Risk importance measures in the dynamic flowgraph methodology

This paper presents new risk importance measures applicable to a dynamic reliability analysis approach with multi-state components. Dynamic reliability analysis methods are needed because traditional methods, such as fault tree analysis, can describe system's dynamical behaviour only in limited manner. Dynamic flowgraph methodology (DFM) is an approach used for analysing systems with time dependencies and feedback loops. The aim of DFM is to identify root causes of a top event, usually representing the system's failure. Components of DFM models are analysed at discrete time points and they can have multiple states. Traditional risk importance measures developed for static and binary logic are not applicable to DFM as such. Some importance measures have previously been developed for DFM but their ability to describe how components contribute to the top event is fairly limited. The paper formulates dynamic risk importance measures that measure the importances of states of components and take the time-aspect of DFM into account in a logical way that supports the interpretation of results. Dynamic risk importance measures are developed as generalisations of the Fussell-Vesely importance and the risk increase factor.