Towards practical pressure-based leakage characterisation of water distribution pipes with a novel pipe condition assessment device

• Main Author: Tanyanyiwa, Craig Tinashe
• Other Authors: Beushausen, Hans-Dieter
• Format: Dissertation
• Language: English
• Published: Faculty of Engineering and the Built Environment 2020
• Subjects: Engineering
• Online Access: https://hdl.handle.net/11427/31785

Leakage detection and management have been proposed as effective ways of mitigating and managing water losses in an age where water scarcity has become prevalent. To this end, several methods have been developed and suggested with different benefits and drawbacks The presently available leakage detection methods, however, fail to identify and characterise the leakage while simultaneously assessing the condition of the water distribution network (WDN). This function is imperative for understanding and addressing leakage. WDN assessments are also important as knowledge of the network parameters helps in reducing water losses through planned infrastructure maintenance programmes. A pipe condition assessment device (PCAD) was thus developed which can detect, characterise leakage and assess the condition of the WDN. However, the efficacy and reliability of this device had not yet been established. In this study, the device was used to characterise leakage and assess system conditions in water networks. Initially, laboratory tests on six known leak types were conducted on a standardised laboratory setup. The leakage characteristics of these pipes were found through regression analysis. The results from the tests established that to 95% level of confidence; the standardised setup can produce repeatable and comparable results to previous studies. The accuracy of the PCAD instrumentation was verified and the device calibrated, the same pipes were then tested on the standardised setup using the PCAD. An overlap of the results from the laboratory experiments and the PCAD revealed that to 95% level of confidence, the device could adequately characterise leakage in pipes. A low variance of less than 4% of the mean parameter, across all tests conducted using the PCAD, informed that the results obtained through using the PCAD are repeatable and reliable. Field tests in the Kensington DMA were done and revealed the limitations of the device, such as its inability to characterise leakage in pipe sections that cannot be successfully isolated. However, in pipelines that were successfully isolated, the PCAD was able to detect and identify leakage characteristics in water networks and aid in conducting maintenance runs. Consequently, this study contributed to the body of knowledge by statistically establishing that the PCAD could adequately, and reliability characterise leakage in real water distribution networks.