Real-world Comparison of Sensor Placement Algorithms for Leakage Localization

Daniela Fuchs-Hanusch, David Steffelbauer

Research output: Contribution to journalConference articlepeer-review

Abstract

In this paper we compare 6 algorithms for optimal pressure sensor placement by applying them to a real world system. The targets of these comparisons were a) to see how robust sensor positions are in terms of different mathematical approaches b) to derive the optimal sensor placement for the purpose of small leakage localisation in a case study distribution system. Therefore, we created artificial leakage scenarios in the real world system by opening fire hydrants at different flow rates. Pressure was measured at all hydrants that were found to be an ideal sensor position by any of the applied methods. Based on these measurements, model-based leakage localization was run using data from the different sensor sets. To compare leakage localization performance the distance between the found to the real world leak was calculated. The results of the sensor placement have shown similarities in sensor positions for graph theory based methods, but the ideal sensor positions derived from the sensitivity based methods varied strongly. The localisation results have shown that leaks down to 0.5 l/s can be located but for different leak positions different sensor sets, mainly those with sensors close to the leak position, led to the best performance.

Original languageEnglish
Pages (from-to)499-505
Number of pages7
JournalProcedia Engineering
Volume186
DOIs
Publication statusPublished - 1 Jan 2017
Event18th International Conference on Water Distribution Systems, WDSA 2016 - Cartagena, Colombia
Duration: 24 Jul 201628 Jul 2016

Keywords

  • hydraulic model
  • leak detection
  • pressure measurements
  • water loss reduction

ASJC Scopus subject areas

  • General Engineering

Fingerprint

Dive into the research topics of 'Real-world Comparison of Sensor Placement Algorithms for Leakage Localization'. Together they form a unique fingerprint.

Cite this