Geologically Realistic Facies Updates for a North Sea Field

Abstract

Summary In the implementation of any Assisted History Matching process, one of the most important things to address is the reservoir geological structure. The natural transition from structural geological uncertainties towards the rock and fluid uncertainties is represented by the facies distribution. In this study we propose a new facies modeling process that can be used in an integrated modeling workflow. These facies instances must be consistent with the prior knowledge about the reservoir geology (the number of the facies types, the possible transition between facies types, the direction of the facies), and honor the hard data (well log data) and the sand probability cubes (from seismic data). The solution proposed here is the Adaptive Plurigaussian Truncation (APT) model. The APT model consists of generating facies distributions conditioned to prior facies probability occurrence maps. In Sebacher et al., 2013 the truncation map is defined through an internal construction of the model, here it is linked directly with he prior information and experts opinion. The proposed process should be updateablebased on all the available data by an ensemble based method. For the estimation purposes and uncertainty propagation and hence, quantification, we use as history matching (HM) method the Ensemble Smoother (ES)( Skjervheim et al., 2011 )