Comparing Ensemble Kalman Filters for Dynamic Petroleum Reservoir Characterization

A relatively recent method for integration of reservoir production data into reservoir geo-model is ensemble Kalman filter, EnKF . In this method, production data are assimilated sequentially inmodel, forcing the model being updated to the most recent observation. In this work, we investigated the applicability of two deterministic Kalman filters with EnKF for history matchingproblem. These two filters are ensemble transform Kalman filter, ETKF, and deterministic ensemble Kalman filter, DEnKF. Our results indicate that deterministic variants are better for characterization process and among themselves, DEnKF preserves system variance better than others meanwhile ETKF provide estimated map similar to true one

Keywords:

Reservoir characterization , Ensemble Kalman Filter , Ensemble Transform Kalman Filter , Deterministic Ensemble Kalman Filter