Leaks In Water Distribution Networks Research Articles

GA-Sense: Sensor placement strategy for detecting leaks in water distribution networks based on time series flow and genetic algorithm

The detection of leaks in time series flow systems is crucial for efficient and integrated industrial processes. This is especially true when daily demand patterns differ, as this results in fluctuations in the snapshots of water consumption that are commonly used as the basis for placing sensors to detect leaks. This paper introduces a novel method in which the genetic algorithm (GA) is applied to find optimal sensor locations and to enhance the accuracy of leak detection in time series flow data. The method consists of two steps. Firstly, the GA is used to identify the optimal sensor locations using a specific fitness function that accounts for flow patterns, system topology, and leak characteristics. The novelty of the proposed method lies in the weighting scheme of the fitness function, which takes into consideration the frequency of events and the magnitude of leaks at potential locations. Secondly, the selected sensor locations are integrated with an advanced time series data analysis to locate leaks. In this technique, the most consistently performing locations are dynamically selected over time, allowing the model to adapt to varying conditions to maintain optimal sensor placement. Experiments were conducted on a simulated time series flow system with known leak scenarios to evaluate the performance of the proposed method. The results demonstrated the superiority of our GA-based sensor placement strategy in terms of leak detection accuracy and efficiency compared to other methods.•We developed a model called GA-Sense for sensor placement strategy by considering flow patterns to maximize leak detection and localization capabilities.•GA-Sense uses time series data to find strategic sensor locations to identify abnormal flow patterns indicative of leaks.•This approach enhances the accuracy and efficiency of leak detection and localization compared to alternative methods.

District metered areas determination for water distribution networks using improved Girvan-Newman algorithm

In general, partitioning networks into district metered areas (DMA) is known as a practical method to manage the pressure and reduce water leakage in Water Distribution Networks (WDNs). Different methods such as engineering judgment, optimization techniques, and graph theory are proposed to determine the DMA. The graph theory is a traditional method for partitioning networks including several algorithms. One of these algorithms is the Girvan-Newman (GN) algorithm which is based on the mathematical parameters without considering the characteristics of networks. In other words, the hydraulic quality and quantity condition of WDNs are not used for DMA determination. Therefore, in this research, a new method is proposed to improve the performance of the GN algorithm for DMA determination considering the quantities of water networks. For this purpose, the average values of nodal pressure and residual chlorine concentration are calculated and used simultaneously for the junction’s weight determination. Then, the edge scores are calculated based on the junction’s weights, and the edges with the maximum scores are removed until to reach the desired number of DMAs. For comparison purposes, here, the Demand-Driven Simulation Method (DDSM) and Head-Driven Simulation Method (HDSM) analyses are used for the analysis of the Poulakis WDN, selected as a case study. A comparison of the results shows that by using the proposed method, the average pressure values are decreased and the average residual chlorine concentration values are increased. In other words, the pressure values are decreased from 16.32% to 26.23% and the average residual chlorine concentration values are increased from 2.5% to 18.37% in comparison with the results of the standard form of the GN algorithm using both DDSM and HDSM analyses. Furthermore, by using the proposed method, in all scenarios, the hybrid reliability values increase in comparison with the GN algorithm. The results indicate the unique performance of the proposed method in comparison with the standard form of the GN algorithm for DMA determination of WDNs.

From Drips to Data: Preventing Unnecessary Leakages in Water Distribution Networks in Slovakia

Water, a critical resource, suffers from significant losses due to leakages in Water Distribution Networks (WDN). These losses present financial, environmental, and public health challenges. With water utility companies amassing vast amounts of data from enterprise information systems, the opportunity for data-driven leakage detection arises. With the increased use of enterprise information systems, water utility companies are generating vast amounts of data that can be valuable for predicting or detecting water leaks early and also for the development of new, automatic, and effective data-driven leak detection techniques. The presented study utilizes this data, by applying anomaly detection methods to heterogeneous time-series data from various components of the WDN in Slovakia. Our results demonstrate that components of the network, such as measured power consumption, water source temperature, and water source levels show significant positive associations with faults in the Water Distribution Network.

Towards AI-Based Condition Monitoring and Predictive Maintenance for Water Smart Pipes: The SANDMAN Approach

Pipes age and corrosion are the main factors of leakage in water distribution networks. According to the World Resources Institute, European countries will face water problems by 2040. If we take Italy as an example, more than 40% of drinking water was lost in 2020 due to leaky aqueducts. Decrepit pipes can lead to environmental concerns, economical losses, and potential public health problems if water gets contaminated. Localizing leakage positions in an accurate way is often a big challenge. On the other side, replacing decrepit pipes is not an easy task and usually costly. An optimal solution to deal with water leakage is to use smart pipes where appropriate sensors monitoring the conditions of the pipes are incorporated in. Digitalization plays a crucial role here. By providing accurate information about the pipes, and use artificial intelligence techniques for data analysis, potential leakages and their corresponding positions can be detected in time, which allows to schedule a maintenance task as soon as possible. The current paper discusses the use of smart pipes combined with predictive maintenance and shows how this combination improves water leakage detection, hence minimizing water waste, and protecting the environment. The solution was validated in an experimental setup put in place by the Italian company EKSO s.r.l in its factory facilities in Rozallo, Italy. The obtained results show the feasibility of the solution and the relevance of using artificial intelligence techniques for predicting degradation in smart pipes.

LOW-COST SYSTEM FOR LEAK DETECTION IN WATER SUPPLY NETWORKS

With the advancement of technology, sensors have been developed for various applications, including detecting leaks in water distribution networks. This work aims to propose a low-cost system for detecting leaks in water networks. The tests were carried out in a simulated network at the Hydraulics and Fluid Mechanics Laboratory of the Faculty of Civil Engineering and Urbanism at UNICAMP, where sensors were installed to measure flow, water level, noise, turbidity, temperature and an extensometer. Real-time readings were received and processed by a microcontroller board (Arduino UNO), and sent to a web page (PHP language) via a Wi-Fi module (ESP8266). An application was also developed, in Java language, to visualize the readings obtained. In laboratory tests, the sensors proved suitable for detecting leaks and triggering audible and visual alarms in the event of leaks

Model-Based Approach for Leak Detection and Localization in Water Distribution Networks: A Literature Survey

Water loss poses a significant problem for water utilities and has received a lot of attention. To fulfill the increasing global demand for water, water supply system operations must be streamlined, making leak detection and location crucial. Water utilities have developed a number of techniques over time for finding leaks in water distribution networks (WDNs). These methodologies range from simple visual inspection to the use of hardware systems and now software using models and algorithms. Data from flow or pressure measurements, which are required for the analysis of water networks, is becoming more readily available with the introduction of intelligent sensor devices. Along with the introduction of geographic information systems (GIS) and supervisory control and data acquisition (SCADA) in the water sector, the deployment of model-driven methodologies for leak detection and localization has found extensive use. This paper aims to provide a concise introductory reference for early researchers in the development of a model-based approach for leak detection in WDNs. Thus, a survey of model-based approaches is presented, along with current research trends and applications of model-driven methodologies for leak detection in water supply networks. Several model-driven approaches and research studies for each case are discussed. Some challenges and research gaps are also discussed.

Locating Multiple Leaks in Water Distribution Networks Combining Physically Based and Data-Driven Models and High-Performance Computing

Locating Multiple Leaks in Water Distribution Networks Combining Physically Based and Data-Driven Models and High-Performance Computing

Investigating the possibility of leakage detection in water distribution networks using cosmic ray neutrons in the thermal region

Water distribution systems can experience high levels of leakage, originating from different sources, such as deterioration due to aging of pipes and fittings, material defects, and corrosion. In addition to causing financial losses and supply problems, leakages in treated water distribution also represent a risk for public health. Despite several techniques for leak detection are already available, there is still a lot of interest in new non-invasive approaches, especially for scenarios where acoustic techniques struggle, such as in noisy environmental conditions.In this work we investigated the possibility of using cosmic ray (CR) neutrons for the detection of underground leakages in water distribution networks, by exploiting the difference in the above ground thermal neutron flux between dry and wet soil conditions. The potential of the technique has been assessed by means of an extensive set of Monte Carlo simulations based on GEANT4, involving realistic scenarios based on the Italian aqueduct design guidelines. Simulation studies focused on sandy soils and results suggest that a significative signal, associated with a leakage, could be detected with a data-taking lasting from a few minutes to a half-hour, depending on the environmental soil moisture, the leaking water distribution in soil, and the soil chemical composition.Finally, a brief description of a new portable and low-cost detector for thermal neutrons, currently under commission, is also presented.

Simultaneous Pipe Leak Detection and Localization Using Attention-Based Deep Learning Autoencoder

Water distribution networks are often susceptible to pipeline leaks caused by mechanical damages, natural hazards, corrosion, and other factors. This paper focuses on the detection of leaks in water distribution networks (WDN) using a data-driven approach based on machine learning. A hybrid autoencoder neural network (AE) is developed, which utilizes unsupervised learning to address the issue of unbalanced data (as anomalies are rare events). The AE consists of a 3DCNN encoder, a ConvLSTM decoder, and a ConvLSTM future predictor, making the anomaly detection robust. Additionally, spatial and temporal attention mechanisms are employed to enhance leak localization. The AE first learns the expected behavior and subsequently detects leaks by identifying deviations from this expected behavior. To evaluate the performance of the proposed method, the Water Network Tool for Resilience (WNTR) simulator is utilized to generate water pressure and flow rate data in a water supply network. Various conditions, such as fluctuating water demands, data noise, and the presence of leaks, are considered using the pressure-driven demand (PDD) method. Datasets with and without pipe leaks are obtained, where the AE is trained using the dataset without leaks and tested using the dataset with simulated pipe leaks. The results, based on a benchmark WDN and a confusion matrix analysis, demonstrate that the proposed method successfully identifies leaks in 96% of cases and a false positive rate of 4% compared to two baselines: a multichannel CNN encoder with LSTM decoder (MC-CNN-LSTM) and a random forest and model based on supervised learning with a false positive rate of 8% and 15%, respectively. Furthermore, a real case study demonstrates the applicability of the developed model for leak detection in the operational conditions of water supply networks using inline sensor data.

Laboratory implementation of the noise correlation methodology for leak location in water distribution systems

Nowadays, reducing the loss of water resources is of primary concern in many countries worldwide. Different methodologies exist for the identification and localisation of leaks in water distribution networks, which can be applied simultaneously at a different scale, to improve success rate. Acoustic correlation has been employed for many years to localise such leakages, based on the post-processing of vibroacoustic signals acquired by two sensors placed on both sides of the pipe where the leakage is supposed to be located. The practical application of this methodology still presents some significant issues. The implementation of the noise correlation methodology was investigated through an experimental test rig by simulating leakages in different conditions of pressure and flow and investigating the effects of background noise. Hydrophonic measurements and recordings were performed, identifying the frequency range affected by the leak noise, and the signal to assess the accuracy of the noise correlation method. A pre-filtering of the signals to reduce the effect of background noise masking was also investigated. The accuracy of acoustic correlation in localising the leakage was assessed by computing the propagation velocity, which depends on the vibroacoustic fluid-structure interaction, through approximated analytical equations

Digital Twin of a Hydraulic System with Leak Diagnosis Applications

This paper presents the design and development of a digital twin to diagnose leaks in water distribution networks. The digital twin allows for the remote operation of the hydraulic system’s actuators using embedded microcontrollers integrated with Internet of Things (IoT) capabilities. Pressure head and flow rate measurements are received online in the operator interface, and hydraulic simulations are performed with a well-calibrated EPANET model of the hydraulic system to estimate the pressure head at nodes without sensors. A genetic algorithm was designed to detect and estimate the size of the leaks online. Different experiments were carried out to validate the online application of the method based on the digital twin and under a multi-leak event.

Using adaptive neuro-fuzzy inference system and imperialist competitive algorithm for leak detection in pipe networks

Finding the position and quantity of leakage in water distribution networks (WDNs) is challenging in cities with old water pipelines. Previous studies proposed physical and data-driven techniques to find the location and quantity of leakage in WDNs. Most approaches rely on large sample of measurements from the WDN which makes them impractical. We propose an efficient model incorporating Adaptive Neuro-Fuzzy Inference System (ANFIS) and Imperialist Competitive Algorithm (ICA) techniques, to reduce the number of required samples. ANFIS approximates the locations and quantities of leakage, while the ICA corrects its estimations. Due to nonlinearity, the application of ICA alone results in long run times, while ANFIS reduces the number of decision variables for ICA, and hence the convergence rate improves. In other words, we reduce the search space leading to reduced computational time and improved accuracy. Results show that the normalized predicted leakage values had a maximum error rate of 10%.

Modeling of pressure-dependent background leakages in water distribution networks

In last decades, several mathematical models have been proposed to simulate background leakages in water distribution networks (WDNs). Some of these models already consider the dependence of leakages to pressure, but they still neglect the gradient of pressure along the pipes. In this article, new models to take into account of this gradient are presented. One of them computes reference background leakage outflows, using a recursive algorithm which discretizes the pipes into sub-pipes until the hydraulic grade line (HGL) along each pipe converges. The other new models consist in gradually refining a state of the art one. All models are then tested and compared on both a single leaky pipe and a WDN derived from a real leaky network. The results of this comparison show clearly the better estimations obtained from our new models when compared to the existing one. Finally, accurate leakage models are essential to estimate the level of leakages and, more generally, the good working order of WDNs. Thus, our new models will help in taking the best decisions for optimal functioning and rehabilitation of the WDNs. Moreover, our recursive discretization approach could be reused for other applications in WDNs, or derived to more general fields of applied mathematics and scientific computation.

Leak localization in water distribution networks using GIS-Enhanced autoencoders

ABSTRACT Water loss due to persistent leakages in water distribution networks remains a substantive problem around the world, all the more so given noticeable trends of increasing global water scarcities. In this paper, we present a data-driven leak localization approach leveraging a connected Geographical Information System together with an autoencoder to perform anomaly detection on time-variable sensor data. Data-driven approaches are able to circumvent many of the uncertainty issues associated with model-based approaches, but they usually require significant amounts of high-quality data, reflecting many different leak scenarios, to perform well. Our approach obviates this requirement by relying only on leakless data during model training. We examine the efficacy of this approach on 19 realistic leak experiments conducted in the field. Based on these evaluations, we were able to achieve average search costs as low as 2.2 kilometers, for a total network length of 215 kilometers.

Employing Robotics and Deep Learning in Underground Leak Detection

Leaks in water distribution networks (WDNs) produce significant economic losses. These leaks from underground pipelines affect the surrounding environment in different ways that can be detected using various technologies. This article introduces an unmanned ground vehicle (UGV) equipped with an infrared temperature sensor to remotely detect thermal anomalies on the surface caused by underground leaking pipelines. An MLX90614 low-cost thermopile infrared sensor was proposed to trace the surface temperature above leaking pipelines and record the corresponding position of each reading. Two positioning techniques were tested independently: a satellite navigation system and odometry using a magnetic encoder. Raw data from the temperature sensor along with the corresponding locations were recorded on a MicroSD card using an Arduino Uno Board through analog-to-digital converters (ADCs) and serial peripheral interface (SPI) bus. Experimental work showed the ability of the robot in detecting thermal anomalies caused by leaks accurately with the required trace resolution. Finally, a 2-D convolutional neural network (CNN) solution is introduced for distinguishing true detections from false alarms.

Stochastic model-based leakage prediction in water mains considering pipe condition uncertainties

Energy loss through leakage in water distribution networks (WDNs) has been estimated to add up to 20% of the freshwater supply in the US. One common practice to battle non-revenue water in WDNs is proactively pinpointing leak sources through both conventional and novel, data-driven methods. Conventional, in-practice leakage detection approaches suffer from costliness and laboriousness, and thus novel model-based techniques are becoming a robust, accurate, and cost-effective alternative. One of the existing limitations in the literature regarding novel leak localization methods and thus the novelty in this study is factoring in the stochastic behavior of water pipe conditions while predicting leakage through consumption and cyber-monitoring data. Hence, this paper presents a direct, data-driven, artificially intelligent predictive tool for leak localization and severity measurement. In this leakage prediction model, uncertain pipe conditions such as pipe diameters and roughness coefficients are stochastically approximated in a Monte Carlo simulations framework using water quality-included empirical formulation. A novel accuracy index is also developed to measure the precision of leak localization by employing a graph connectivity algorithm. Prediction results and sensitivity analyses on a large WDN benchmark demonstrate that the proposed stochastic leakage prediction model offers robustness in locating leak sources and measuring severity in various leak test cases. Not only does this research offer new numerical insights into the probabilistic nature of water pipelines’ dynamic parameters amid leakage, it also pushes the boundary of smart leak detection approaches at water utilities by offering cost-effective and data-driven horizons.

GIS-Based Identification of Locations in Water Distribution Networks Vulnerable to Leakage

The detection of leakages in Water Distribution Networks (WDNs) is usually challenging and identifying their locations may take a long time. Current water leak detection methods such as model-based and measurement-based approaches face significant limitations that impact response times, resource requirements, accuracy, and location identification. This paper presents a method for determining locations in the WDNs that are vulnerable to leakage by combining six leakage-conditioning factors using logistic regression and vulnerability analysis. The proposed model considered three fixed physical factors (pipe length per junction, number of fittings per length, and pipe friction factor) and three varying operational aspects (drop in pressure, decrease in flow, and variations in chlorine levels). The model performance was validated using 13 district metered areas (DMAs) of the Sharjah Electricity and Water Authority (SEWA) WDN using ArcGIS. Each of the six conditioning factors was assigned a weight that reflects its contribution to leakage in the WDNs based on the Analytic Hierarchy Process (AHP) method. The highest weight was set to 0.25 for both pressure and flow, while 0.2 and 0.14 were set for the chlorine and number of fittings per length, respectively. The minimum weight was set to 0.08 for both length per junction and friction factor. When the model runs, it produces vulnerability to leakage maps, which indicate the DMAs’ vulnerability classes ranging from very high to very low. Real-world data and different scenarios were used to validate the method, and the areas vulnerable to leakage were successfully identified based on fixed physical and varying operational factors. This vulnerability map will provide a comprehensive understanding of the risks facing a system and help stakeholders develop and implement strategies to mitigate the leakage. Therefore, water utility companies can employ this method for corrective maintenance activities and daily operations. The proposed approach can offer a valuable tool for reducing water production costs and increasing the efficiency of WDN.

A microservice architecture for leak localization in water distribution networks using hybrid AI

Abstract Up to 30% of all drinking water is wasted due to leaks in water distribution networks (WDNs). In times of drought and water shortage, wasting so much drinking water has a considerable environmental and financial cost. In this paper, we present a microservice architecture for leak localization in WDNs, where heterogeneous sources of data consisting of sensor measurements, Geographic Information System (GIS), and Customer Relationship Management (CRM) data are used to feed a leak monitoring solution which combines hybrid data-driven and model-based leak detection and localization methodologies. The solution is validated using in-field leak experiments in an operational WDN. The final leak probabilities are presented in a visualization dashboard. The search zone for most leaks is reduced to a few kilometers or less. For other leaks, the solution is able to indicate a larger search zone to reflect its higher leak prediction uncertainty.

Application of Machine Learning for Leak Localization in Water Supply Networks

Water distribution networks (WDNs) in urban areas are predominantly underground for seamless freshwater transmission. As a result, monitoring their health is often complicated, requiring expensive equipment and methodologies. This study proposes a low-cost approach to locating leakages in WDNs in an urban setting, leveraging acoustic signal behavior and machine learning. An inexpensive noise logger was used to collect acoustic signals from the water mains. The signals underwent empirical mode decomposition, feature extraction, and denoising to separate pure leak signals from background noises. Two regression machine learning algorithms, support vector machines (SVM) and ensemble k-nearest neighbors (k-NN), were then employed to predict the leak’s location using the features as input. The SVM achieved a validation accuracy of 82.50%, while the k-NN achieved 83.75%. Since the study proposes using single noise loggers, classification k-NN and decision trees (DTs) were used to predict the leak’s direction. The k-NN performed better than the DT, with a validation accuracy of 97.50%, while the latter achieved 78.75%. The models are able to predict leak locations in water mains in urban settings, as the study was conducted in a similar setting.

Leakage detection in water distribution networks using machine-learning strategies

Abstract This work proposes a reliable leakage detection methodology for water distribution networks (WDNs) using machine-learning strategies. Our solution aims at detecting leakage in WDNs using efficient machine-learning strategies. We analyze pressure measurements from pumps in district metered areas (DMAs) in Stockholm, Sweden, where we consider a residential DMA of the water distribution network. Our proposed methodology uses learning strategies from unsupervised learning (K-means and cluster validation techniques), and supervised learning (learning vector quantization algorithms). The learning strategies we propose have low complexity, and the numerical experiments show the potential of using machine-learning strategies in leakage detection for monitored WDNs. Specifically, our experiments show that the proposed learning strategies are able to obtain correct classification rates up to 93.98%.