Uncertainty in fault tree analysis: a fuzzy approach

Abstract In fault tree analysis, the uncertainties in the failure probability and/or failure rate of system components or basic events can be propagated to find the uncertainty in the overall system failure probability. The conventional approach is Monte-Carlo simulation by assuming a probability distribution for the failure probability. In addition, a new methodology based on fuzzy set theory is also being used in the fault tree analysis for quantifying the basic event uncertainty and for propagating it. However, identification of the components which contribute maximum to the system failure probability is also important in fault tree analysis. Similarly, ranking the components based on their contribution of uncertainty to the uncertainty of the system failure probability is also very important. This paper presents a comparative study of probabilistic and fuzzy methodologies for top event uncertainty evaluation. Further, it explains a new approach to rank the system components or basic events depending on (1) their contribution to the top event failure probability and (2) their uncertainty contribution to the uncertainty of the top event based on fuzzy set theory.