Extremely Sensitive Anomaly Detection in Pipe Networks Using a Higher-Order Paired-Impulse Response Function with a Correlator

Abstract

Detection of anomalies in pipe networks (leaks, blockages, and wall deterioration) is critical for targeted pipe section replacement and maintenance in water distribution systems. A hydraulic signal-processing approach, termed the paired-impulse response function (paired-IRF), has been previously proposed for anomaly detection by transforming the persistent principal wave reflections by anomalies into distinctive paired spikes. In this paper, a new higher-order paired-IRF has been derived, which considers both principal and higher-order wave reflections by the anomalies. A correlator has then been designed (and incorporated into the higher-order paired-IRF) to highlight anomaly-induced spikes and suppress noise. A looped pipe network with realistic background noise was assembled in the laboratory to examine the efficacy of the new methods. According to the experimental results, it is observed that (1) the higher-order paired-IRF is an extremely sensitive detection technique and clearly identifies anomalies inducing wave reflections as small as 0.5% of the injected wave magnitude; (2) its sensitivity is sufficiently accurate when using micropressure waves as small as 20 mm in magnitude and contaminated by 2-m background pressure fluctuations; and (3) the proposed advanced correlator highlights the anomaly-induced spikes in the paired-IRF trace in a noisy environment.