Optimal Parameter Selection in a Polymer Alternating Gas PAG Process

Abstract

Abstract As the majority of conventional reservoirs are reaching maturity, the attention is gradually shifting towards unconventional and heavy oil reservoirs. Steam flooding, Miscible/Immiscible CO2 flooding, Polymer flooding and Water Alternating Gas (WAG) are the most common EOR techniques currently being employed in the industry. However, it has been recently proved that Polymer alternating gas (PAG) is a better alternative to the other EOR processes as it provides improved sweep efficiency (Zhang et. al. 2010, Li and Schecter, 2014, Kong et. al., 2015). A PAG process alternately inject miscible CO2 and water mixed with polymer. Different parameters directly affect the performance of this method. Thus, it is important that these parameters are carefully selected to increase the recovery along with the profitability. In this paper, an attempt has been made to optimize a PAG process for five production and five injection wells using two different global optimization algorithms. One injection cycle of a PAG process constitutes of two stages – (i) Injection of miscible CO2 (ii) Injection of polymer. CO2 dissolves in oil, swelling it and decreasing the viscosity thereby increasing the mobility of oil. The injection of polymer increases the sweep efficiency and reduces viscous fingering. In this paper the operational parameters that have been selected for optimization are well locations, number of injection cycles required, production BHP, CO2 injection rate, CO2 injection time, polymer injection rate and polymer injection time and the concentration of the polymer that needs to be injected. The algorithms used are the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) and the Particle Swarm Optimization (PSO). A synthetic reservoir model was used to study the performance of the method. The Net Present Value (NPV) was used as the objective function to gauge the suitability of the solution and hence optimize the different parameters to obtain the highest NPV. Both used optimization algorithms yielded NPV that were well within the acceptable range. Additionally, sensitivity studies were conducted which showed that PAG gave a higher recovery efficiency when compared to the other EOR processes. Although some case studies have been performed to show the performance of the PAG process, to the best of our knowledge, no studies have been conducted on the use of global optimization algorithms for the estimation of the operational parameters of the PAG process.