Application of Relaxation Distribution Function in Diagnostic of Power Transformer Insulation

In this paper the relaxation distribution function, which is obtained from measuring the polarization and depolarization current (PDC) on oil-pressboard samples, will be presented. The samples have been aged at 115 °C with Fe, Cu and Zn contents as catalyst in order to simulate the stress conditions. Four samples have been aged for 250, 500, 1000, 2000 hours. Tests have been made also on four moisturized samples. Depolarization current curves are decomposed to basic relaxation functions. The results obtained show classification of samples to dry and wet is possible. Also, the quality of oil can be determined from the relaxation distribution function. Aging state of the samples has an impact on the measuring result and the degree of aging of the insulation can be distinguished. INTRODUCTION Cellulose insulation is used as the main insulation system of power transformers. In more or less all utilities power transformers exist with over 40 years of use. Nowadays cellulose-aging detection becomes an important trend in transformer engineering due to the need of longer use of old transformers at a stress close to the nominal value. Cellulose is aged by thermal, electromechanical, chemical and dielectric stresses. Cellulose consists of glucose molecule chains. The length of molecule chains determines the mechanical strength of the pressboard or paper. Degree of Polymerization (DP) shows the length of molecule chains. During degradation processes of oil and cellulose, the DP is decreased. Water and 2-furfural are produced. Water act as a catalyst for the degradation process of cellulose, therefore 49-E-TRN-747 Application of Relaxation Distribution Function in Diagnostic of Power Transformer Insulation يمهدزون قرب يللملا نيب سنارفنك ن نارهت 1383 2 detection of water in cellulose is very important. Water content of oil and cellulose in transformers are in equilibrium after a certain time at constant temperature. Water content of oil can easily be measured, however only less than 1 % of water in transformers is stored in the oil [1]. Dielectric spectroscopy in time or frequency domain has been used for diagnostics purpose in power transformers, because dielectric characteristics of material are related to the molecular properties of the material. The dielectric response function is directly proportional to the depolarization current [2]. In several investigations [3,4,5], PDC has been measured for power transformer diagnostic reasons and can be used for the assessment of the transformer insulation. This paper presents the PDC analysis as a method for the detection of aging status of the insulation system of power transformers. REVIEW OF CONCEPT Figure 1 shows an extended Debye model for oil-pressboard insulation. R∞ is the insulation resistance and Cg the geometric capacitance of the electrodes. Each series circuit represents a relaxation process in the insulation, shown as R1-C1, R2-C2, ..., Rn-Cn.