Biobjective optimization for general oil field development

• Developed a new biobjective optimization procedure for general field development.
• Method is based on a recent MINLP approach that includes global and local search.
• Successfully applied new methodology to biobjective field development problems.
• Method used to identify optimal tradeoff between NPV and cumulative recovery.
• Biobjective strategy also applied for optimization under geological uncertainty.

The optimization of oil field development and production planning typically requires the consideration of multiple, possibly conflicting, objectives. For example, in a waterflooding project, we might seek to maximize oil recovery and minimize water injection. It is therefore important to devise and test optimization procedures that consider two or more objectives in the determination of optimal development and production plans. In this work we present an approach for field development optimization with two objectives. A single-objective product formulation, which systematically combines the two objectives in a sequence of single-objective optimization problems, is applied. The method, called BiPSOMADS, utilizes at its core our recently developed PSO–MADS (Particle Swarm Optimization–Mesh Adaptive Direct Search) hybrid optimization algorithm. This derivative-free procedure has been shown to be effective for the solution of generalized field development and well control problems that include categorical, discrete and continuous variables along with general (nonlinear) constraints. Four biobjective field development and well control examples are solved using BiPSOMADS. These examples include problems that consider the maximization of both net present value and cumulative oil production, and the maximization of both long-term and short-term reservoir performance. An example that highlights the applicability of biobjective optimization for field development under geological uncertainty is also presented. This usage of BiPSOMADS enables us to maximize expected reservoir performance while reducing the risk associated with the worst-case scenario.