Quantification of fluid phase probability in prospect evaluation using a Bayesian approach

Abstract

This paper discusses a method to quantify fluid phase probabilities by integrating rock property trend models with seismic attribute maps. Rock property trend models for the reservoir sandstone and surrounding shales, with quantified uncertainties were used to model the near and far angle seismic response. Forward modelling of many realistic reservoir models was done in close cooperation with the petrophysicist and geologist. Brine, oil and gas filled reservoir sandstones were modelled. A statistical method was used to calculate fluid phase probability maps from calibrated fluid factor or near/far angle attribute maps. These probability maps may be used to rank undrilled prospects, and as a part of risking the fluid phase for each prospect in the area under investigation.