Porosity Volume Attribute by Bayesian Estimation

We propose a new volume attribute (residual porosity), consisting of a porosity perturbation generated from different saturation assumptions in a Bayesian porosity estimation framework. This idea explores the strong coupling existing between porosity and saturation. Using Bayesian inference and Gassmann model, porosity may be estimated using two different conditions for the reservoir water saturation: initial water saturation (e.g., Swi = 15 %) and fully water saturated (Sw = 100 %). The residual porosity volume attribute is further obtained by subtracting the two porosity volumes. This attribute should highlight the oil region of the reservoir, once the parameters for the Gassmann model are calibrated for oil reservoir condition from well-log data. We present an application to real data sets from Campos Basin, Brazil. A comparison of seismic volume inversion with well-log data inversion shows the usefulness of this approach for highlighting oil-saturated intervals.