Abstract

Geo-statistical inversion technique can be applied to quantitatively relate well and seismic data, analyze quantitatively the resulting map and estimate the probability of success directly from the available data. The aim of the present work is to delineate the extent of discrete reservoir facies of Panna formation in B157-127 area of Western offshore Basin, India. Panna is mainly dominated with shale sequences with intercalations of sand and coal facies. Reservoir characterization of hydrocarbon bearing sands is a challenging task in this area as sands are not having similar elastic properties in all the drilled wells. To tackle the above issue a geostatistical pre-stack inversion methodology was adopted. Rock physics modelling were carried out in a few wells and in remaining wells, wherever required, S-sonic as well as P-sonic logs were predicted using machine learning. The pre-stack geostatistical inversion outputs helped in mapping of sand dispersal pattern in Panna formation with a good confidence and matching with the well observations.