Topological analysis of water distribution networks for optimal leak localization

Conference Article


International Conference on Hydroinformatics (HIC)





Doc link


Download the digital copy of the doc pdf document


This paper introduces two methodologies to provide an optimum sensor deployment layout, one based on a model-based approach and the other entirely data-driven. The first method is formulated as an integer optimization problem, an optimization criterion consisting of minimizing the average topological distance. The second method is a new methodology to provide an optimum sensor placement regarding how many sensors to install without using hydraulic information but just exploiting the knowledge of the topology of the Water Distribution Networks. The method uses the Girvan-Newman clustering algorithm to ensure complete coverage of the network and the study of the installation of pressure sensors in the central nodes of each group, selected according to different metrics of topological centrality. The approach is illustrated in the Modena network.


artificial intelligence.

Author keywords

data-driven, leak localization, sensor placement, water distribution network

Scientific reference

D. Alves, J. Blesa, E. Duviella and L. Rajaoarisoa. Topological analysis of water distribution networks for optimal leak localization, 2022 International Conference on Hydroinformatics, 2022, Bucharest, Romania, Vol 1136 of IOP Conference Series: Earth and Environmental Science, pp. 012043.