Reduced-Order Model for CO2 Enhanced Oil Recovery and Storage Using a Gravity-Enhanced Process

Abstract

Abstract CO2 flooding offers a means to recover significant amounts of oil while simultaneously sequestering CO2. Estimates of recovery and geologic storage by CO2 enhanced oil recovery (EOR) are highly uncertain owing to reservoir heterogeneity and other reservoir design parameters. One way to optimize the design and characterize the uncertainty of such processes is to introduce a response surface model (RSM), which offers a simple-to-use and useful tool to forecast expected recoveries and storage potential for reservoirs with a broad range of reservoir and design parameters, with potential utility in development of high-level (i.e., regional and national scale) prospective resource assessments. Injectivity is important to economics of CO2 flood enhanced oil recovery, and horizontal wells can greatly increase the injectivity and, when coupled with horizontal production wells, the sweep efficiency, recovery, and storage efficiency in many horizontal and dipping reservoirs. This paper develops a reduced-order model (ROM) for continuous CO2 flooding in heterogeneous oil reservoirs when a horizontal injector and producer are used in a specific optimal pattern. The ROM aims to quickly estimate oil recovery and CO2 storage efficiency based on values of key dimensionless scaling groups. The key scaling groups for first contact miscible (FCM) two-phase (hydrocarbon and water) flow are derived, and their impact on performance metrics of oil recovery and CO2 storage potential are analyzed. First contact miscibility between the hydrocarbon phases is modeled based on the assumption that semi-gravity or gravity stable processes will offer sufficient time for CO2 to contact oil. In this research, a response surface analysis using both a Box-Behnken (BB) experimental design and Latin hypercube sampling (LHS) was also made to predict recovery and storage, and results of that analysis were used to validate the important scaling groups. Monte Carlo simulation with these response surfaces was then used to predict the P10, P50, and P90 expected recoveries and storage efficiencies. The results show that the processes of CO2 EOR and storage in heterogeneous oil reservoirs can be effectively scaled using 11 groups: effective aspect ratio, CO2- or water-oil mobility ratios, buoyancy ratio, normalized initial oil saturation, buoyancy number, residual hydrocarbon saturation, residual water saturation, Dykstra-Parsons coefficient, and correlation length coefficients in x- and z-directions. Furthermore, the results show that oil recovery and CO2 storage efficiencies of 82 to 91% can be obtained in this process for continuous reservoirs with no water influx. This is the first paper that considers a gravity-enhanced process using all horizontal wells, where the process may not be completely gravity stable, to efficiently recovery incremental oil and store CO2. The approach represents a "next generation" CO2 EOR scenario to yield a large volume of CO2 stored per bulk volume of reservoir given unlimited CO2 injection capability, little water influx, and future economic benefits of sequestration and high enhanced oil recovery.