Common cause, or common mode failures are a major concern in analyzing system reliability. A careful and elaborate treatment of common mode failures has been studied. Up to date, Binomial Failure Rate model has been used as a tool for common mode failures. This model treats binary type, that is, success or failure in describing component``s events. The simply on-off typed methodology can not analyze complex phenomena such as partial failures, incipient failures, potential failures, command faults, etc., which really exist.
Trinomial Failure Rate model has been developed to analyze such ambiguous events.
Maximum likelihood estimators method is adopted to unfold this model. The results of Trinomial Failure Rate model are compared with those of Binomial Failure Rate model using assumed data for gray events.
By identifying the ambiguous situations and subdividing the events which are used as an input data, uncertainty which may come from the analyst``s hesitation in judging complex events, can be reduced, and really existent phenomena can be revealed quantitatively.