Optimal maintenance—replacement policy under imperfect maintenance

Abstract During a given planning period, in order to have the system working at or below a fixed failure rate, it may be necessary to replace the system to minimize the expected total cost. Preventive maintenance (PM) or replacement should be performed whenever the system reaches the maximum failure rate. This paper formulates a cost model and a branching algorithm to determine the number of PM interventions to be performed between each replacement and the number of such replacements during the planning period to minimize the expected total cost. Adjustment of the failure rate for the system degradation after each PM intervention and inflationary trends in cost factors are also incorporated. Numerical illustration and computational times have been presented.