Understanding and analysis on frequency dielectric parameter for quantitative diagnosis of moisture content in paper-oil insulation system

Abstract

The ageing rate of cellulose insulation materials (paper/pressboard) in oil-impregnated transformers is largely determined by moisture. In this paper, several experiments were conducted under controlled laboratory conditions to quantitatively diagnose the amount of moisture in cellulose insulation using frequency domain spectroscopy (FDS) technique, which has been proven to be an effective tool to predict the insulation condition in oil-impregnated power equipment. The frequency dielectric response in oil-impregnated pressboards, moisture content as well as test temperature, was monitored carefully, and a new dielectric characteristic parameter for the diagnosis of moisture content was proposed. In addition, the master curve technique was introduced to calibrate the temperature effect on FDS measurement results. These experiments revealed that a strong correlation exists between frequency dielectric parameter and moisture content in cellulose insulation. Such correlation can be utilised for better prediction of moisture content in paper-oil insulation. A modified assessment method for the diagnosis of moisture content in paper-oil insulation was investigated and its effectiveness was also demonstrated using a case.