Disturbance Extraction for Burst Detection in Water Distribution Networks Using Pressure Measurements

Abstract

Pipe bursts in water distribution networks cause considerable water losses and lead to potential environmental hazards. Effective burst detection methods enable water companies to repair broken pipes in a timely manner and minimize damage and disruption. A data-driven detection method is developed and proven for real-time leak and burst detection in water distribution networks. The method uses a unique integration of disturbance extraction and isolation forest techniques to enable detection of subtle burst signals from normally noisy pressure data. Verification and validation progress from synthetic data to the application to real-life data from large-scale open networks. The method is shown to generate comparable levels of detection, with low false positive rates, to other published techniques but without the need to introduce closed valves and expensive flow meters of district meter area structures and the loss of resilience these cause. The method offers the potential to effectively detect bursts from pressure data alone, helping to better manage the distribution of increasingly stressed water resources through aging pipe networks.Plain Language Summary Bursts and leakage are an inevitable but unacceptable consequence of the aging pipe networks used to supply drinking water. This paper presents a novel burst detection method which extracts burst-induced pressure variations from pressure data alone. The advance presented negates the need to introduce complex zonal structures using closed valves that limit network resilience and without the need for expensive flow meters.

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

“…For example, while the authors in [ 44 ] only consider flow data, the sampling rate in this case is 15 min with a data transmission rate of 30 min, similar to the authors of [ 17 ], who use a sliding window of 2 h, which is considered a substitute for the transmission interval. Other works use sampling rates between 1 min [ 54 ], 5 min [ 55 , 56 , 57 , 58 ], 10 min [ 59 ], or 15 min [ 11 , 60 , 61 , 62 , 63 , 64 , 65 , 66 ], which in one case were resampled [ 62 ]. Choi et al [ 67 ] propose a Kalman-Filter-based methodology making use of adaptive sampling rates between 1 min and 1 h, which would directly impact sensor device runtime.…”

Comparison of model-based techniques for pipe burst location in water distribution networks

“…Although efficient in leakage assessment (i.e., quantifying the amount of lost water), these techniques might not be suited for an early response regarding the burst location; in these situations, more efficient and lower-cost burst location methods are required. Such methods have been actively developed by the scientific community and can be roughly divided into three groups, depending on the type of data and analysis being used [4]: transient-based approaches [5][6][7][8], model-based approaches [9][10][11], and data-driven approaches [12][13][14][15].…”

Flowrate Time Series Processing in Engineering Tools for Water Distribution Networks

“…Regardless of the acquisition and transmission settings, the obtained raw flowrate time series may contain measurement errors, such as missing, repetitive or even false readings (Mounce et al., 2010; Xenochristou et al., 2020). These errors can have their origin in sensor or logger malfunctioning, faulty transmission system due to battery failure, inadequate acquisition range (e.g., above or below meter range or bidirectional flow) and data storage limitations (Loureiro et al., 2016; Machell et al., 2014; Xu et al., 2020). These measurement errors, typically known as outliers or anomalous values (Kirstein et al., 2019), should be detected and corrected before they can be used in engineering applications (e.g., hydraulic modeling, calibration, leak detection) (Romano et al., 2014).…”