COMPARATIVE STUDY OF WAG-CO2 INJECTION PERFORMANCE IN A DEAD OIL RESERVOIR USING THE BLACK OIL MODEL WITH MRST AND CMG SIMULATORS

Abstract

WAG-CO2 is a promising EOR technique that also contributes to reducing global emissions. Numerical simulation is a crucial tool for capturing complex fluid flow and phase interactions within reservoirs. Since most commercial simulators operate as black boxes, this study compares the performance of an open-source simulator (MRST) with a commercial one (CMG), using a black oil fluid model, with pseudo-miscibility represented through the Todd-Longstaff mixing approach. Five recovery scenarios are simulated: natural depletion, water-injection, CO2-injection, WAG-CO2 injection with yearly alternating cycles, and WAG-CO2 with three-phase segmented injection. Results show varying degrees of deviation in oil production between the two simulators. Compared to CMG, MRST overpredicts oil recovery factor by 0.13% for natural depletion, 0.02% for water-injection, 4.24% for CO2-injection, 1.01% for standard WAG-CO2 injection, and 0.46% for segmented-phase WAG-CO2 injection. In terms of sweeping efficiency, MRST tends to exhibit a broader fluid distribution pattern, although the overall flow behavior remains similar. These findings suggest that MRST slightly overestimates recovery, particularly in gas injection scenarios due to the pseudo-miscibility effect. Nevertheless, MRST offers flexibility and transparency for early EOR evaluations, particularly when numerical control and simplified assumptions are required.