The loss of large volumes of treated and frequently pumped water from water distribution systems (WDSs) is environmentally and economically damaging. Cost-effective reduction of water loss through bursts and leakages is however a challenging task for water utilities. New and more efficient methodologies are required for both timely detection and location of bursts and leaks. The recently developed methodology for the automated detection of bursts/leaks at the District Metered Area (DMA) level makes use of the data collected by real-time pressure and/or flow sensors and several Artificial Intelligence (AI) techniques and statistical data analysis tools, including: (i) Wavelets, (ii) Artificial Neural Networks (ANNs), (iii) Statistical Process Control (SPC), and (iv) Bayesian Inference Systems (BISs). The above detection methodology is further developed here with the aim to determine the approximate location of bursts/leaks within the DMA. The new location methodology works by processing in real-time the output information generated by the detection methodology, by means of geostatistical techniques. The novel bursts/leaks location methodology is demonstrated and tested on a case study from a real-life DMA in the United Kingdom with simulated (i.e., engineered) burst events. The results obtained illustrate that the new detection and location system can successfully approximately locate the bursts within a DMA (in addition to detecting the associated events in a fast and reliable manner).
展开▼