Abstract:
This work focuses on the integration of Seismic and petrophysical data and insights drawn out of them for reservoir characterization. The field of study is Bonanza, onshore field, lying in the northern Barmer basin. The field is characterized as low net-to-gross reservoir having very low fluid sensitivity. In that case, lithofacies classification is the key scheme to identify pay zones. Rock physics analysis using rock physics template and well logs have been used to provide preliminary knowledge about petrophysical properties of field. Sand and Shale are the two main types of facies which can be distinguished from each other using efficient methods.
To classify litho-facies in the entire seismic volume, classification by regression has been adopted: one deterministic approach. Another novel approach is by probabilistic analysis. For that, facies dependent statistical probability density functions (PDFs) have been built using well data as training data. Further, Bayesian classification has been used to make sand probability volume. The uncertainty in the output has also been quantified using a parameter called Information Entropy. The entire process of probabilistic analysis has been accomplished via R programming as it is very competent in statistical computations. Ultimately, lithofacies classification has been done for lower region of producing formation.
