Radiometric detection and analysis of arcing faults

Rachel M. Harris, Martin D. Judd, Philip J. Moore, James Livie

Research output: Contribution to journalArticlepeer-review

22 Citations (Scopus)

Abstract

A novel method for long-range detection and potential location of arcing faults on electrical distribution networks is described and investigated. When a network fault induces an electrical arc it is accompanied by the radiation of electromagnetic transients that can be detected remotely as a radio frequency signal. By capturing its wavefront at more than one monitoring location the time-difference-of-arrival between monitoring station pairs can be used to establish the origin. This approach is used to locate lightning arcs where the emissions are referred to as sferics. A key advantage of detecting faults by capturing the radiometric emissions from their arcs is that no direct connection is required to the plant being monitored, eliminating the need to take it out of service for either installation or maintenance of the radiometric monitoring equipment. This approach also has the potential for wide area coverage as opposed to monitoring a single specific circuit. In this paper, we describe the development and trialing of a monitoring network comprising 4 geographically separate receiving stations with an inter-station distance of less than 20 km. The utilization of GPS steered time-stamps overcame the fundamental challenge of providing a common, accurate, synchronized time-stamp for the transient signals received at each station. As well as providing the basis for fault location, this information allowed signals to be correlated with the network fault records. The system was proven to capture emissions from fault-induced arcs, but none of the captures were made at more than a single monitoring station, hence the efficacy of the system at locating fault-induced arcs could not be evaluated. However, lightning strikes were captured simultaneously at several monitoring stations and successfully located; demonstrating that the data capture methodology does allow signal origins to be determined. Further study is needed regarding the propagation distances of fault induced radiometric emissions, particularly their rate of attenuation with distance.

Original languageEnglish
Article number7116350
Pages (from-to)1547-1558
Number of pages12
JournalIEEE Transactions on Dielectrics and Electrical Insulation
Volume22
Issue number3
DOIs
Publication statusPublished - 1 Jun 2015

Keywords

  • distribution networks
  • fault location
  • radiometry
  • sferics
  • time difference of arrival

Fingerprint

Dive into the research topics of 'Radiometric detection and analysis of arcing faults'. Together they form a unique fingerprint.

Cite this