Acoustic emissions caused by the corona discharge in an oil-tank

Measurements of sonic and ultrasonic waves caused by corona discharges were performed based on an acoustic emission (AE) technique, to diagnose the deterioration of insulation and its location in an oil filled pole transformer. The detected signals in these measurements were analyzed by means of a fast Fourier transform (FFT) and their properties were discussed. It was found that the AE signals due to the Lamb waves caused in a thin steel tank can be easily recognized by the FFT analyses of initial stages of the detected AE signals. This indicates that the accuracy on location of the AE sources in the transformer can be improved. Also, it was shown that a low pass filter was effective to improve the accuracy of location.