Probabilistic common-cause failures analysis

Abstract

Common-cause failures (CCF) are simultaneous failures of multiple components within a system due to a common-cause or a shared root cause. CCF can contribute significantly to the overall system unreliability. Therefore, it is important to incorporate CCF into the system reliability analysis. Traditional CCF analyses have assumed that the occurrence of a common-cause results in the deterministic/guaranteed failure of components affected by that common-cause. In practical systems, however, the occurrence of a common-cause may result into failures of different components with different probabilities of occurrence. This behavior is termed as probabilistic CCF (PCCF). In this paper, we present a combinatorial method for the reliability analysis of systems subject to PCCF. The approach is represented in a dynamic fault tree model by a proposed probabilistic CCF gate. Basics of the proposed approach and effects of PCCF on the system reliability are illustrated through the detailed quantitative analysis of an example system.