Reconciling Geology with Geophysics: Estimating A-Priori Facies Probabilities for Seismic Amplitudes Inversion. Part 1, Theory

Abstract

An approach to quantitative incorporation of geological knowledge into the seismic inversion process is presented. Information about depositional environments and geological evolution of the sedimentary basin along with well logs interpretation and petro-elastic modeling data are used not only for background model building and inversion regularization, but also for inversion results interpretation and reservoir properties prediction. The method is based on the parameterization of the geological model using so-called “generalized geological variables” or G-Factors. These variables provide a quantitative description of the range of observed or expected facies. Topological and metric properties of the model are defined by a set of reference sedimentary environments and estimates of facies transitions probabilities. The method aims at solving a well-known problem of a-priori facies probabilities estimation required within the Bayesian framework. It can be applied in workflows involving either deterministic or stochastic inversion algorithms.