The marginal cost analysis and its application to repair and replacement policies

The concern of this paper is twofold: (i) elaboration on the role of the marginal cost analysis (MCA) method in the study and optimization of maintenance policies; (ii) the study of a repair and replacement policy where the decision whether to repair a failed item or replace it is based on the repair costs: those encountered now and those anticipated in the future (the policy is an extension of another one which bases this decision solely on the present repair costs). The situation of incomplete information on the underlying life distribution is also considered in which case a Bayesian inference procedure is associated with the failure process and its results are incorporated into the repair or replace decision process at failures. These two goals are interlinked in that on one hand the repair and replacement policy is analyzed and optimized by using the MCA, and on the other hand this provides yet another illustration for the application of the MCA and its adaptability to model and policy generalizations.