Burst Detection and Location in Water Distribution Systems

Water Distribution Systems Analysis 2010, 2011

In this paper we present a technique for detecting and locating burst events in pipelines. The proposed method uses wavelet analysis of the high-rate pressure data to detect pipe burst events. Multiscale wavelet analysis of the pressure signal will be shown to be robust to impulsive noise encountered in the physical phenomena under observation. The wavelet coefficients also allow us to obtain additional information about the nature of the signal feature detected, which can used for further feature classification. A local search method is also proposed to accurately determine the arrival time of the pressure front associated with the burst event. The detection performance of these algorithms is verified through leak-off experiments performed on the WaterWiSe@SG test bed deployed on the water distribution system in Singapore. We also propose a graph-based search algorithm which uses the arrival times of the pressure front at different locations within the water distribution system to determine the actual location of the pipe burst event.

Procedia Engineering, 2014

This paper presents a burst detection methodology that utilizes distributed real time sensor data in a district metered area using a data assimilation method and a hydraulic model. A sensitivity analysis was applied to evaluate the performance of various burst detection metrics under different conditions, and to identify appropriate thresholds for online burst detection using artificial generated burst events. It was found that the best performing metrics are the corrected flow residual and normalized corrected flow residual. This metric can be effective to detect bursts in a timely and reliable manner within a district metering area under assumed test conditions.

2014

The residual of the Kaiman filter applied to flow measurements in water distribution systems has been found in recent research to strongly correlate with bursts. A positive residual represents an unexpected extra flow which is likely caused by an event such as a burst in the downstream network. However there is a certain level of uncertainty in the flow measurements leading to a noisy and fluctuating residual during normal and abnormal network operations. This work investigates how the change of a residual threshold affects the success rates of burst alarming during burst periods and non-burst periods respectively, and presents a statistical method to automatically select a suitable residual threshold to tolerate the uncertainty in flow measurements. In addition, this work also investigates how the number of burst alarms is influenced by using different flow measurement sampling frequencies and different averaging window sizes on the data sets. Engineered tests with 3 simulated burst events were carried out to validate the methods. The results showed that the threshold proposed in this paper can produce a relatively high success rate on burst alarming. In addition, the results also showed that the current sampling interval of 15 minutes is suitable for burst detection from flow data, and the use of a window averaged over the last several hours can reduce the continual on/off alarm states.

Journal of Signal Processing Systems, 2013

In this paper we present techniques for detecting and locating transient pipe burst events in water distribution systems. The proposed method uses multiscale wavelet analysis of high rate pressure data recorded to detect transient events. Both wavelet coefficients and Lipschitz exponents provide additional information about the nature of the signal feature detected and can be used for feature classification. A local search method is proposed to estimate accurately the arrival time of the pressure transient associated with a pipe burst event. We also propose a graph-based localization algorithm which uses the arrival times of the pressure transient at different measurement points within the water distribution system to determine the actual location (or source) of the pipe burst. The detection and localization performance of these algorithms is validated through leak-off experiments performed on the WaterWiSe@SG wireless sensor network test bed, deployed on the drinking water distribution system in Singapore. Based on these experiments, we also present a systematic analysis of the sources of localization error.

Journal of Water Resources Planning and Management, 2012

The following work presents a multivariate statistical technique applied to the control of water inflows into district metering areas (DMA) of urban networks. This technique, called Principal Component Analysis (PCA), allows for a sensitive and quick analysis of the inflows into a DMA without hassling mathematical algorithms. PCA technique simplifies the original set of flow rate data recorded by the SCADA system, synthesizing the most significant information into a statistical model which is able to explain most of the behaviour of the water distribution network. PCA technique also allows for the establishment of control charts that help system operators in the identification of anomalous behaviours regarding water use, bursts or illegal connections. The described technique has been proved to offer high detection sensitivity to bursts or other unexpected consumptions.

Journal of Pipeline Systems Engineering and Practice, 2011

Automatic burst and leak detection in water distribution systems plays an important role in water saving and management. This research develops a novel burst detection method of using adaptive Kalman filtering on hydraulic measurements of flow and pressure at district meter area (DMA) level. Adaptive Kalman filtering is used to model normal water usage (or alternatively water pressure), so the residual of the filter (e.g., the difference between the predicted flow and the measured flow) represents the amount of abnormal water usage relating to the bursts (or newly occurred leaks) in the downstream network. The results from a series of engineered tests which simulated flushing show that the size of the bursts and leaks strongly correlates with the residual of the filter. Finally, the method was applied to data from several real DMAs in the north of England, and the results show that the detected bursts correspond well to known historical operational information such as customer complaints’ records and work management (repair) data. The results suggest that flow measurement data are more sensitive to a burst or leak than the pressure measurement data.

Water

A large volume of the water produced for public supply is lost in the systems between sources and consumers. An important—in many cases the greatest—fraction of these losses are physical losses, mainly related to leaks and bursts in pipes and in consumer connections. Fast detection and location of bursts plays an important role in the design of operation strategies for water loss control, since this helps reduce the volume lost from the instant the event occurs until its effective repair (run time). The transient pressure signals caused by bursts contain important information about their location and magnitude, and stamp on any of these events a specific "hydraulic signature". The present work proposes and evaluates three methods to disaggregate transient signals, which are used afterwards to train artificial neural networks (ANNs) to identify burst locations and calculate the leaked flow. In addition, a clustering process is also used to group similar signals, and then tr...

Applied Sciences

Developing data-driven models for bursts detection is currently a demanding challenge for efficient and sustainable management of water supply systems. The main limit in the progress of these models lies in the large amount of accurate data required. The aim is to present a methodology for the generation of reliable data, which are fundamental to train anomaly detection models and set alarms. Thus, the results of the proposed methodology is to provide suitable water consumption data. The presented procedure consists of stochastic modelling of water request and hydraulic pipes bursts simulation to yield suitable synthetic time series of flow rates, for instance, inlet flows of district metered areas and small water supply systems. The water request is obtained through the superimposition of different components, such as the daily, the weekly, and the yearly trends jointly with a random normal distributed component based on the consumption mean and variance, and the number of users ag...

World Water Congress 2005: Impacts of Global Climate Change - Proceedings of the 2005 World Water and Environmental Resources Congress, 2005

Authors may post the final draft of their work on open, unrestricted Internet sites or deposit it in an institutional repository when the draft contains a link to the bibliographic record of the published version in the ASCE Civil Engineering Database. "Final draft" means the version submitted to ASCE after peer review and prior to copyediting or other ASCE production activities; it does not include the copyedited version, the page proof, or a PDF of the published version

Procedia Engineering, 2015

The common existence of pressure transients in operational water distribution systems (WDS) requires their characterisation and assessment of their impact. This paper performs such characterisation by evidencing the occurrence and the differences in pressure transient behaviour in complex WDS. Ten samples of continuously recorded high resolution pressures from diverse networks and sources were analysed. The presented pressure traces show regular and occasional pressure transient waves in various complex networks. Histogram analysis of the rate of change of head provides some insight into transient behaviour in these sites. Although there was no distinct correlation between network characteristics (ie. length, diameter, age) and transient behaviour, network complexity was observed to change the transient characteristics. Transient characteristics were observed to be strongly influenced by likely sources, in particular commercial customers. The data highlights the need to understand, quantify and characterise transients and hence link to possible impacts, such as structural or water quality failures.