Field Development Optimization under Uncertainty Using an Ensemble-based Simulation Approach

Field development optimization under geological uncertainties defines a major challenge for reservoir management. Historical production data is used to calibrate reservoir simulation models and thereby reduce the variability of geological uncertainty distributions. The practice of using a single history matched model to evaluate forecast scenarios is questionable, as they prove to be a poor basis for including uncertainty descriptions. In recent years ensemble-based approaches have been increasingly applied in history matching workflows to account for consistent uncertainty descriptions. Any uncertainty description becomes relevant for field development planning, however, it is less obvious how to integrate the ensemble into a field development optimization process. In this paper we use an Ensemble Kalman Filter technique (EnKF) to generate a history matched ensemble. A workflow is developed to identify representative ensemble members based on key performance indicators. All selected cases are included in an integrated field development optimization process. A risk-averaged objective function definition is used to evaluate field development scenarios for multiple ensemble members. Full uncertainty distributions for optimized field development scenarios are derived on the bases of the full history matched ensemble. The workflow is applied to a real field case and results are discussed.