Use of Full-Field Simulation to Design a Miscible CO2 Flood
- F.P. Brinkman (Exxon Co. U.S.A.) | T.V. Kane (Exxon Co. U.S.A.) | R.R. McCullough (Exxon Co. U.S.A.) | J.W. Miertschin (Exxon Production Research Co.)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Evaluation & Engineering
- Publication Date
- June 1999
- Document Type
- Journal Paper
- 230 - 237
- 1999. Society of Petroleum Engineers
- 5.3.2 Multiphase Flow, 5.4.2 Gas Injection Methods, 5.4.1 Waterflooding, 4.6 Natural Gas, 4.1.2 Separation and Treating, 5.1.5 Geologic Modeling, 5.5.2 Core Analysis, 5.4 Enhanced Recovery, 1.6 Drilling Operations, 4.9 Facilities Operations, 5.4.9 Miscible Methods, 4.1.5 Processing Equipment, 2.2.2 Perforating, 5.5.8 History Matching, 5.1.1 Exploration, Development, Structural Geology, 5.2 Reservoir Fluid Dynamics, 1.6.9 Coring, Fishing, 4.3.4 Scale, 4.1.9 Tanks and storage systems, 5.7.2 Recovery Factors, 5.8.7 Carbonate Reservoir, 5.5 Reservoir Simulation, 5.2.1 Phase Behavior and PVT Measurements, 5.6.1 Open hole/cased hole log analysis
- 1 in the last 30 days
- 646 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
A study using full-field reservoir modeling optimized the design of a miscible CO2 flood project for the Sharon Ridge Canyon Unit. The study began with extensive data gathering in the field and building a full-field three-dimensional geologic model. A full-field simulation model with relatively coarse gridding was subsequently built and used to history match the waterflood. This waterflood model highlighted areas in the field with current high oil saturations as priority targets for CO2 flooding and generated a forecast of reserves from continued waterflooding. Predictions for the CO2 flood used an in-house four-component simulator (stock tank oil, solution gas, water, CO2. A full-field CO2 model with more finely gridded patterns was built using oil saturations and pressures at the end of history in the waterflood model. The CO2 model identified the best patterns for CO2 flooding and was instrumental in selecting a strategy for sizing the initial flood area and in determining the size, location, and timing of future expansions of the CO2 flood.
The Sharon Ridge Canyon Unit (SRCU) is located in West Texas, about 70 miles northeast of the city of Midland. The Unit covers 13,712 acres. Fig. 1 shows the Horseshoe Atoll, a trend of more than 40 oil fields covering several West Texas counties. SRCU is geologically continuous with the Diamond M Unit and the giant Kelly-Snyder Field (SACROC Unit) to the northeast. Production is from the Canyon Reef formation, a thick carbonate buildup of late Pennsylvanian Canyon and Cisco age, and occurs at an average depth of 6600 feet. There are active CO2 floods in this formation at SACROC, Reinecke, and the Salt Creek field.
Sharon Ridge was discovered in 1949 and developed on 40 acre spacing by 1953 with about 340 wells. The reservoir initially contained undersaturated oil at 3135 psi. Production was by expansion drive until 1952 when pressure fell below the bubble point of 1850 psi over most of the field. In 1955, the field was unitized and a peripheral waterflood was started to stabilize reservoir pressure. The waterflood is now at a mature stage with oil recovery approaching 50% of the original oil-in-place (OOIP). There has been limited infill drilling with 22 wells drilled at 20-acre spacing.
Screening studies identified SRCU as a good candidate for a miscible CO2 flood project. These studies included core flood displacements, pattern element simulation models, and detailed evaluations of similar fields with CO2 floods. Laboratory core displacements showed a remaining oil to waterflood of over 40% with subsequent injection of CO2 reducing oil saturation to less than 10%. Simulations with small element models have also shown significant incremental oil recovery from injection of CO2 at SRCU. SRCU has reservoir properties similar to SACROC which has reported significant additional oil recovery from miscible CO2 flooding (Ref. 1).
The goal of full-field modeling was to design a miscible CO2 flood with maximum economic potential. Key issues for project design include the amount and location of remaining oil, reservoir sweep efficiency, flood rate, gas injection volume, strategy for handling increased produced gas, and projection of continued secondary operations. To address these issues, we built three different full-field three-dimensional (3D) models: geologic model, coarse-grid waterflood model, and fine-grid CO2 flood model. Recent advances in computer technology made this approach possible as opposed to the prior approach of running type-element models and scaling up those results to field rates. The approach of using field-scale simulation models to study optimizations for another CO2 flood in West Texas has been reported in Ref. 2. Thus, advancing technology and prior experience led us to embark on this ambitious approach to use full-field modeling to design our CO2 flood.
The reservoir is a thick carbonate buildup that is predominately limestone. Fig. 2 shows the structure on the top of the reservoir. Geographic areas of the field have been named: North End, South End, and Southeast Pinnacle. The topography is extremely variable, with the hydrocarbon column averaging 115 feet and ranging to a maximum of 450 feet in the South End area of the field. A large portion of the North End has over 90 feet of gross reservoir thickness above the original oil-water contact. Table 1 is a summary of reservoir rock and fluid properties.
The reservoir has been subdivided into five depositional sequences or zones, four of which are shown in Fig. 3. The lower zones (4, 5) are found over almost the entire field while upper zones (1, 2, 3) are more areally restricted. Zones are usually separated by intervals of low porosity limestone with few shales in the reservoir. Most wells drilled during initial field development did not penetrate the entire reservoir, thus limiting description of the lower zones. A more detailed discussion of the geologic setting and depositional facies is available in Ref. 3.
Building a full-field 3D geologic model of SRCU presented several unique challenges, including having modern porosity logs on only a few wells and only 90 full penetrations of the reservoir. To address this problem of limited data, an extensive data acquisition program was implemented. This program included deepening 19 wells, coring 11 wells, and obtaining 49 miles of new two-dimensional (2D) seismic lines. After gathering these data, all new and old core, well log, and seismic data were integrated to develop a sequence stratigraphic reservoir framework.
|File Size||549 KB||Number of Pages||8|