Summary
This paper proposes two seismic characterization workflows based on the results obtained after a supervised seismic facies analysis. Two methodologies have been designed in the frame of an integrated reservoir study. They are based on a real data set and pre-stack inversion results (P and S impedance volumes), and a comprehensive well data analysis. These workflows exploit the two outputs of the supervised facies analysis, namely the facies attribution and the probabilities of good assignment. One workflow uses the geographical distribution of the facies areas, with a strong hypothesis of continuity, while the other is based on the good assignment probabilities. Both methods aim at providing a constraint for a geomodel with a spatial continuity consistent with the geology. These two approaches have successfully been controlled and compared. Finally, the results were integrated in the geological model infilling.
