Leak Localization in Water Distribution Networks Using Pressure and Data-Driven Classifier Approach

Abstract

Leaks in water distribution networks (WDNs) are one of the main reasons for water loss during fluid transportation. Considering the worldwide problem of water scarcity, added to the challenges that a growing population brings, minimizing water losses through leak detection and localization, timely and efficiently using advanced techniques is an urgent humanitarian need. There are numerous methods being used to localize water leaks in WDNs through constructing hydraulic models or analyzing flow/pressure deviations between the observed data and the estimated values. However, from the application perspective, it is very practical to implement an approach which does not rely too much on measurements and complex models with reasonable computation demand. Under this context, this paper presents a novel method for leak localization which uses a data-driven approach based on limit pressure measurements in WDNs with two stages included: (1) Two different machine learning classifiers based on linear discriminant analysis (LDA) and neural networks (NNET) are developed to determine the probabilities of each node having a leak inside a WDN; (2) Bayesian temporal reasoning is applied afterwards to rescale the probabilities of each possible leak location at each time step after a leak is detected, with the aim of improving the localization accuracy. As an initial illustration, the hypothetical benchmark Hanoi district metered area (DMA) is used as the case study to test the performance of the proposed approach. Using the fitting accuracy and average topological distance (ATD) as performance indicators, the preliminary results reaches more than 80% accuracy in the best cases.

Assessing the Potential of LPWAN Communication Technologies for Near Real-Time Leak Detection in Water Distribution Systems

“…On the other hand, in many cases where real data sets are used, multivariate time series representing normal operation and at times containing engineered leaks are used as provided by a water utility. While in other instances, simulated, artificial datasets, generated with a hydraulic model are applied [ 47 ], in some cases combined with data from engineered leaks for model calibration, validation and performance evaluation [ 43 , 46 ]. For data-driven approaches, data preparation and the structure of the provided data is of particular importance, as the quality of the underlying dataset is the basis for model development and performance [ 48 ].…”

An Advanced Sensor Placement Strategy for Small Leaks Quantification Using Lean Graphs

“…The computational cost can grow significantly when the mathematical modelling is applied to quantifying leaks in large-scale networks. Machine learning algorithms extract an unseen model from a set of data [5] used to quantify leaks, such as genetic algorithms [6,7], support vector machines [8] and artificial neural networks [9,10]. Most of the algorithms require a complete knowledge of network conditions to gain their best performances.…”

Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest Classifier

“…However, it must be noted that the requirement for additional measurements can extend reaction time in case of a pipe burst. In the work of Sun et al [28], a classification approach was utilized where pressure measurements in network nodes with no pressure sensors were estimated using the Kriging method. The Hanoi water distribution network was considered with a wide range of sensors, and it was reported that in the average case 70% accuracy was achieved; however, for some sensor layouts the reported accuracy was below 20%, which is believed to be due to the Kriging interpolation error.…”