Multi-objective deterministic and robust models for selecting optimal pipe materials in water distribution system planning under cost, health, and environmental perspectives.

Abstract By the urban growth in the societies, water distribution systems have been particularly noteworthy. Regional water authorities face a wide range of decisions related to development and replacement of water distribution systems. In this context, decisions about pipe expansion, replacement, and material selection are some of the most important decisions because they have long-term effects and it is costly and difficult to change them. In this paper, a multi-objective mathematical model is constructed based on the needs of a regional water authority and is verified by its experts. The proposed model considers environmental and health considerations in addition to costs of purchasing and installing pipes. In order to cope with uncertainties in the parameters of the model, several robust counterparts have been proposed that takes into account various types of uncertainties. The proposed models have been implemented in a case study to show their applicability and usefulness in practice. Finally, a comparison is carried out between the corresponding results. One of the proposed robust models is suggested which considers all possible uncertainties and also holds the second position in providing reasonable objective values. This research helps the decision makers in pipeline expansion and replacement projects by proposing appropriate pipe materials regarding cost, health, and environmental aspects of candidate materials even under different uncertainties. Highlights • A new optimization model is proposed for pipeline expansion & replacement decisions. • The proposed model determines the optimal material for each location of pipeline network. • Besides piping costs, environment aspects and health-related costs are also considered. • Several robust optimization counterparts are proposed for coping with possible uncertainties. • The applicability and usefulness of all proposed models is tested in a case study. [ABSTRACT FROM AUTHOR]