Battle of the Leakage Detection and Isolation Methods

Abstract

A key challenge in designing algorithms for leakage detection and isolation in drinking water distribution systems, is the performance evaluation and comparison between methodologies using benchmarks. For this purpose, the "Battle of the Leakage Detection and Isolation Methods" (Bat-tLeDIM) competition was organized in 2020 with the aim to objectively compare the performance of methods for the detection and localization of leakage events, relying on SCADA measurements of flow and pressure sensors installed within a virtual water distribution system. Several teams from academia and the industry submitted their solutions, using various techniques including time-series analysis, statistical methods, machine learning, mathematical programming, meta-heuristics and engineering judgment, and were evaluated using realistic economic criteria. This paper summarizes the results of the competition and conducts an analysis of the different leakage detection and isolation methods used by the teams. The competition results highlight the need for further development of methods for leakage detection and isolation, and also the need to develop additional open benchmark problems for this purpose. Vrachimis et al., May 24, 2022 network; and leakage localization, which aims to provide an approximate location of leakages given the available measurements. A recent review paper (Chan et al. 2018) classifies leakage detection methodologies into Passive and Active methods. Passive methods (also referred to as equipmentbased, hardware or external methods) require the deployment of specialized equipment, such as acoustic sensors or ground penetrating radars, at areas that are suspect of leakage. Active methods(also referred to as internal or software methods) are methods that are based on the presence of permanently installed sensors which continuously monitor the system for leakages. The latest developments in hydraulic sensor technology and on-line data acquisition systems have enabled water companies to deploy a larger number of more accurate pressure and flow devices with less cost. These data can be used to monitor the system in real-time and develop methodologies that use the data to detect and pre-localize leaks using Active methods. Pre-localization is the process of defining an area in which the leak exists, instead of pin-pointing exactly its location. This research area has witnessed a significant interest, as indicated in recent review papers (

Online leakage detection system based on EWMA-enhanced Tukey method for water distribution systems

“…The L-Town WDS (Figure 4) is a small hypothetical town with around 10,000 people (Vrachimis et al 2020). The distribution network consists of 905 pipes, 782 junctions, 3 pressure reduction valves (PRV), 2 reservoirs, and 1 water tank.…”

“…The leak area A i,max is defined as a circular orifice with a diameter D i . The relationship of the size of the leak hole change with time is assumed to follow a quadratic function so that the growth rate of leakage is increased as the leak area increases (Vrachimis et al 2020).…”

A Comparison of Model-Based Methods for Leakage Localization in Water Distribution Systems

“…Moreover, an EPANET hydraulic model containing accurate pipe roughness values and the exact water consumption patterns at a 5-min time step for the same two-year period is available. Furthermore, 26 leakages that occurred during the same two-year period are known, and, for each of them, the position in terms of the pipe, the starting time step, the possible repair time step, and the flow rate at a 5-min time step are likewise known (Vrachimis et al 2020). In particular, 15 were bursts, characterized by a steady flow rate ranging between 5 m 3 ∕h and 35 m 3 ∕h .…”

PathoInvest: Pathogen Contamination Investigations during Emergencies

“…The scenario utilizes the benchmark network "L-Town", created for the needs of the BattLeDIM (Battle of the Leakage Detection and Isolation Methods) competition [29] and it is based on a realistic network in Cyprus. The network has been suitably modified for security purposes.…”