Stochastic ageing and maintenance models for unavailability quantification of complex multi-component systems

Main objective of this paper is research and development of the previously developed original methodology so that to be used for unavailability calculation of a complex maintained system with maintained (both preventively and correctively) and ageing input components with optional distributions of the time to failure. Original computing methodology results from the stochastic alternating renewal process. Models I-III (for non-repairable component, repairable with both apparent and hidden failures) are in this paper completed by new Model IV with calendar based preventive maintenance policy in which the component is renewed as soon as it reaches a prescribed deterministic age. These models represent all frequently used component models with both preventive and corrective maintenance. Models were solved both analytically and numerically. Found component unavailability functions are used to quantify unavailability of a complex maintained system, here is used real power distribution network. System is represented by the use of directed acyclic graph, which proved to be very effective system representation for computing highly reliable systems.