Power plants are complex systems consisting of thousands of heterogeneous components. The maintenance policy of such systems includes grouping the maintenance actions of different components. This grouping of the maintenance actions together with functional interactions between the components results in system level models that are highly coupled. Accurate models of the maintenance process would allow plant managers and system designers to generate cost-effective maintenance policies. A system level model with explicit accounting of all components, interactions and maintenance policy groupings would provide the desired accuracy, but the creation of such a model is infeasible or very expensive. In this paper, a bottom-up modeling procedure that allows for the creation of the maintenance model is introduced. This approach reduces the complexity of the modeling process while still capturing all the relevant characteristics. Each component type present in the system is modeled separately, but each model includes an aggregated representation of the effects of the rest of the system on the modeled component type. This procedure applied to both a midsize-system and to a gas turbine engine. Simulations are used to illustrate how the created models are used and to verify the level of accuracy by comparing the results to system-level models.

No abstract

No abstract

No abstract