Seismic Driven Probabilistic Classification of Reservoir Facies and Static Reservoir Modeling

A complete workflow for reservoir characterization is hereby proposed to derive seismic driven facies classification by means of an integrated probabilistic framework which includes a multistep inversion of seismic data and seismic facies classification. The proposed methodology overcomes some common assumptions in reservoir modelling such as Gaussian distribution of rock properties by means of more flexible Gaussian Mixtures. Moreover the method is based on robust physical models for each step of the workflow: inversion of seismic data, estimation of petrophysical properties from seismic attributes and facies classification from seismic derived information. In particular the so obtained seismic facies preserve the link and the discriminability with both elastic data and petrophysical properties. Seismic driven facies and the associated probability can be directly integrated as a prior trend into reservoir properties geostatistical simulations. An example of application of such workflow is hereby presented. The studied case is a clastic reservoir in Barents Sea.