A framework for comprehensive reliability assessment has been developed based on Markov
models of system components and identification methods of events that contribute to
unreliability. Specifically, an improved wind-chime approach is described coupled with an
improved power system model. The model is based on the single phase quadratic modeling
approach that provides superior performance in two aspects: (a) faster convergence, (b) ability to
model complex load characteristics, and classes of loads such as interruptible load, critical load,
etc. The same model has been extended for identifying events that contribute to the unreliability
of the system. The method is integrated in the wind-chime scheme for the quick identification of
critical events and effects analysis of the critical events. Since the methods are based on the
Markov state space approach, probability, frequency and duration indices are computed, such as:
(a) probability of customer interruption, (b) frequency of customer interruption and (c) duration
of customer interruption. The advanced load modeling capability enables: (a) a realistic
evaluation of industry practices such as load management programs on system reliability, and (b)
a realistic evaluation of load characteristic on voltage problems and their impact on reliability.
Examples illustrating the capabilities of the approach are provided.