PMF series for availability and reliability probabilistic assessments

PMF series (a contiguous family of probability mass functions) is a new method for probabilistic assessment of availability and reliability of both repairable and non- repairable systems. The same general methodology is applied to both availability and reliability. This allows availability assessment of complex systems with both repairable and non- repairable subsystems. Binominal and Weibull can be used in narrow circumstances, and in those cases results match those of PMF series, thus validating the more general PMF series. The method uses small quantities of system data that is always available for important operating systems, whereas Monte Carlo requires substantial data or assumptions. The probability tables that are developed are analogous to the capacity outage tables of the electric power industry. However, PMF series is more broadly applicable as it also applies to single plants, the more common situation in manufacturing. Probability tables are embedded in specific user application software for efficient day-to-day business and operating decisions. These may include applications such as product inventory optimization and the absolute and relative risk difference between operating strategies. This practical methodology using available data promises to greatly expand the use of probabilistic assessments. In the future, probabilistic risk may routinely be integrated into business and operating decisions. Practical methods have not heretofore been available for these decisions; therefore, decision makers are unaware that probabilistic risk can be used in their applications. Educating the decision makers in this technology is necessary. The ability to measure and analyze risk that previously was unmeasured and unrecognized leads to surprise. Management and Reliability Engineers are generally unaware of the actual risk in meeting the production goals for their plants. When the risks are recognized, they are invariably larger than supposed. As a consequence, the reliability performance gap in any manufacturing company is perhaps significantly larger than single-valued measurement of mean availability, reliability and capacity would suggest.