Sensitivity Analysis of Maintained Systems Using a Population Model: A Case Study

A probabilistic methodology is discussed for the evaluation of maintenance requirements and expected costs for large‐scale logistics systems. A multitude of identical system components that are used to support operations characterise large‐scale systems. These identical components can be grouped together, and “representative” maintenance characteristics used in lieu of individual component tracking to perform a strategic analysis. The combination of these component “groupings” forms a population. The benchmark results are presented from a previous study of a population of US Navy tugs. Benchmark parameters are then varied in ten different maintenance scenarios to demonstrate the use of population models to analyse strategically the trade‐off between expected population costs and system availability.