Simplified Reliability Calculations for Complicated Systems

This paper discusses methods for calculating reliability and availability of systems. The example used assumes the exponential distribution, but the method of Section III is even more useful in other cases. Well-known general methods are briefly reviewed in Section II. A little-used but sometimes highly practical method is presented in Sections III and IV for use when the time period involved is (relatively) short. It involves a systematic investigation of the various possible patterns of failures and repairs, beginning with the simplest patterns that can cause system failure and progressing only as far as necessary to attain the accuracy required, calculating the contributions each makes to unreliability and the possible contributions of all the not-yet considered patterns.