An improved optimization procedure for production and injection scheduling using a hybrid genetic algorithm

This paper proposes an improved optimization workflow for oil production and water injection allocation for oil reservoirs under waterflooding by optimizing both oil production and water injection rates. Suitable initial estimations of water injection rates (from streamlines injection efficiency) and optimization of injection allocation are included in this work in addition to the work of Azamipour et al. (2016) to increase oil production. Besides, genetic algorithm (instead of simulated annealing) coupled with polytope search is used in two steps; a model with coarse grid blocks and a model with fine grid blocks for the reservoir description. NPV is considered as objective function. The results of applying proposed workflow for a field sector model used from (Azamipour et al., 2016) show that the reservoir recovery is increased by optimized oil production and water injection rates. The total oil production is changed from 391.2 Msm3 to 447.4 Msm3. The comparison to the previous work, shows the superiority of both initial guess for injection rates and optimized injection well rates. Finally, the proposed method is implemented in another case study with a different heterogeneity to show the generality of the proposed approach. Besides, the differences of simultaneous and separate optimization for scheduling of oil production and water injection are discussed.