Quantitative Risk Assessment Techniques Based on Uncertainty Theory for Natural Gas Distribution Station

Pipeline stations, as an important part of long-distance pipeline systems, include lots of facilities which are highly concentrated and always operate continuously. Risk assessment is an important foundation work for the risk management of these stations. Since various uncertainties exist during the quantitative risk assessment (QRA), this paper explores the theories and approaches of QRA for station accidents, and also introduces some specific mathematical theories for quantification and dealing with uncertainties. This paper combines uncertainty theory effectively with the QRA for gas distribution stations, analyzes the uncertain factors in the QRA of gas distribution station, and establishes Bayesian update model for estimating basic events’ failure rates and probabilities of failure on demand based on generic failure data and plant-specific data. And it also offers conversion method among conjugate prior distribution of different types. Besides, probabilistic estimation model is set up by the combination of fuzzy set theory, expert judgments and fuzzy group decision making. The paper builds Fuzzy Bow-Tie quantitative model for distribution station under dependency relationships, and proposes the sensitivity analysis method for the accident model based on fuzzy importance index, fuzzy uncertainty index and minimal cut sets importance index.