperature of 46.1°C (115°F). All the core floods in the Trembley, Wahrman, and Muddy Creek leases are listed in the final report by the National Energy Technology Laboratory (NETL 2012a).

Acknowledgments

The work described in the article was supported by awards from the US Department of Energy (DOE)/NETL. The author acknowledges permission given by DOE to write the paper. The paper is based on reports presented by the principal investigators: Randy Seright, New Mexico Tech; Stan McCool, University of Kansas; Jeff Harwell, University of Oklahoma; and Beverly Seyler, Illinois Geological Survey. These and numerous other researchers contributed to the information in the paper.