Geolocalization of water-waves origin within water distribution networks using time reversal of first event detection. (15th February 2023)

• Record Type: Journal Article
• Authors: Guibert, R.; Bayle, A.; Plouraboué, F.
• Is Part Of: Water research. Volume 230(2023)
• Publication Date: 2023-02-15
• Subjects: 00-01 -- 99-00; Water distribution network -- Water-hammer -- High-frequency detector -- Geolocalization; Water -- Pollution -- Research -- Periodicals; 363.7394
• DOI: 10.1016/j.watres.2022.119538
• Languages: English
• ISSNs: 0043-1354

Abstract

Abstract: Drinking water distribution networks in urban areas are daily subject to fast propagating pressure waves resulting from routine operations. These water-hammer waves lead to structural aging and facility damages, the origin of which is not easy to find but are sometimes of high managerial interest. In this contribution, we demonstrate that using a reasonable number of high-frequency pressure detectors distributed within the network combined with a proper post-processing method permits a close geolocalization of the damaging wave origin. The method is first tested and validated on a real water distribution network having approximately 26 000 pipes, whereas considering a known, prescribed waveorigin, so that the sensitivity to sensor number (sensor spatial density), sensor location and signal-to-noise ratio on the geolocalization robustness are analyzed in detail. It is then applied and illustrated over real sensor recordings the result of which are validated on the field from history matching. This paper thus presents the first field-scale geolocalization of water-hammer events origin test as well conditions for which, given sensor density and signal-to-noise ratio, the geolocalization success is to be expected. Graphical abstract: Highlights: Geo-localization of water-hammer waves origins on water distribution networks. Network method using a limited number of high-frequency pressure sensors. Sensitivity study of the method to sensor number, locations and signal/noise ratio. Application and history matching on distribution network of a large city.