Optimization of well placement and assessment of uncertainty

Determining the best location for new wells is a complex problem that depends on reservoir and fluid properties, well and surface equipment specifications, and economic criteria. Various approaches have been proposed for this problem. Among those, direct optimization using the simulator as the evaluation function, although accurate, is in most cases infeasible due to the number of simulations required. This study proposes a hybrid optimization technique (HGA) based on the genetic algorithm (GA) with helper functions based on the polytope algorithm and the kriging algorithm. Hybridization of the GA with these helper methods introduces hill-climbing into the stochastic search and also makes use of proxies created and calibrated iteratively throughout the run, following the idea of using cheap substitutes for the expensive numerical simulation. Performance of the technique was investigated by optimizing placement of injection wells in the Gulf of Mexico Pompano field. A single realization of the reservoir was used. It was observed from controlled experiments that the number of simulations required to find optimal well configurations was reduced significantly. This reduction in the number of simulations enabled the use of full-scale simulation for optimization even for this full-scale field problem. Well configuration and injection rate were optimized with net present value maximization of the waterflooding project as the objective. The optimum development plan for another real world reservoir located in the Middle East was investigated. Optimization using the numerical simulator as the evaluation function for the field posed significant challenges since the model has half a million cells. The GA was setup in parallel on four processors to speed up the optimization process. The optimal deployment schedule of 13 predrilled wells that