Skip to Main Content
Leakages are present to some extent in all water-distribution systems. This paper compares two model based methodologies for leakage isolation in water distribution networks. Both are based on the pressure measurements and pressure sensitivity analysis of nodes in a network. Simulations of the network in presence and absence of leakage may provide an approximation of this sensitivity. The difference between both methodologies relies on how the information of the sensitivity matrix is handled. In the first approach, this information is binarised using a threshold. The resultant binary matrix is assumed as a signature matrix for leakages. One of the main issues in the binarisation process is the threshold selection. Even with the best selection of a threshold, binarisation implies loss of information. The second method is based on the use of the sensitivity matrix without any transformation in order to avoid loss of information. Results obtained on a real network (a District Metered Area (DMA) of Barcelona water distribution network) using both methods are compared. Finally, some discussions and conclusions about the limitations of both techniques and future work are presented.