Has the Largest Field Been Discovered Yet? PETRIMES and GRASP 25 Years Later

Abstract

Assessment of undiscovered oil and gas resources has been an important component of energy policy for the governments of the United States and Canada for many years. A pool-size-by-rank statistical procedure is a centerpiece of the Geological Survey of Canada’s Petroleum Exploration and Resource Evaluation System (PETRIMES) and of the U.S. Department of Interior’s Geological Resource Assessment Program (GRASP). Both employ discovery process modeling to make inferences about the number of pools in a play and about parameters of the play’s pool size distribution. The pool-size-by-rank procedure implemented in these two systems abandons a key primitive postulate on which modern discovery process models are based—sampling proportional to pool size and without replacement. This logical disjunction has consequences: the predictive distribution of number of pools remaining to be discovered and the predictive distribution of undiscovered pool sizes generated by use of pool-size-by-rank procedures differ substantially in shape, location and spread from predictive distributions that incorporate sampling proportional to size. Uncertainty about total undiscovered oil and gas in a play is diminished.