Methodology for Ranking the Elements of Water-Supply Networks

This paper presents a scenario-based methodology for the ranking of the elements of a water-supply network according to their value to the network's owner. The failures of the elements due to random causes and to malevolent acts are considered. The methodology is based on multiattribute utility theory and a graph theory-based network analysis algorithm. This methodology extends approaches proposed in the literature by taking into consideration the capacity of the infrastructure's elements and their mean time to repair. The water-supply infrastructure of a midsize city serves as a case study. We model the infrastructure system as a network and create scenarios to evaluate the consequences of the failure of each of its elements. For each scenario, we evaluate the supply level to the various users considering the capacity of their connection to the available resources. We then evaluate the disutility of this supply level to the decision makers using multiattribute utility theory. The accident scenarios are ranked according to their expected disutility. The vulnerabilities to malevolent acts are ranked using a subjective combination of the disutilities and the scenario susceptibility to attack. The results are provided to the decision makers for evaluation and risk management. The limitations of the proposed methodology and its applicability to other infrastructures are discussed.