Field-scale simultaneous optimization of well rate and polymer concentration is challenging due to the geologic heterogeneity, large number of wells, and operational constraints. This paper proposes a multi-stage optimization strategy to achieve a multi-objective goal: maximize cumulative oil production and minimize cumulative polymer injection. The proposed optimization strategy follows a sequential step: well rate optimization followed by polymer concentration optimization. In the first stage, streamline-based method is used to adjust the water injection rate and liquid production rate for hundreds of wells. This method equalizes the well pair efficiency that quantifies the amount of oil recovered per barrel of water injection, to maximize oil recovery. Key injectors are selected from the first stage and included in the multi-objective optimization for the second stage. A novel stochastic approach, multi-objective global and local surrogate-assisted method (MO-GLS), is utilized to optimize polymer concentration, to maximize cumulative oil production while minimizing cumulative polymer injection. This method considers a global prescreen and local search to obtain the trade-off solutions along the pareto front based on a dominance relation. Proxy model is built inside and updated during the iteration to improve optimization efficiency and particle swarm optimization is adopted as the optimizer for a fast convergence rate. After multi-objective optimization, multiple trade-off solutions can be found along the pareto front, which have lower parameter uncertainty compared to the initial population. The optimization results are visualized using streamlines, which provide insights into the improvement in sweep efficiency from rate and concentration optimization. The sequential optimization workflow conducted in this paper provides theoretical basis and operational recommendations for the optimization of well rate and polymer concentration. The practical feasibility of the approach is demonstrated through a large-scale field application.
