Acoustic feature-based leakage event detection in near real-time for large-scale water distribution networks. Issue 2 (31st January 2023)

• Record Type: Journal Article
• Authors: Chew, Alvin Wei Ze; Kalfarisi, Rony; Meng, Xue; Pok, Jocelyn; Wu, Zheng Yi; Cai, Jianping
• Is Part Of: Journal of hydroinformatics. Volume 25:Issue 2(2023)
• Page Start: 526
• Page End: 551
• Publication Date: 2023-01-31
• Subjects: acoustic signals -- data analytics -- leakage detection -- near real-time analysis -- smart water grid; Hydrology -- Data processing -- Periodicals; Geographic information systems -- Periodicals; Geographic information systems; Hydrology -- Data processing; Electronic journals; Periodicals; 551.480285
• DOI: 10.2166/hydro.2023.192
• Languages: English
• ISSNs: 1464-7141

Abstract

Abstract: Acoustic sensors are widely deployed to detect hidden leakages in water distribution networks (WDNs). However, few studies have been conducted to quantitatively understand the dominant leakage acoustic characteristics, which are usually mixed with unknown environmental noises, coupled with the constraint of sparse deployment of acoustic sensors. In this paper, a comprehensive approach, that performs acoustic data feature analysis, is developed to detect pipe leakages in near real-time via a series of systematic analyses, namely: (1) data quality assessment; (2) features identifications; (3) outlier detection and event classification; and finally (4) near real-time leakage detection. The proposed solution has been tested on two major WDNs in Singapore having around 1, 000 km of water pipelines installed with 74 permanently installed hydrophone sensors. The leakage detection results obtained from our case study demonstrate that the dominant leakage acoustic characteristics can be captured in lower intrinsic mode functions (IMFs), to within the frequency range of 100–750 Hz approximately, by decomposing the original acoustic signal. Systemwide leakage event detection and classification models are subsequently trained and tested on acoustic datasets collected over 13 historical months, where more than 70% F1-scores can be obtained from the emulated near real-time leakage detection analysis. HIGHLIGHTS: Developed acoustic feature-based methodology for near real-time leakage event detection and classification. Verified methodology with 74 permanently installed hydrophones in two water supply zones having 1, 000 km water pipelines in Singapore. Achieved F1-score of >70% for leakage event detection and classification analyses on imbalanced acoustic datasets collected over 13 months.