This paper describes a novel methodology to integrate well logs and 3D pre-stack seismic data. The objective is to assess lateral continuity and spatial extent of lithology and fluid units penetrated by a well. Well logs exhibit a radial length of investigation shorter than 3 m and hence provide limited indication of lateral extent and continuity of reservoir flow units. In the past, post-stack seismic data have been used to fill the spatial gap between sparse well locations. However, post-stack seismic data respond to acoustic impedance (the product of bulk density and P-wave velocity and, therefore, cannot always uniquely discriminate between spatial variations of porosity, thickness, shale concentration, and fluid saturation. Pre-stack seismic data, on the other hand, are sensitive to S-wave velocity and bulk density in addition to P-wave velocity. This provides additional degrees of freedom to uniquely interpret lateral variations of seismic amplitude in terms of variations of petrophysical properties and flow-unit thickness.
We have developed a novel approach to integrate petrophysical well logs such as litho-facies, shale concentration, porosity, and fluid saturation, with 3D pre-stack seismic data. This approach is based on a stochastic global inversion method that concomitantly honors the well logs and multiple angle stacks of seismic amplitude data. Inversion results consist of 3D spatial distributions of litho-facies and petrophysical parameters between wells that exhibit a vertical resolution intermediate between that of well logs and 3D seismic data.
Examples of the application of this technique are shown using high-quality data acquired in the deepwater Gulf of Mexico. Reservoir units consist of amalgamated turbidite sands. The lateral continuity of flow units is conditioned by extensive faulting. Conventional petrophysical interpretation based on well logs and rock-core data was performed for 7 wells. Petrophysical and litho-facies logs we constructed and correlated with elastic parameters inferred from P- and S-wave sonic logs to assess the sensitivity of elastic parameters to variations in porosity, shale concentration, and fluid saturation. Both petrophysical logs and elastic-petrophysical correlation cross-plots, together with four angle stacks of pre-stack seismic amplitude data, were entered to the stochastic inversion algorithm to produce 3D distributions of litho-facies, porosity, and fluid saturation, as well as their corresponding uncertainty. Results successfully describe the spatial continuity of sand units and of their porosity and saturating fluids away from wells. These are the first results ever published on the successful quantitative integration of well logs and pre-stack seismic data.