Thesis

UHF monitoring of partial discharges in high voltage transformers

Creator
Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2006
Thesis identifier
  • T11888
Qualification Level
Qualification Name
Department, School or Faculty
Abstract
  • The thesis considers using measurements of Ultra High Frequency (UHF) signals from Partial Discharges (PD) to detect and characterise defect signals in multi-source environments. PD is the result of partial bridging of the insulation medium between electrodes, and can be a precursor to eventual flashover. UHF monitoring techniques became predominant in monitoring Sulphur Hexafluoride (SF6) filled Gas Insulated Substations (GIS). There are some particular characteristics of a PD that make monitoring in this range of signals advantageous, such as fast propagation velocities and low attenuation. These attributes make the technique very suitable to determining the type of PD source and location of the defect with accuracy. The research presented in the thesis uses three UHF sensors to detect PD signals from defects. By simultaneously capturing the UHF signals on three channels of a high bandwidth oscilloscope the time of arrival and signal strengths at the three sensors can be used to form Partial Discharge Cluster Maps (PDCMs), which are one of the main contributions of this research. These enable multiple defect sources to be identified and the defect location to be accurately determined. The application that will be considered in most detail is the use of UHF techniques to characterise defects in High Voltage (HV) oil-filled transformers. This is a relatively new area of research and the thesis discusses the challenges the complex internal structure of a transformer presents. The capability to locate and characterise multisources is shown to be successful. An on-line system was developed for plant assessment. Using this system, the technique was successfully tested on operational transformers and these results are also presented in the thesis.
Resource Type
DOI
EThOS ID
  • uk.bl.ethos.443122
Date Created
  • 2006
Former identifier
  • 760311

Relations

Items