Multi-objective Optimization using Integration of Experimental Design Methods, Particle Swarm Optimization, and Fuzzy Logic, Case Study: Polymer Injection for Enhanced Oil Recovery

Single-criterion techniques in which just a single objective is considered cannot offer the perfect solution because they cannot take into account the trade-off between conflicting technical and economic conditions. In this study, a multi-criteria algorithm was developed based on experimental design methods, particle swarm optimization, and fuzzy logic. It was able to solve the optimization problem via considering different objectives simultaneously, finding the optimum values of effective factors. To evaluate the efficiency of the workflow, a case study was done in which influential parameters (water flooding duration, polymer concentration, duration of polymer injection, and polymer adsorption) for the design of an enhanced oil recovery operation of polymer flooding in a sandstone reservoir were optimized considering technical (cumulative oil production) and economic (net present value) objectives. The results were compared to the results of the base-case scenario as well as a single objective algorithm (particle swarm optimization). Compared to the base-case scenario, cumulative oil production increased more than 58% and net present value rised from $ 6.9 to 13.1 MM as well. Although the optimum scenario proposed by single-criterion optimization algorithm based on technical objective produced more oil compared to the best solution of the multi-purpose algorithm, a severe reduction was observed in the economic objective simultaneously. Finally, the results of this study demonstrate that multi-objective algorithms are more applicable to precise and realistic decision-making.