TABLE OF CONTENTS For complex multi-component systems, for example such as those employed in the nuclear, chemical, process and aerospace industries, it is important to analyze the possible mechanisms of failure and to perform probabilistic analyses for the expected frequency of such failures. Often, each such system is unique in the sense that there are no other identical systems (same components interconnected in the same way and operating under the same conditions) for which failure data have been collected: therefore a statistical failure analysis is not possible. Furthermore, it is not only the probabilistic aspects of failure of the system which are of interest but also the initiating causes and the combination of events which can lead to a particular failure.
The engineering way to tackle a problem of this nature, where many events interact to produce other events, is to relate these events using simple logical relationships (intersection, union, etc.) and to methodically build a logical structure which represents the system.
In this respect, Fault tree analysis is a systematic, deductive technique which allows to develop the causal relations leading to a given undesired event. It is deductive in the sense that it starts from a defined system failure event and unfolds backward its causes down to the primary (basic) independent faults. The method focuses on a single system failure mode and can provide qualitative information on how a particular event can occur and what consequences it leads to, while at the same time allowing the identification of those...
