A Decomposition Method to Analysis Complex Fault Trees

This paper describes a new method for decomposing a coherent fault tree F(X) of n variables into a set of simpler functions. Each simple function is analyzed independently to determine the minimal cut sets (MCS) and the probabilistic parameters of interest, i.e. top event failure probability and importance measures of basic events. These results are suitably re-combined to obtain the results at F(X) level such that i) the MCS of F(X) is the union of the MCS of simple functions and ii) the probabilistic parameters at F(X) level are obtained by simply combining the results of the probabilistic analysis of all simple functions. It is shown that in applying the cut-off techniques for determining the most important MCS, to a decomposed fault tree, the truncation error can be correctly determined. An example is described to show the potential of the proposed method.

CONTINI Sergio; 

2009-01-05

Taylor & Francis

JRC44425

http://www.esrel2008.com,    https://publications.jrc.ec.europa.eu/repository/handle/JRC44425,