State estimation and leak detection in water distribution networks

Leakages in the water distribution networks (WDNs) are real problems for utilities and other governmental agencies. Timely leak detection and location identification have been challenges. In this paper, an integrated approach to geospatial and infrared image processing was used for robust leak detection. The method combines drops in flow, pressure, and chlorine residuals to determine potential water leakage locations in the WDN using Geographic Information System (GIS) techniques. GIS layers were created from the hourly values of these three parameters for the city of Sharjah provided by the Sharjah Electricity, Water, and Gas Authority (SEWA). These layers are then analyzed for locations with dropped values of each of the parameters and are overlaid with each other. In the case where there were no overlaying locations between flow and pressure, further water quality analysis was avoided, assuming no potential leak. In the case where there are locations with drops in flow and pressure layers, these overlaying locations are then examined for drops in chlorine values. If overlaying locations are found, then these regions are considered potential leak locations. Once potential leak locations are identified, a specialized remote sensing technique can be used to pinpoint the leak location. This study also demonstrated the suitability of using an infrared camera for leak detection in a laboratory-based setup. This paper concludes that the following methodology can help water utility companies in the timely detection of leaks, saving money, time, and effort.

Detecting and mitigating leaks in water distribution networks are vital for water conservation. Machine-learning-based (ML) acoustic leak detection models were introduced as effective alternatives for leak management. However, ML model training requires sufficient labeled data, which hinders related development. To address this challenge, this study employed contrastive learning (CL) for leak detection using limited labeled signals. Experimental results indicate that flip-x and amplitude scaling are optimal combinations for contrastive learning. Besides, ablation and t-distributed stochastic neighbor embedding (t-SNE) results reveal that increasing the model depth does not always yield performance improvement, and five convolutional blocks are more suitable for the leak detection problem in this study. Comparison experiments demonstrate that contrastive learning outperforms supervised learning (SL) when trained with insufficient labeled data. The out-of-sample validation results indicate that the proposed leak detection model is robust and effective in unexplored pipelines. The proposed framework significantly advances ML-based leak detection research and supports sustainable water management practices.

Various leakage detection algorithms have been proposed in the literature to pursue prompt leakage identification in water distribution networks using sensor data. Some use stochastic process control (SPC) methods, where the leak identification is considered an anomaly detection problem. However, several practical challenges emerge from SPC in leakage detection, including outliers, random fluctuations, and non-normally distributed and autocorrelated data. Here, we contribute a comparative analysis of seven advanced SPC techniques based on cumulative sum for leakage identification in water distribution networks, selected based on an extensive literature review of approximately 100 publications. We test their usability for leakage detection by integrating them into a state-of-the-art data-driven detection method and demonstrating them on the L-Town benchmark network. Ultimately, this study offers actionable recommendations to guide the selection of change point detection methods for leakage detection: our results indicate that transformations are best combined with robust statistics. Nonparametric and adaptive methods are suited to retrieve reliable alarms under heterogeneous conditions. Lastly, incorporating a decorrelation step improves the detection performance if irregular water demand patterns occur in the network.

One great challenge for waterworks is effective leakage detection. This paper presents a method based on the self-organising map for leakage detection in a water distribution network. The data used for training and validating the test results consist of vectors of the flow meter readings and knowledge of reported leak locations. The most important factor facilitating the self-organising-map-based modelling of leaks is the developed leak function. The results of the experiments presented show that the model trained on flow data can detect leaks in a defined distribution network area.

The present Ph.D. thesis is devoted to the experimental and numerical analysis of the Pumps As Turbines (PATs) application in hydraulic systems and, specifically, in Water Distribution Networks (WDNs) to generate Energy Recovery (ER) in both urban and rural areas. The use of PATs in WDNs represents a relevant topic in the field of the innovative Best Management Practices (BMPs) of WDNs because addressed to both the pressure regulation and the hydropower generation. Thus, it could represent an alternative approach to the use of Pressure Reducing Valves (PRVs), benefitting of the exceed pressure to produce sustainable energy, otherwise dispersed. The PATs application consists in the use of pump models commercially available running in reverse mode, by inverting the flow direction and making use of the motor as an electrical generator. Their employment can be intended as an effective alternative to the traditional micro-turbines in WDNs, having lower investment and maintenance costs and allowing the model selection among a wide set of pump models available in the market. Conversely, PATs usually assure lower operative ranges, however expandable by arranging proper hydraulic and/or electrical regulations. In addition to the evaluation of the effective opportunities for pressure reduction in water systems, one of the main issue, concerning the PATs use for hydropower generation, regards the limited knowledge about their performance curves, rarely made available by manufactures. In the literature, several experimental, theoretical and numerical models have been proposed, aiming at both predicting the PAT performances and reproducing the internal fluid dynamics. In this regard, Computational Fluid Dynamics (CFD) models are widely applied, because able to both simulate the fluid dynamics of actual configurations and analyse the performance improvement obtainable by modifying the geometric properties of the considered turbo-machine. However, the current knowledge about PAT performances, available in the literature, does not allow to characterize the wide set of pump models, which are on the market, useful for running as turbines. With the aim of making this limitation smaller, in the present study an experimental and numerical investigation of centrifugal PAT models was performed, presenting an operative procedure for the optimal selection of centrifugal PATs in WDNs. A laboratory set-up was specfically installed at the Department of Civil, Architectural and Environmental Engineering (DICEA) of the University of Naples (IT) Federico II where several centrifugal PATs, having different geometric configurations and motor equipment, were tested. The experiments were performed in wide flow rate ranges, typical of WDNs supplying small-medium sized urban and rural centers, by varying the PAT rotational speed from 300 to 3000 rpm. A total of about 7400 operative configurations were investigated, in compliance with the laboratory set-up potentialities, achieving generated head drops up to about 70 m and produced electrical powers up to 16.3 kW. From the experimental analysis, with the aim of characterizing the investigated PATs, comparison with models from the literature was performed and, in order to provide effective tools for predicting the performances of centrifugal PATs, analytic relationships were derived, as a function of the geometric configuration, the number of stages and the motor efficiency class equipment. The attention was specifically drawn to the experimental characterization of: a) a classical Horizontal Axis Single-Stage PAT running at wider operative ranges than those discussed in the literature; b) PAT models, not adequately taken into account in the literature, such as centrifugal Single-Stage and Multi-Stage Vertical Axis PATs. Taking advantage of the experimental results, an operative procedure was developed for the optimal selection of PATs in WDNs, able to maximize the potential ER in WDNs. It could be intended as a basic tool for implementing a Decision Support System (DSS) for hydropower generation in WDNs. Finally, the development of a numerical CFD model, by applying the ANSYS® Fluent™ code, allowed to test the reliability of numerical simulations to both predict the PAT performances and investigate the fluid behaviour across centrifugal PATs.

Objective: This study aimed to implement a model for detecting and locating leaks in water distribution networks, using artificial intelligence (artificial neural networks). Background: Traditional leak detection methods, such as visual inspections and flow and pressure analysis, are being progressively complemented by solutions based on smart sensors and machine learning techniques. Method: Using the method of characteristics to solve the momentum and continuity equations, along with nodal equations, a hydraulic transient simulator was created for water distribution networks with leaks. For the leak simulation, the orifice law (Torricelli's Law) was used. Results and Discussion: The developed model successfully located the leak with an error of approximately 5.65%, corresponding to a distance of approximately 61.22 m for the studied network. Research Implications: Deteriorated infrastructure in water distribution systems can lead to leakage losses, reduced water transport capacity, component failures, increased maintenance and operating costs, frequent system interruptions, and decreased reliability. Thus, the exploration of rapid leak detection and location methods has increased, aiming to mitigate environmental impacts and the losses of natural and financial resources. Originality/Value: The methodological distinction lies in using pressure wave travel times as input data, rather than absolute pressure values. This approach eliminates the need for detailed hydraulic network calibration, reduces sensitivity to operational noise, and broadens the practical applicability of the model.

This chapter presents a data-driven based approach for detection of leaks in water distribution networks in which the demand is formed by a known periodic pattern plus a stochastic variable. The leak detection method is based on an adaptation of the dynamic principal component analysis (DPCA), and it is assumed that only pressures at selected consumption nodes are measured. Since the variables of water distribution networks (WDNs), even in normal conditions, are nonstationary and time-correlated the data are preprocessed with a periodic transformation previous to the application of DPCA. The proposed approach is validated with the Hanoi network model. The performance is evaluated with three indexes: the leak detection rate, the false alarm rate, and the delay of the detection with respect to the leak’s occurrence time. All of them are satisfactory for diverse leaks’ scenarios, and the proposed approach presents an improvement in the leak detection rate of approximately \(70\%\) as compared with the traditional PCA and DPCA methods.

In this paper, a pragmatic approach for leakage detection and localization is presented to solve the problem introduced within the framework of the Battle of the Leakage Detection and Isolation Methods (BattLeDIM). The method is based on the application of the hydraulic model of the water distribution network and the comparison of simulated pressures against the corresponding values observed in field. In particular, it consists of two phases: (1) calibration of the hydraulic model of the network; (2) detection and localization of the leakages affecting the water distribution network through the application of engineering judgment and the adoption of an enumerative procedure. The method was applied to the case-study network of L-Town, enabling 16 of 23 leakages to be efficiently detected and 11 of these also accurately localized. The proposed method is simple and transparent and can aid water utilities in water leakage management.

The measurements given by Supervisory Control And Data Acquisition (SCADA) systems are generally insufficient for Distribution network State Estimation (DSE). The accuracy of DSE seriously relies on the pseudo measurements, which are generally inaccurate. In historical scene, measurements given by smart distribution transformer meter (SDTM) can be used to enrich the measurements for DSE. The hybrid measurements of SCADA and SDTM can satisfy the observability of DSE with certain redundancy. By using rectangular coordination and introducing intermediate variables and quadratic equalities, the SCADA/SDTM hybrid DSE is modeled as quadratic constrained quadratic estimation (QCQE) problem, which can be solved by traditional state estimation methods with sparse techniques efficiently and conveniently. Simulations on the IEEE 33-bus system show the accuracy of the hybrid DSE. The proposed DSE has been implemented in distribution networks in Zhejiang and shows good efficiency and accuracy.

An accurately calibrated hydraulic model of a water distribution network (WDN) is crucial for tasks such as leakage detection, resilience assessment, and rehabilitation planning. Since model calibration relies on real data, increasing monitoring locations for pressure and flow raises both accuracy and cost. Therefore, selecting suitable data collection points is crucial for efficient calibration. This research proposes a practical, graph-theory-inspired approach for determining pressure sensor locations based on the network topology and the demand distribution in the WDN. The method employs a customized metric of Node Demand Betweenness Centrality (NDBC) as a criterion to estimate flow distribution, and thereby identifies influential paths within WDNs. Sensor locations are then optimized using the nondominated sorting genetic algorithm II (NSGA-II), which identifies nodes located along the most influential paths in WDNs while maximizing network coverage by increasing the graph-based hydraulic distance between sensors. This methodology is further supported by additional graph analysis methods, which serve as a faster alternative and are compared against a widely utilized hydraulic model-based approach. The proposed approach was applied to both the Modena and L-Town WDNs. In both case studies, the calibrated models based on pressure data from the sensor locations proposed in this research demonstrated satisfactory accuracy.

In this research, reliability indicators of water distribution networks are evaluated under pipe failure conditions. The case studies include two benchmark and one real-life water distribution networks in Iran with more hydraulic constraints. Some important reliability indicators are presented such as resilience index, network resilience, modified resilience index and minimum surplus head index. GANetXL is used to do one-objective and two-objective optimization of the previously mentioned water distribution networks in order to not only minimize the cost, but also maximize the reliability indicators. Moreover, the results of a statistical analysis for each pipe is used to determine the sensitive pipes that are of the most failure probability. GANetXL is an optimization tool in Excel environment and works based on Genetic Algorithm. GANetXL has the capability of being linked to EPANET (Hydraulic simulation software). The results obtained clearly show that network resilience index is of poor performance when compared with the other indexes under pipe failure conditions, especially in real-life networks that include small pipe diameters. It was also showed that if a water distribution network was optimized only in terms of cost, there would be an unacceptable pressure drop at some nodes in case of pipe failure.

In this paper a new approach to state estimation in distribution networks is proposed. This approach is more robust against large uncertainties of the state estimation inputs than the conventional method. Traditionally, the goal of state estimation was to estimate the exact value of network parameters, such as voltages and currents. This works well in transmission networks where many real time measurements are available. In distribution networks, however, only few real-time measurements are available. This means that the estimated state can be significantly different from the actual network state. Therefore, the focus of the proposed robust state estimation is shifted from estimating the exact values of the network parameters to the confidence that these parameters are within their respective constraints. This approach is able to provide useful results for distribution network operation, even if large uncertainties are present in the estimated network state.

Many Machine Learning models are vulnerable to adversarial attacks: There exist methodologies that add a small (imperceptible) perturbation to an input such that the model comes up with a wrong prediction. Better understanding of such attacks is crucial in particular for models used in security-critical domains, such as monitoring of water distribution networks, in order to devise counter-measures enhancing model robustness and trustworthiness. We propose a taxonomy for adversarial attacks against machine learning based leakage detectors in water distribution networks. Following up on this, we focus on a particular type of attack: an adversary searching the least sensitive point, that is, the location in the water network where the largest possible undetected leak could occur. Based on a mathematical formalization of the least sensitive point problem, we use three different algorithmic approaches to find a solution. Results are evaluated on two benchmark water distribution networks.

This paper represents the initial steps in the development of a new system for leak detection and localization in the water distribution network. This system is based on wireless underground sensor networks. First, we start with determining an optimal location of leak control points in the water distribution network especially in areas which are difficult to control. Then, we move to the development of a distributed algorithm to detect and locate in real time the small and the large leaks in the water distribution system. This algorithm is accompanied by a data aggregation method which minimizes the amount of transmitted information and avoids false alarms. In addition, we present a new cross-layer method in order to optimize the performance of the underground wireless sensor network.

Leakage presents a significant challenge in water distribution network (WDN) planning and management. This study introduces a novel methodology for hydraulic model calibration and leak detection based on MCMC-Statistical Distance. The central hypothesis posits that the presence of leaks induces fluctuations in estimated pipe roughness coefficients (PRCs) as the pipe flow changes, reflecting the altered behavior of leaking pipes in energy dissipation. The proposed model comprises two distinct algorithms: (1) PRC estimation using MCMC and (2) a leakage detection algorithm employing a Kolmogorov–Smirnov test. Demonstrated in a simple water distribution network with various scenarios, the results illustrate the model’s potential for real-time leak detection.