Bayesian Seismic Inversion for the Joint Estimation of Facies and Elastic Properties of the Lula Field

Abstract

Summary In this work we apply a Bayesian framework for the joint estimation of facies and elastic properties from seismic data. The statistical model accounts for three-dimensional spatial correlation. Due to the high dimension of the problem, the algorithm is based on a MCMC method for drawing multiple realizations of the reservoir properties. The main objective of this work is to evaluate the capability of the algorithm to distinguish the different carbonate facies. The method is applied to a large seismic data set of Lula field along with six wells used to define the prior distributions and to evaluate the inversion results.