Electronic International Standard Serial Number (EISSN)
In this paper, a multichannel instrumentation system for the location of partial discharges (PDs) in power transformers is presented. It is based on the detection of the acoustic emissions from PDs in oil with several piezoelectric (PZT) and fiber optic sensors. An acoustic detection and location approach is proposed based on a time reference given by one fiber optic sensor installed inside the tank and the times of arrival to several PZT sensors installed outside the tank (two in front of each phase windings of a three phase transformer for a typical application). The signal processing includes digital denoising techniques and time-difference lookup table-based 3-D location algorithms; these algorithms have been implemented with virtual instrumentation. The system is tested in an acoustic experimental setup and the location accuracy is evaluated. Finally, the error propagation from the times of arrival and the influence of the number of sensors and their sites in the 3-D location algorithm were analyzed.