A Life Cycle Oriented Multi-objective Optimal Maintenance of Water Distribution: Model and Application

Aiming at trading off several conflicting criteria in practical maintenance in a deteriorating water distribution network, a life cycle oriented multi-objective optimization model of water distribution network maintenance is developed, which is composed of seven interrelated sub-models with different functions. This model can provide decision support for preventive maintenance decision, including identifying the pipeline that needs to be maintained, judging the time point for maintenance, determining the type of maintenance technology, calculating the economic cost of maintenance, and presenting the impact under different maintenance strategies. Based on the life cycle of each pipeline, multiple effects in the water distribution can be dynamically evaluated, such as pipeline age, failure rate, hydraulic reliability health level etc. Based on special design of chromosome gene encoding, the algorithm of elitist Non-dominated Sorting Genetic Algorithm (NSGA-II) is incorporated to achieve multi-objective optimization solution effectively. With application of a county in Zhejiang province in China, three strategies including empirical decision single-objective optimization decision and multi-objective optimization decision are evaluated and compared to the baseline systematically. Although the annual maintenance cost of strategy III is not the lowest among those three strategies, the pipeline age, failure rate, hydraulic reliability, and health level of the water distribution network under the strategy are at the best level. With multiple objectives considered simultaneously, the results of strategy III are recommended as the optimal maintenance implementation arrangements. This model can promote to find an optimal maintenance strategy, and provide a technical support for the planning, design and implementation of maintenance arrangements of water distribution network in a long-term period.