WAG Optimization of Egg Model with Brazilian Pre-Salt Fluid Using PSO Algorithm

Abstract

Abstract Water Alternating Gas (WAG) is a consolidated enhanced oil recovery method that outperforms secondary recovery methods such as waterflooding or gas injection. Optimization of the WAG parameters can increase the cost function and improve revenues, but the algorithms usually implemented require high computational resources and time. In this context, a bio-inspired algorithm, Particle Swarm Optimization (PSO) is used to determine the best candidate for only two parameters, water and gas injection time, resulting in the determination of the WAG cycle and WAG ratio, thus drastically reducing the complexity of the problem. The proposed long-term optimization algorithm was applied in a modified version of a well-known reservoir benchmark Egg model, in which the fluid composition was adapted to resemble Brazilian pre-salt reservoir fluids and WAG injection. Moreover, it is shown that the best results obtained consider longer periods of water injection and improve the NPV by 5.5%.