Gas injection is one of the most commonly used enhanced oil recovery (EOR) methods. However, there are multiple problems associated with gas injection including gravity override, viscous fingering, and channeling. These problems are due to an adverse mobility ratio and cause early breakthrough of the gas resulting, in poor recovery efficiency. A Water Alternating Gas (WAG) injection process is recommended to resolve these problems through better mobility control of gas, leading to better project economics. However, poor WAG design and lack of understanding of the different factors that control its performance might result in unfavorable oil recovery. Therefore, this study provides more insight into improving WAG oil recovery by optimizing different surface and subsurface WAG parameters using a coupled surface and subsurface simulator. Moreover, the work investigates the effects of hysteresis on WAG performance.
This case study investigates a field named Volve, which is a decommissioned sandstone field in the North Sea. Experimental design of factors influencing WAG performance on this base case was studied. Sensitivity analysis was performed on different surface and subsurface WAG parameters including WAG ratio, time to start WAG, total gas slug size, cycle slug size, and tubing diameter. A full two-level factorial design was used for the sensitivity study. The significant parameters of interest were further optimized numerically to maximize oil recovery.
The results showed that the total slug size is the most important parameter, followed by time to start WAG, and then cycle slug size. WAG ratio appeared in some of the interaction terms while tubing diameter effect was found to be negligible. The study also showed that phase hysteresis has little to no effect on oil recovery. Based on the optimization, it is recommended to perform waterflooding followed by tertiary WAG injection for maximizing oil recovery from the Volve field. Furthermore, miscible WAG injection resulted in an incremental oil recovery between 5 to 11% OOIP compared to conventional waterflooding. WAG optimization is case-dependent and hence, the findings of this study hold only for the studied case, but the workflow should be applicable to any reservoir. Unlike most previous work, this study investigates WAG optimization considering both surface and subsurface parameters using a coupled model.