Leakage in water distribution network (WDN) leads to a large amount of water loss and water pipe pollution and affects residents’ domestic water supply. Therefore, network leakage detection significantly saves water resources. The traditional model approach has ample search space in solving large distribution network applications, a challenging and complex leakage detection process and a low detection accuracy. For the above problems, this study proposes a new method of leakage area detection based on gray wolf optimization (GWO). First, the extensive WDN is divided into several virtual areas. Then, the leakage is simulated by the additional water demand of nodes, and the node demand of the distribution network is calibrated based on the GWO algorithm. Finally, the leakage area is identified, and the size of the leakage in that area is estimated. The method was experimented on in two cases, simulating single-point leakage and multi-point simultaneous leakage, respectively. The results show that the method estimates the size of leakage in the corresponding area based on accurate identification of leakage areas, and the detection error of leakage is within 17.14%. The method provides water workers with guidance on leak detection, significantly reducing staff time to repair pipes.

  • Use a small number of monitoring sensors to detect the leakage status of pipeline networks, identify leakage areas and estimate the leakage amount.

  • A multi-leakage sun detection method based on gray wolf optimization is proposed, which can effectively simulate the actual pipe network leakage situation.

  • A method of dividing multiple virtual areas for leak detection is proposed, which improves the accuracy of leak detection.

Graphical Abstract

Graphical Abstract
Graphical Abstract
This content is only available as a PDF.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).