APPLICATION OF BAYES' MODEL TO OIL/GAS PROCESSING NETWORKS FAILURES

摘要

This paper seeks to present the uncertain events that occur during the assessment and management of corrosion data survey by using Bayes' Model. In decision making, uncertainty about the outcomes of various situations is an important element of the analysis of the alternative strategies from which the choice must be made. The two main interpretations of probability in this paper are objective and subjective probabilities. Bayesian conditional probability theory is the approach adopted in the assessment of subjective probability. The Bayesian (or subjectivist) interpretation views the probability of an event as a subjective degree of belief, which taken into account all relevant knowledge and information. The cornerstone of Bayesian probability theory is the inversion formula due to Bayes. This paper therefore models pipeline oil/gas failures in line with Bayesian methods of diagnostic inference and evidence patterns along six broad categories of oil/gas failures: stress rupture, lack of quality control, environment, sand erosion, corrosion, and natural cause. The data generated in this study are from survey, which runs from Makaraba, Utonana, Abiteye, Dibi, Olero Creek, Opuekeba, and terminating at the Escravos Tank Farm of the Chevron western swamp facilities.