A Bayesian Procedure for Resource Evaluation of Petroleum Provinces in the Early Stages of Exploration

Medova, E. (1981). A Bayesian Procedure for Resource Evaluation of Petroleum Provinces in the Early Stages of Exploration. IIASA Working Paper. IIASA, Laxenburg, Austria: WP-81-066

Preview

Text WP-81-066.pdf Download (1MB) | Preview

Abstract

A major open problem in quantitative methods for petroleum resource evaluation concerns the provision of statistical techniques for geological provinces in the early stages of exploration-for example, if all exploratory wells drilled to date have been found to be dry or if only a few fields have been discovered. A considerable body of literature exists concerning statistical methods for mature provinces in which discovery volumes are on a general declining trend; all these methods use exploration history within the province to project total petroleum resources. In the early stages of exploration however, historical data is an insufficient basis for resource estimation and a simple method must be found to transfer relevant information from explored provinces with similar geology.

Following a survey of the existing quantitative methods for petroleum resource evaluation, this paper describes a new method for Bayesian updating of discovery probabilities corresponding to a rough field size classification in the light of current dry hole data. It is based on spatial Poisson dry hole sampling distributions fitted to geologically similar provinces. The method is applied to and the underlying assumptions statistically tested on some typical partly-explored provinces in Brazil. Finally, a Monte Carlo method for the resource assessment of immature provinces, based on revised discovery probabilities and building on earlier work (Medova, 1980) will be suggested for future development.

Actions (login required)

View Item