Study On Condition Assessment Of Oil-Paper Insulation Based On Gray Relational Analysis

Oil-immersed power transformers as part of the core of power transmission equipment,play a decisive role in the power system.The overall life of an oil-immersed transformer depends primarily on the aging of its internal oil-paper insulation.Oil paper insulation will be affected by electricity,heat,mechanical stress in operation life.The dielectric properties of oil-paper insulation in the aging process vary a lot and can be affected by degree of aging and moisture content.It's important to solve the problem that finding time domain dielectric characteristics of insulation oil and paper separately based on gray correlation analysis.This paper tested the polarization/depolarization current of oil-paper insulation samples with different aging days.Then steady oil aging charge quantity and steady oil-paper interfacial polarization charge quantity were performed as aging characteristics of transformer oil and oil-paper interfacial polarization.And extracted peak value and peak time of time domain dielectric spectroscopy as aging characteristics of insulation paper.Finally verified the feasibility of the above described aging characteristics by field test.Steady oil aging charge quantity increases with increasing of samples aging days and there is a linear fitting relationship between them.Steady oil aging charge quantity can represent strength of dipole polarization in oil.Peak value and peak time of time domain dielectric spectroscopy can represent the strength and response speed of interfacial polarization in paper caused by cellulose cracking.There is exponential negative correlation between Peak value of time domain dielectric spectroscopy and degree of polymerization.Steady oil-paper interfacial polarization charge quantity increases with increasing of sample aging days and there is a linear fit relationship between them.Steady oil-paper interfacial polarization charge quantity can represent strength of oil-paper interface polarization.Oil-paper insulation samples with different thermal aging and different moisture contents were prepared by accelerated thermal aging test and moisture absorption test.Then system pole of the largest time constant branch of oil-paper insulation is proposed as the state characteristic based on the analysis of polarization/depolarization current curve.The results show that the polarization and depolarization currents of the samples are shifted to the upper left and the initial value and steady value of the polarization current increase with the increase of water content.The increasing of water content also shorten stable time of polarization and depolarization currents.The system pole is so sensitive to the moisture content and the degree of aging that it can be used as the characteristic value of the oil-paper insulation state.But the system pole can't distinguish the influence of aging and moisture.This paper introduced the birth,development and current situation of gray relational analysis theory,and selects Deng's correlation degree as the most suitable algorithm of oil-paper insulation state assessment.Then introduced the precautions of dielectric test of field transformer and obtained the reference vector before and after normalization by summarizing the time domain dielectric characteristics of oil-paper insulation.Finally,the state of the transformer is evaluated by calculation of correlation degree.The assessment result by using the gray correlation analysis is consistent with the actual operating experience of the oil-paper insulation status of the transformer.The next step will enrich the test conditions such as add influence of test temperature or diminish the gradients of the moisture content and the degree of aging to eliminate the possible error of the test electrode and the type of cardboard.It's better to establish a complete database of state parameters embedded in software to obtain the evaluation results more accurately.