Oil Production Control Optimization During Foam Injection in a Water Flooded Reservoir Using Response Surface Methodology

Abstract

Foam injection has been reported as an efficient Enhanced Oil Recovery (EOR) technique for optimizing sweep efficiency and avoiding gas channeling in mature reservoirs. Nevertheless, the financial performance of this process is significantly influenced by the operational control parameters. This study optimized the production of foamy oil from a heterogeneous reservoir by employing response surface methodology and multi-objective genetic algorithms. 300 parameter realizations were generated through a Central Composite Design in MATLAB. The outputs obtained from simulation runs were used to develop surrogate models that predicted Net Present Value (NPV), Return on Investment (ROI), and Payout Time (POT). The prediction accuracy was verified by employing regression analysis with coefficient of determination R2 values greater than 0.995. The Multi-objective Genetic Algorithm (MOGA) was used to identify the optimal parameter set that results in the maximization of economic performance with minimum payout time. The results obtained show that substantial economic performance improvements can be achieved with the optimization of injection and production controls, even without the addition of any new infill wells. A maximum NPV of $10.3067 billion and ROI of 23.4491 were predicted at a minimum POT of 0.0419 years. This proved that a surrogate optimization approach is an effective tool for managing reservoir production as it reduces computational cost with acceptable prediction accuracy.