Component replacement and reordering policies for spare parts experiencing two-phase on-shelf deterioration

Spare parts provisioning strategies for critical components are essential for maintenance management in a wide range of industrial applications. In practice, some types of spare parts may suffer from on-shelf deterioration, which will not only affect the reliability and availability of the parts themselves but also the overall operational costs. In this context, a natural question that arises is: “Should we use new parts or degraded ones first for component replacement?” Indeed, developing mathematical models to answer this question has both theoretical and practical values. This article studies component replacement and spare parts reordering policies for a system with spare parts experiencing on-shelf deterioration. A two-phase continuous-time Markov chain is utilized to model the on-shelf deterioration process for each spare part, and two different part consumption strategies, i.e., Degraded-First (DF) and New-First (NF) strategies, are formulated. An optimization framework and a solution algorithm are developed for the two strategies to determine the optimal order intervals and order quantities of spare parts for a fixed planning horizon. The monetary performance measures of the two strategies are studied and compared to a random selection (RS) alternative. Numerical examples show that when the overall system-wise replacement demand rate is approximately independent of the part consumption strategies, the DF strategy leads to the biggest savings compared to the RS strategy while the NF strategy results in the highest expected cost among the three alternatives.