Optimization of lifetime maintenance strategies for deteriorating structures considering probabilities of violating safety, condition, and cost thresholds

Modern maintenance strategies for deteriorating structures are typically based on minimizing total cost. Extensions of this concept into the nondeterministic range allow to take into account the uncertainty in total cost. In this paper, the probability of exceeding a total cost threshold is included in the structure maintenance optimization problem. In addition, the probabilities of violating acceptable levels of safety and condition are computed. Based on this, an optimization task can be formulated and solved in which these probabilities can be included into the formulation of the objective function and the constraint conditions. Smoothing of the Monte-Carlo-based results is carried out utilizing a smoothing procedure by regression methods which are well-known for the response surface approach. Both time-based maintenance and performance-based maintenance are taken into account. The total maintenance cost is minimized with respect to parameters of both types of maintenance. Taking into account the cost of failure, optimal values for the threshold safety level and the time interval between subsequent maintenance applications are obtained. It is shown that different design parameters can lead to the same total cost thus allowing for a trade-off between time-based maintenance and performance-based maintenance at the discretion of the structure maintenance manager. (c) 2005 Elsevier Ltd. All rights reserved.