Algorithms for leak detection, estimation, isolation and localization in open water channels $ Nadia Bedjaoui à , Erik Weyer Department of Electrical and Electronic Engineering, University of Melbourne, Parkville, VIC 3010, Australia article info Article history: Received 5 November 2009 Accepted 24 June 2010 Available online 27 July 2010 Keywords: Fault detection and isolation Open water channels Fault detection and diagnosis Irrigation channels Environmental systems abstract In this paper, a methodology for leak detection and isolation in open water channels is proposed. It consists of on-line leak detection and off-line leak isolation, estimation and localization. The on-line leak detection is carried out pool by pool using local information only, and a comparison between three leak detection methods in terms of detection performance and ease of implementation is presented. Information from neighboring pools is used to distinguish the effect of a leak from sensor faults. The proposed methods were tested on real experimental data from the Coleambally Channel no. 6 in Australia and showed good performance. & 2010 Elsevier Ltd. All rights reserved. 1. Introduction Water is becoming an increasingly scarce resource in many parts of the world, especially in Australia which is a dry continent and where there is a real need for saving water. Irrigation accounts for about 70% of water usage, and the distribution of water is often achieved via a network of open channels. Water is drawn from a source e.g. a river, lake or reservoir and distributed by a main channel to secondary channels which deliver water to farms. Along the channels, there are gates which control the flow of water, and the stretch of a channel between two gates is referred to as a pool. Management of irrigation channels must take into account the required level of service that has to be provided to the farmers while minimizing water losses. Large water losses are mainly due to a poor management caused by oversupply of water. It has been shown (e.g. Cantoni et al., 2007; Dulhoste, Georges, & Besancon, 2004; Gomez, Rodellar, & Mantecon, 2002; de Halleux, Prieur, Coron, d’Andrea-Novel, & Bastin, 2003; Litrico, Fromion, Baume, Arranja, & Rijo, 2005; Malaterre, Rogers, & Schuurmans, 1998; Ooi & Weyer, 2008; Schuurmans, Hof, Dijkstra, Bosgra, & Brouwer, 1999; Weyer, 2008) that implementation of automatic control systems for regulation of the flows and water levels can significantly improve water delivery efficiency. The presence of leaks (Bedjaoui, Litrico, Koenig, Malaterre, & Bruno, 2009; Weyer & Bastin, 2008) and seepage lead to additional water losses that can be reduced by channel repair and lining. Leaks can be due to a breakdown of the channel wall or the failure of an escape gate. A typical example is a gate to an escape channel not sealing properly and letting water through even when it is fully closed. As reducing water losses is a main priority, it is important to detect the presence of a leak, to estimate the size of the leak to quantify the losses and to determine the location of the leak in order to repair the channel. This paper proposes a leak detection methodology that does the above. The proposed methodology consists of an on-line leak detection algorithm while leak isolation, estimation and localiza- tion are done off-line. The leak detection is carried out pool by pool based on local information only. Three methods based on different residual signals are compared. The comparison is in terms of detection performance and computational cost. All the methods are model-based and generate residuals that are zero in the absence of a leak and which become non-zero when a leak occurs. This is a standard change detection problem that can be solved by e.g. the well known cumulative sum (CUSUM) algorithm. In the case of detection, an off-line algorithm for the estimation of the amount of water lost is activated. This is followed by a routine that also uses data from neighboring pools to distinguish between the effect of leaks and sensor faults. If it is determined that a leak is present, an off-line leak localization method is used to determine the position of the leak within the pool. The complete methodology is tested on data from the Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/conengprac Control Engineering Practice 0967-0661/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.conengprac.2010.06.008 $ This research was supported by Rubicon Systems Australia and the Australian Research Council under the Linkage Grant Scheme, Project LP0349134 and LP0989497. à Corresponding author. E-mail addresses: nadiab@unimelb.edu.au (N. Bedjaoui), ewey@unimelb.edu.au (E. Weyer). Control Engineering Practice 19 (2011) 564–573