Reliability modeling of parallel systems under multiple common-cause failures

For most complex systems, CCF (Common Cause Failures) are their common features. It will lead to considerable errors or even misleading conclusion by neglecting failure dependence and assuming independence when evaluating the system reliability. Most of the existing CCF models focused on systems under single common cause. For systems under multiple common causes, these models are not applicable. In this paper, a new implicit modeling method is proposed to address the effect of CCFs in the reliability evaluation of static parallel systems based on failure mechanism analysis, GLSI model, conditional probability and Boolean truth table/Minimal path sets. This method enables the evaluation of multiple common causes that affect components simultaneous. An illustrative example is given in the end. The calculation result based on the proposed CCF model is lower than that of independent model, and the CCF model is verified to be more applicable by using Monte-Carlo simulation.