Concepedia

TLDR

Water distribution systems suffer from leaks that waste treated freshwater, worsening the imbalance between demand and supply. The study proposes a leak‑detection method that continuously monitors changes in the correlation of surface acceleration measured at discrete pipeline locations. A leak detection index based on cross‑spectral density of pipe surface accelerations is formulated, requiring minimal human intervention and operating under normal pipeline conditions, and is demonstrated on a 76 mm PVC test system with varying leak severities. Preliminary results are promising, suggesting the method works but warranting further investigation.

Abstract

Exacerbating the imbalance between demand for freshwater and available water resources is the sub-optimal performance of water distribution systems, which are plagued with leaks that cause significant losses of treated freshwater. This paper presents an approach for leak detection that involves continuous monitoring of the changes in the correlation between surface acceleration measured at discrete locations along the pipeline length. A metric called leak detection index is formulated based on cross-spectral density of measured pipe surface accelerations for detecting the onset and assessing the severity of leaks. The proposed non-invasive approach requires minimal human intervention and works under normal operating conditions of the pipeline system without causing any operational disturbances. The approach is demonstrated on a 76 mm diameter polyvinyl chloride pipeline test system considering varying leak severities. The preliminary results presented in this paper seem promising and lead to several interesting questions that will require further research.

References

YearCitations

Page 1