University of Sheffield
An Acoustic Velocity Vector Based Approach to Robotic Pipeline Inspection
Abstract
dc:description.abstractWater loss due to damaged, pressurised pipelines is a significant and ongoing problem in the UK and across most of the world. As of 2022, over 20% of treated water in the UK was lost due to leaks, barely improving on loss rates in 1995. This thesis aims to diagnose some of the causes of current issues with leak detection and suggest an improved methodology. A comprehensive review identified several areas within the leak detection sphere worth investigation: acoustic propagation and attenuation in plastic pipes, the variability in leak-generated noise, and moving beyond monoaxial sensing, given that this captures only a fraction of the available acoustic information. Motivated by these findings, experimental and numerical studies were undertaken considering the propagation and attenuation of sound in plastic pipes and examining the influence of features such as joints, corners, and burial conditions. Results showed that attenuation was strongly frequency dependent, with higher frequencies attenuating considerably over short distances, limiting the viability of traditional correlation-based leak localisation methods for plastic networks. To address these challenges, this work explores the novel use of acoustic velocity vector sensors (AVVSs) on an in-pipe platform for defect detection. Numerical models demonstrated that radial components of the acoustic velocity are particularly sensitive to small wall defects, while the pressure and axial velocity showed no deviation from background levels close to the defects. This prediction was confirmed experimentally using an AVVS on a test pipe. A robotic platform carrying multiple AVVSs further validated the feasibility of this approach. This research establishes acoustic vector sensing as a promising new diagnostic method for pressurised water pipes, allowing smaller and incipient leaks to be detected and laying the groundwork for intelligent, mobile in-pipe monitoring systems that can enhance predictive and preventative maintenance of critical infrastructure.
Degree
thesis:*- Name dc:type.qualificationname
- Ph.D
- Level dc:type.qualificationlevel
- doctoral
- Grantor dc:publisher.institution
- University of Sheffield
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Watts, Joanna
- Advisors dc:contributor.advisor
-
- Horoshenkov, Kirill
- Krynkin, Anton
- Tait, Simon
- ORCID person:identifier.orcid
Rights
- Language dc:language.dcterms_iso639_2
- eng
Identifiers
dc:identifier.*- Dc Identifier Grantnumber
- EP/S016813/1