Component Risk Trending Based on Systems Thinking Incorporating Markov and Weibull Inferences

This paper uses systems thinking to present a power utility asset management system as a system in spatial transition. Next, it integrates inferences from Markov processes, the Weibull distribution, and the bathtub curve analysis to develop a quantitative risk trending model. A risk factor (RF) is used to define a quantitative measure of risk. A set of failure data is applied to compute the maximum likelihood estimates of Weibull parameters that are fitted into the RF. MATLAB algorithms are used to simulate sensitivities of the RF to changes in the number of components renewed and the number admitted to or relieved from a high-operating-load regime during the life cycle. The model determines the impacts of renewal strategies on failure risk by trending risk profiles associated with these sensitivities. In addition, it provides modeling equations for the systems thinking approach which has, traditionally, used qualitative models. Furthermore, it is flexible since the computed parameters are unique to the set of data. These parameters thus generate equally unique plots of probability distribution functions that are required for analysis of reliability and risk. This is primarily intended to be used in risk management, but it can also be applied in performance-based compensation schemes for workers.