Techniques to estimate the size of the undiscovered hydrocarbon resource base and produce discovery rate forecasts are reviewed and assessed. Accurate assessments of the resource base are needed by exploration firms and the government in comparing the potential of competing areas of exploration and by policy makers in developing appropriate policies to address possible hydrocarbon shortages. The techniques reviewed include judgemental prediction, extrapolative methods, discovery process models,

more »

... econometric models, and probabilistic models. A cn'tical distinction exists between econometric and geologically-based approaches. The geological approaches aim at explicitly determining the size of the undiscovered resource base and its frequency-size distribution, producing discovery rate forecasts as a by-product of the adopted methodology. The econometric approaches aim explicitly at forecasting future discovery rates, producing estimates of the resource base as byproduct. eet article examine et evalue les techniques qui servent a estimer Ie volume des ressources non decouvertes d'hydrocarbures et afaire des previsions quant aux taux de decouverte. Des evaluations precises du volume de la ressotlrce sont utiles aux entreprises exploratrices et nux gouvernements pour comparer Ie potentiel de diffhents sites d'exploration, ainsi qr/aux deddeurs pour preparer les nlesures a prendre en cas d'eventuelles pemln'es d'hydrocarbures, Les techniques examinees conlprennent la prediction au jugt, les methodes d t extrapolation, les modeles economitriques et Ies 1rwdeles de decouverte, Ies modeles econometriques et les modeles probabilistes. Une nette distinction est faite entre les methodes econometriques et geologiques. Les methodes geologiques, telies que Ies modeles probabilistes et ceux de decouverte, visent expressement adeterminer Ia taille de Ia ressource non decouverte et sa distribution frequence-taille, d'orl sont derivees des previsions de taux de decouverte. Quant aux methodes fuhas de decouverte, les estinzations du volume de la ressource etant un resultant de la methode adoptee. M. Power and J.D. Fuller are in the Department of Management Sciences at the University of vVaterloo. The authors wish to acknowledge the contributions of both the referees, whose suggestions led to substantial improvements in this paper, and the Editor, for his encouragement.