Life-cycle maintenance of deteriorating structures by multi-objective optimization involving reliability, risk, availability, hazard and cost

In recent years, several probabilistic methods for assessing the performance of structural systems have been proposed. These methods take into account uncertainties associated with material properties, structural deterioration, and increasing loads over time, among others. When aging phenomena have significant effects on the life-cycle performance of the structure, it becomes essential to perform actions to maintain or improve structural safety, in agreement with the system requirements and available funds. Various optimization methods and performance indicators have been proposed for the determination of optimal maintenance plans for simple and complex systems. The aim of this paper is twofold: (a) to assess and compare advantages and drawbacks of four different performance indicators related to multi objective optimization of maintenance schedules of deteriorating structures, and (b) to assess the cost-efficiency of the associated optimal solutions. Two annual performance indicators, annual reliability index and annual risk, and two lifetime performance indicators (i.e. availability and hazard functions) are used in conjunction with total maintenance cost for evaluating Pareto fronts associated with optimal maintenance schedules of deteriorating structures. Essential maintenance actions are considered and optimization is performed by using genetic algorithms. The approach is illustrated on an existing deteriorating bridge superstructure.