Analysis of Relation between UHF Signals and Discharge Magnitude from Typical PD Defects in GIS

Abstract: The relation between UHF signals and discharge magnitude remains a difficult problem among partial discharge UHF researches. In order to analyze the relation between detection results by UHF method and parameters by impulse current method, the paper compares UHF signal energy, quadratic integral, peak voltage, peak-to-peak value and wave area with discharge quantity or square of discharge magnitude through four typical defect models of partial discharge. It is concluded that the highest coefficient of determination exists in the linear relation between UHF signal energy and square of discharge quantity. Besides, the best regression curve between UHF signal energy and discharge quantity is obtained. 

Key words: UHF method, impulse current method, GIS, PD, quantity of discharge, signal energy   

1. Introduction

The GIS is widely used in the present grid transformation and construction thanks to many advantages, including small floor area, stable and reliable operation, free from external influence and less maintenance. Although the reliability of GIS is high, inevitable internal defects can still cause the breakdown, and finally, major accident occur. There have been studies of partial discharge (PD) at home and abroad as it is one of vital factors, which cause GIS breakdown. The available mature detection methods are impulse current method and ultra-high frequency method (UHF).    

The  impulse current method is a common method for detecting partial discharge. The method is to measure the impulse current generated by partial discharge to detect corresponding impulse voltage at both ends of impedance. Its advantages are as follows: good transmission characteristics are kept within broad frequency range; the sensitivity is high and quantity of discharge can be calibrated. However, it has disadvantages. For instance, measuring frequency is low; it is easy to be interfered by the electromagnetism; it cannot be suitable for on-site monitoring.