Pigott, John D.¹, David L. Harn¹ 
(1) University of Oklahoma, Norman, OK

ABSTRACT: Hindcasting Offshore Gulf of Mexico Reserves through Monte Carlo Continuous, Monte Carlo Discrete, and Fuzzy Arithmetic Risk Analysis

Risk analysis for future hydrocarbon reserves remains inductive until the last reserve is produced. Therefore, in order to examine the accuracy and efficacy of risk forecasting techniques, unrisked hindcasting was employed for historic data from two offshore Gulf of Mexico fields: Tiger Shoal and Starfak. Data for two fault blocks from Tiger Shoal and three fault blocks from Starfak were analyzed using three distinct methods for reserve determinations with increasing knowledge of the population probability parameters: Fuzzy Arithmetic (minimal knowledge and estimation of the population possibility parameters), Monte Carlo continuous (some knowledge), and Monte Carlo discrete (more knowledge). 
Comparisons of these simulations with the known reserves produced found: 1) Significant differences between predictions and reality made initially by the operator without recourse to these three methods; 2) Differences in some cases imply by-passed reserves; 3) Minimization of differences (read: increased accuracy and decreased cost) with systematic implementation of the three methods with increasing reservoir information content: Fuzzy Arithmetic for rank wildcat exploration, Monte Carlo continuous for exploration, and Monte Carlo discrete for development.

AAPG Search and Discovery Article #90026©2004 AAPG Annual Meeting, Dallas, Texas, April 18-21, 2004.