Condition Evaluation Method Of Transformer Oil-Paper Insulation Based On Fusion Of PDC And RVM

It has become concerned and urgent technical problems of power industries andrelated research departments to improve operating safety level of power transformers,especially the ones which have been on service for more than20years. ResponseRecovery Voltage (RVM) method and Polarization and Depolarization Current (PDC)method based on the principle of time-domain dielectric response have thecharacteristics of simple operation, convenient and time-saving measurement, richobtained information, so they are increasingly applied in the field of transformeroil-paper insulation condition assessment. Nevertheless, current researches on dielectricresponse in time domain are limited to applying a single time-domain measuringtechnique to assess the moisture content in insulation pressboard and conductivity ofinsulating oil in domestic and foreign, and reports about combination of twotime-domain measurement techniques to assess the insulation aging status are quiteinfrequent. Studying the combination of two time-domain dielectric responsemeasurement methods and the conversion technology between RVM and PDCmeasurement results can not only enrich the characteristic parameters, improve thediagnosis accuracy of the insulation condition, but also can avoid the error caused byrepeated testing, reduce the economic losses caused by the a long time outage oftransformer, which can greatly save resources.The results of PDC and RVM mesurement technology are connected by extendedDebye model according to circuit principle analysis, and time-domain analysis softwareis coded by Matlab. Experiments are designed to investigate the impact of aging degree,temperature and oil changing on time-domain characteristic parameters by measuringthe polarization and depolarizing current and calculating time-domain characteristicsparameters. Criterion of qualitative evaluating the aging status of transformer maininsulating is proposed using variation law of characteristic parameters in time domainunder different aging degree. And finally, validity of this method is verified by themeasured results of the main transformer model and the real transformer. The mainconclusions of the paper are as following:â‘ Combination of time-domain dielectric response measurement techniques(PDC and RVM) is achieved by extended Debye model of transformer oil-paperinsulation. RVM curves are derived and calculated by measured PDC data. Ten time-domain characteristic parameters including polarization spectrum, insulationresistance, and insulation conductivity are extracted. The Graphical User Interface oftime-domain dielectric response is designed to simplify the data processing and formdevice terminal for preparations of engineering application.â‘¡Thermal aging time of oil-paper insulation shows greater impact on the testingresults of PDC, by which time-domain characteristic parameters calculated can reflectthe degree of aging sensitively. Especially, when aging degree increases, IPS changegradually from a single peak to double. Thus, shape of IPS at the end of aging issignificantly different with that at initial.â‘¢Oil changing and measurement temperature have great influence on thetime-domain characteristic parameters and the shape of IPS is quite dissimilar fromeach other before and after the oil change. IPS of oil-paper insulation movesprogressively to the left and the charging time corresponding to the peak reduces withthe increasing of temperature, showing the double-peak characteristic.â‘£According to change law of time-domain characteristic parameters against withthe aging degree, the qualitative assessing criteria of transformer oil-paper insulationaging status based on time-domain dielectric response is suggested, and the qualitativeassessment of the main insulation system in transformer was realized preliminary.