Opportunity-based block replacement (Q1176825)

Preventive maintenance is widely accepted within industry as an effective means to reduce the number of failures. Preferably, it is planned at those moments in time when units are not required for production. In the process industry this may cause problems as most units are continuously used and downtime costs are high. Sometimes, there may be shortlasting interruptions of production for a variety of reasons. During these interruptions some other units are not required and these can be maintained preventively without costs for downtime being incurred, in which case one speaks of maintenance opportunities. Unfortunately, these opportunities cannot be predicted in advance. In this paper the authors give theoretical as well as computational results for general continuously distributed lifetimes and times between opportunities. In the theoretical part they focus on establishing an optimality equation without restrictive assumptions and provide an interpretation. In the computational part they present for a special class of opportunity distributions (including the exponential) an exact optimization algorithm and show how this method can be used as an approximative method for general distributions. Simulation studies are carried out to check the approximations.