Probabilistic availability risk assessments without simulation

Two methods for calculating availability and capacity probability distributions are discussed. One gives exact results, but has narrow application. The other is approximately exact but is a general method applicable to time varying data. Each method uses high level system data that is both available and accurate. The two methods can be implemented using personal computers and spreadsheets. Templates with cell equations are provided so the user can build their own working spreadsheets. The first method presented is capacity outage probability tables (COPT) used routinely in the electric power industry. It calculates exact empirical instantaneous availability and capacity distributions for limited but practical applications. The second method is probability mass function (PMF) series, a general method for measuring accumulated system performance with a dense family of empirical distributions. COPT is mathematically exact and PMF series is essentially so. PMF series is computer intensive with accuracy increasing with increasing computations. Excellent results are achievable on personal computers. In an example, COPT and PMF series results are shown to be comparable in the special case in which these totally different concepts can be compared. Two non-conventional applications will be discussed as examples. One of these is personal injuries recorded by an industrial maintenance contractor. The probability and risk associated with any performance-based contract safety goal can be determined. If risk (measured in units of recorded injuries) is high, it can easily be identified and managed down by changing the \\“system\\” that generates injuries. Alternatively, new contract commitments can be negotiated, again with the new risk calculated. The other nonconventional application is a business system that delivers a medical services product. The risk associated with management´s goals for the next month or the next quarter can be assessed. Risk-based goals can b- > - > e set, and once goals are established, unacceptable risk can be managed by changing the business system. Computer simulation models or other traditional methods are not now being routinely used for these non-conventional applications. Likewise there are numerous risk assessment applications residing within engineered physical systems, such as manufacturing plants, that are underserved. Also, traditional applications, such as capacity assessment of manufacturing plants, can benefit due to no modeling requirement and the availability of high quality data.