A new approach in well placement optimization using metaheuristic algorithms

Abstract

The need to determine the optimum well placement which causes maximum cumulative oil production has dramatically raised in recent years. Hence, there is a need for a reliable and accurate tool to diagnose the possible solutions in a cost- and time-effective manner. Two of the most commonly used algorithms in this area are Particle Swarm Optimization (PSO) and Genetic Algorithm (GA). These two algorithms, the similar used methodologies for finding the optimum well location, suffer from local optimization issues. These algorithms are slow, and their found points is not accurate. Therefore, in this study, a new meta-heuristic optimization algorithm known as the Black Widow Optimization (BWO) is employed to optimize the location of five production wells. The results are compared to those of PSO and GA. To this end, the cumulative oil production is intended as the objective function, and a simplified model of an actual reservoir is simulated. Afterward, the well placement is optimized. The results revealed a higher improvement of the objective function for the BWO algorithm compared to PSO and GA. The new algorithm was faster and resulted in a better optimum point. Concisely, the BWO algorithm reduced the associated time complexities up to 9.19% and 18.29% compared to GA and PSO, respectively. The PUNQ-S3 reservoir model was also employed to validate the results.