Quantitative Chance Assessment of Geological Risk Factors in Oil and Gas Exploration

Geological risk evaluation for oil and gas prospects depends on individual geological factors, such as reservoir beds, traps, seals, source beds, and so forth. Despite the simplicity and rapidity of Delphi guessing of chance quantification, its major disadvantage is the lack of direct influence of frequency data, once the estimation of probabilities of the factors enters into decision systems as poorly constrained assumptions. The goal here is to outline some procedures that provide chance measurement derived from statistical inferences. The chance can be interpreted from geological and geophysical interpretation, basin modeling, analysis of probability of success and uncertainty. The logical link between exploration-oriented databases and statistical tools is one of the most critical aspects in defining chance during risk analysis. In order to establish a systematic to quantify the chances to be further used during economic evaluation, we propose organize data to derive statistical weights for each risk factor. The relevance of the statistical procedures comes from the needs of the intensive use of scientific knowledge, in order to represent numerically the subjective thinking to model the complexity and diversity of petroleum accumulations. It is assumed that probability updating based on post-mortem is a technique that makes it possible to incorporate uncertainties in the chance evaluation. Statistical inventory of petroleum system risk factors are considered in order to quantify chance. The probability distributions are derived from logical operations that allow a spatial correlation between the existence/absence of petroleum system elements and producing/dry wells. In the suggested method, to each risk factor is applied a probability that represents the multiplication of its necessity and sufficiency values. The necessity of a risk factor is proportional to its absence in a dry hole, and can be measured by the conditional probability P(nV|nA), where nA represents non-accumulation in a particular drilled site. Its sufficiency, in turn, is proportional to its presence in producing well contexts, measured by P(V|A). Ideal variables would show high values for both probabilities, i.e., its presence accounts for petroleum accumulation over a given area, whereas its absence in turn accounts for non-accumulation over the same area. This work presents the results of the application of this methodology in a hydrocarbon producing basin from Brazil.

AAPG Search and Discover Article #90100©2009 AAPG International Conference and Exhibition 15-18 November 2009, Rio de Janeiro, Brazil