SACROC unit location in horseshoe atoll. (Images courtesy of Kinder Morgan Production Co. LP)

The technical and commercial success of miscible CO2 flooding as a technique for enhancing oil recovery has been demonstrated in various floods. Ironically, not much is known, or at least, little is available in the public domain, about the hydrocarbon gas-oil ratio (HCGOR) profile during miscible CO2 flooding.

The SACROC Unit in Scurry County, Texas, encompasses about 95% of the Kelly-Snyder field. The field was discovered in 1948 with approximately 2.75 billion bbl of original oil in place (OOIP). It produces from the Pennsylvanian aged Cisco and Canyon formations on the eastern half of the Horseshoe Atoll (Figure 1).
The early performance of the field indicated that primary production would produce 20% of the OOIP.

In 1972, Chevron initiated a pattern waterflood and tertiary recovery in the form of CO2 injection. Following Penzoil and Devon, Kinder Morgan (KM) took over operatorship of the SACROC unit in June 2000. At the time, the field was near an economic limit, producing 8,500 b/d oil. With large scale miscible CO2 flood, production rose to a peak production near 34,000 b/d of oil and 15, b/d of NGL.

Discussion

The nature of the recovery mechanism in a miscible CO2 flood is by multiple contact miscibility with the hydrocarbon system. This multicontact mechanism causes lighter components of the oil to vaporize into the dense phase CO2. More mixing occurs, and generally, a lighter oil is mobilized. The multicontact mechanism also causes a change in the amount and quality of the produced hydrocarbon gases. This can be demonstrated by investigating the profile of the HCGOR under a CO2 flood compared to primary and secondary recovery processes, like water flooding. In these processes, when pressure is maintained above the bubble-point, a generally constant HCGOR is observed.
In addition to investigating the character of the HCGOR profile during a miscible CO2 flood, other reasons for carrying out this study were that it would:
• Aid in designing gas processing facilities for a CO2 flood;
• Aid in forecasting theoretical recoverable natural gas liquids yield; and
• Improve scoping economics for a miscible flooding project.

The injection scheme KM employs requires that a continuous slug of CO2 be injected for a pre-determined hydrocarbon pore volume (HCPV), followed by water alternating gas (WAG) injection, with the WAG ratio continually being adjusted (getting wetter) until finally chase water injection is reached.

Results from the study are presented as a plot of the ratio of: versus cumulative total HCPV injected (see Figure 2).

HCGORKM refers to the HCGOR during the KM era (2000 to date), and HCGORWF refers to the HCGOR during the waterflood era (1954 to 1972). For this study, the constant HCGORWF during the waterflood era was determined for each pattern to be approximately 1,000 scf/stb.

Patterns with less than 10% cumulative total HCPV injected were excluded from this study as being too young. The generality of the remaining HCGOR profiles reviewed showed remarkably similar profiles. Each profile could be divided into three distinct characteristic segments denoted as:

Early Time. The period beginning from initial CO2 injection to the first water injection in a WAG injection scheme. The end of the early time is denoted by the S-point in Figure 2.

Mid Time. The period beginning from the first water injection in a WAG injection scheme to the point of the last CO2 injection or start of chase water injection. The end of the mid time is denoted by the C-point in Figure 2. The period between the S-point and the C-point is termed the mid time.

Late Time. The period after the start of chase water injection. Due to the in-sufficient maturity of the patterns used in this study (cumulative total HCPV injected), it was not possible to determine the value to which the ratio plotted on the ordinate will reduce (i.e. to the level recorded during waterflood or below that). However, one fact was obvious.

In all of the patterns that had gone on chase water injection, at the inception of chase water injection, the ratio plotted on the ordinate began to decline.

Future study

With similar motivation and methodology for carrying out the HCGOR project, the NGL yield during a miscible CO2 flood was investigated. Preliminary results were reported as a plot of the theoretical yield (bbl NGL/MMcf HC gas) against cumulative total HCPV injected.

The average waterflood theoretical NGL yield for each pattern used in this study was found to be constant at approximately 220 bbl/MMcf HC gas. Preliminary results from patterns used in this study (those still on WAG injection, but not yet on chase water injection) showed a profile similar to the HCGOR profile. Two distinct segments were apparent:

Early Time. Represents the “hump” section in the yield profile. During this period, the yield increases, then decreases, tending to level off to an almost constant value.

Mid Time. Marks the period in the profile where the yield has stabilized to a nearly constant value above the yield during the waterflood era. At this time, it is not clear if the start of WAG injection has any relationship to the transition from early time to mid time in these profiles. More patterns will need to be reviewed.

It is anticipated that upon the initiation of chase water injection, the yield profile will behave in a manner similar to that of the HCGOR profile. It is however unknown at this time, if value will drop below that observed during the waterflood era.

Understanding the results

Figure 3 shows a segmented set of equations representative of an average pattern in the unit. The equations predict HCGOR as a function of flood maturity during the period of modern miscible CO2 flooding by Kinder Morgan. The results of the study reveal that the HCGOR does indeed change after proceeding from a waterflood to a CO2 flood, increasing to a value of nearly twice the value observed during the waterflood era.


Acknowledgement

The author would like to thank the management of Kinder Morgan Production Co. for permission to publish this paper. Particular thanks to Scott C. Wehner and Robert J. Boomer for their valuable suggestions.

This article is based on a presentation given at the 12th Annual CO2 Flooding Conference December 7-8, 2006 titled, “Impact of Miscible CO2 Flooding on HCGOR, A SACROC Unit Case Study.” The full presentation is available at www.spe-pb.org/ attachments/contentmanagers/100/1_4_OlumayowaFamodimu_GORSacroc_KM.pdf