An Algorithm for Preventive Maintenance Policy

Abstract

Simple preventive maintenance (Maintenance type 1P) and preventive replacement (maintenance type 2P) are scheduled in such a way that the system does not drop below a minimum reliability. Failure rate after maintenance type 1P lies between "good as new" and "bad as old". The degree of improvement in failure rate after maintenance type 1P is cailed the improvement factor. A set of curves for the improvement factor as a function of cost for maintenance type 1P and age of the system is proposed. The cost rate for a system is formulated as a ratio of an average cost for a cycle (time between replacements) to an average cycle length. An optimum number of type 1P maintenance actions before type 2P maintenance is obtained by minimizing the cost rate when the failure times are Weibull distributed. The optimum solutions are a function of improvement factors and predetermined upper limit of failure rate.