Abstract :
Scope and Purpose—There presently is no accurate and efficient method to determine the reliability and availability of service at various points in a large-scale public utility or electro/mechanical system. Up to now, large-scale systems have been analyzed by either approximate empirical methods or by mathematical reliability/availability analysis, which requires severely limiting (unrealistic) assumptions about component failure and repair time distributions and about component linkages. The novel simulation modeling approach presented here is very accurate for determining time-between-failure (TBF) and time-to-repair (TTR) distributions without any limiting assumptions; however, it is presently computer time intensive.
Discrete simulation modeling is shown to be of value in determining the time-between-failure (TBF) and time-to-repair (TTR) distributions at various points in series/parallel and non-series/parallel (cross-linked) systems. Most real world systems, such as power, telephone, gas, water, and electro/mechanical systems, are extremely large and have cross-linked components that are not “well-behaved” statistically, which make them impossible to analyze accurately and efficiently using mathematical reliability/availability analysis.
