Well Placement Optimization Enhancement by Implementation of Similarity-Based Mating Scheme in Non-dominated Sorting Genetic Algorithm-II

An important step in the early stages of an oilfield development is the accurate well placement (production and injection) which has a significant impact on productivity and profitability of the reservoir. Well placement optimization is a complicated task as it is a function of several contributing factors including reservoir heterogeneities and economical constraints. The complexity of this task increases when it is considered as a multi-objective problem rather than a single-objective one as it raises the evaluation time. In this work, three mating procedures, as part of multi-objective optimization algorithm utilized for this purpose, are examined in order to improve the algorithm’s performance in terms of the computational time. Here, for the first time, Similarity-Based Mating Scheme (SBMS) is applied in well placement problem which has been proven to be a viable alternative to conventional mating arrangements.

It is revealed that among the mating schemes evaluated, SBMS, demonstrates a substantial reduction in the number of the generation which is needed to create final Pareto front. Convergence analysis of this technique indicates that this strategy dramatically decreases the time and number of function evaluations by 40 and 150 percent compared to tournament and random selection methods, respectively.