Detecting Pipe Bursts Using Heuristic and CUSUM Methods

AbstractPipe bursts in a drinking water distribution system lead to water losses, interruption of supply, and damage to streets and houses due to the uncontrolled water flow. To minimize the negative consequences of pipe bursts, an early detection is necessary. This paper describes a heuristic burst detection method, which continuously compares forecasted and measured values of the water demand. The forecasts of the water demand were generated by an adaptive water demand forecasting model. To test the method, a dataset of five years of water demand data in a supply area in the Western part of the Netherlands was collected. The method was tested on a subset of the data (only the winter months) in which 9 (larger) burst events were reported. The detection probability for the reported bursts was 44.4%, at an acceptable rate of false alarms of 5.0%. The results were compared with the CUSUM method, which is a general statistical process control (SPC) method to identify anomalies in time series. The heuristic and CUSUM methods generated comparable results, although rate of false alarm for the heuristic method was lower at the same detection probability