Reliability of Infrared Thermography in Detecting Leaks in Buried Water Reticulation Pipes

Peter M. Bach, Jayantha K. Kodikara

Research output: Contribution to journalArticleResearchpeer-review

13 Citations (Scopus)

Abstract

The failure of pipelines in water distribution networks results in a significant volume of pristine quality water being lost. Failing infrastructure is often identified through different condition monitoring methods and then maintained. Increasing pressures on a rapidly growing water supply network call for more efficient leak detection techniques. Infrared (IR) thermography has received little attention in the literature, with little effort spent on understanding and assessing its feasibility, pertaining to the detection of buried pipe leaks. The present study develops and tests a passive leak detection methodology using IR thermography (informed by scientific literature) for small diameter buried reticulation pipes. The method was assessed in the field at 27 sites comprising water leaks beneath soil and grass nature strips. Results showed that the passive leak detection method by means of IR thermography was able to clearly identify 59% of leaks, with an additional 22% of sites having notable, yet inconclusive, thermal signatures. Based on image interpretation of field data, the authors propose a set of four types of surface thermal signatures likely to be detected from leaking water reticulation pipes, and a conceptual representation of subsurface processes that would produce these signatures. The study creates opportunities for improving the practicality of IR thermography as a leak detection tool and its combination with other techniques for better automation and data fusion.

Original languageEnglish
Pages (from-to)4210-4224
Number of pages15
JournalIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Volume10
Issue number9
DOIs
Publication statusPublished - 1 Sep 2017

Keywords

  • Engineering
  • Leak detection
  • Pipelines
  • Water resources

Cite this