This paper describes a seismically-driven reservoir characterisation study as a key integrated GG secondly, a simultaneous elastic inversion process and last, a seismic facies classification using Bayesian inference to generate, from inverted data, a set of probabilistic cubes of the reference model. The probabilistic petro-elastic facies volumes were then used as input information for conditioning reservoir property distribution, with the aim of capturing the spatial heterogeneity of the reservoir system, describing quantitatively the lateral facies variations and supporting the interpretation of sedimentological characteristics and corresponding petrophysical behaviour in a 3D domain. This quantitative approach has shown the effectiveness for integrating mutli-disciplinary and multi-scale data to support the decision making process, by optimising the location of new production wells and running more accurate volumetrics.
