An efficient hybrid methodology for optimization of CO2 Huff-n-Puff EOR and sequestration in tight oil reservoirs

Abstract

CO2 Huff-n-Puff is one of the effective enhanced oil recovery (EOR) methods in tight oil reservoirs, which is also attractive to sequestrating CO2 underground. Therefore, CO2 Huff-n-Puff EOR and sequestration project is particularly important both for oil and gas industry and environmental protection field. However, the technique has not gained industry-wide adoption partly due to knowledge gaps in efficient optimization of CO2 Huff-n-Puff EOR and sequestration using compositional numerical simulation. The purpose of this work is to propose an efficient hybrid optimization methodology which is called proxy-model-based optimization with initialization constraint. Two cases for single objective optimization and multi-objective co-optimization of CO2 Huff-n-Puff EOR and sequestration with two objective functions were investigated. Our findings suggest that the optimization performance of proxy-model-based optimization is highly dependent on the relative errors of the proxy model compared with simulator-based optimization. The optimization performance of proxy-model-based optimization with initialization constraint is 4∼10 % lower than that of simulator-based optimization, but the reduction in the number of simulations and improving the robustness of a stochastic optimizer are very attractive. In addition, different objective functions and optimization mode have different sensitive parameters. To the best of our knowledge, this is the first time to conduct single/multi-objective optimization of CO2 Huff-n-Puff EOR and sequestration in tight oil reservoirs based on numerical simulation technology, method of response surface experimental design and single/multi-objective particle swarm optimization algorithm. We believe the proposed methodology in this work is efficient especially for large reservoir simulations and rapid prediction of optimization schemes.