Optimal Repair Policies for Infrastructure Systems with Life Cycle Cost Minimization and Annual Cost Leveling

Abstract

One of the goals of infrastructure asset management research is to find appropriate repair policies for infrastructure systems. The annual repair cost of the system may vary when a repair policy meant only to minimize life cycle cost is applied to each facility in the system. Such variance in the annual repair cost leads to difficulty in securing the budget for system managers; thus, a desirable, practicable repair policy should reduce not only life cycle cost but also the dispersion of annual repair cost. In this study, we first formulate a network-level repair policy optimization problem for an infrastructure system comprising a finite number of facilities to minimize the total cost over the planning period, which is defined as the weighted sum of the repair cost and its variance. Then we propose two methods for solving it: (1) the exact one based on the Markov decision process; and (2) an approximate one using preventive repair rules. The former can be used in small-scale systems, whereas the latter can be used to simplify the repair policy regardless of the size of the system and to determine an approximate repair policy for large-scale systems. The proposed methodology is applied to two numerical investigations of (1) a small-scale infrastructure system; and (2) a large-scale infrastructure system. In the first case, we find the Pareto frontier of the repair cost and the variance in annual repair costs by the exact solution method and show that the preventive repair rule–based method provides a near-optimal solution. In the other case, the preventive repair rule–based method leads to a superior policy on aggregation of the optimal solutions independently found for each decomposed subsystem.