4D Seismic History Matching With Ensemble Kalman Filter - Assimilation on Hausdorff Distance to Saturation Front

Abstract

Abstract Distance to water front parameterization has been implemented and capability of ensemble kalman filter to handle large number of measured data on top of the uncertain parameters has been deployed to benefit from the richness of 4D seismic observation as spatial information in the reservoir. Distance to the observed and predicted water fronts has been calculated using an accurate Fast Marching method and the Hausdorff distance has been calculated to provide the misfit term to perform EnKF updates during the assimilation steps that the 4D observation is available. Need for proper scheme of localization has been revealed in this study to both ameliorate the quality of the history match and also the posterior geological models. The proposed workflow has been compared to the direct approach of using cell by cell inverted water saturation as a 4D seismic derived observation to perform the history match. Result showed not only an important gain in the obtained match but also significant advantage in preserving the geology of post match models.