Insulation analysis of frequency-dependent dissipation factor in high voltage model transformer

It is known that power transformers are of great importance in today's energy world. These devices are quite expensive and require constant maintenance. The most maintenance-requiring zone is insulation systems. Oily paper insulation systems can be breakdown even during normal operation. If these failures cannot be noticed in time, they can lead to disaster, such as fire, as well as unexpected failures and unplanned power outages. There are many old transformers still in use today. If these transformers are capable of working, it is important to extend their life and ensure that they continue to work. The way to do this is to increase the insulation quality of the transformer. If we can learn the isolation status of the transformer, we can have better information about the life of the transformer. That's why researchers found some methods to learn the insulation status of transformers (RVM, PDC). These methods have some advantages and disadvantages compared to each other, and research has been continued. Finally, FDS (frequency domain spectroscopy) method was developed. The FDS method has advantages over the other two methods. In this study, the transformer windings, transformer protection, transformer failures and the reasons of transformer failures are mentioned. Some protection devices have been introduced. In addition, the advantages and disadvantages of the aforementioned methods are mentioned comparatively. A pancake model transformer with different oil ratios has been developed. The oil content of this pancake transformer ranges from 85% to 0%. The dissipation factor and power factor measurements were taken in Pancake test transformer by using FDS method. These measurements were analyzed. On pancake test transformer under constant voltage, loss factor, power factor, imaginary capacity and real capacity measurements were taken at different frequencies. These values for DG-CH are power factor 0.0053475, dissipation factor 0.0053476, real capacity 2.8319E-9 and imaginary capacity is 1.5138E-11. Values for E-DG are power factor 0.005547, the dissipation factor is 0.0055471, the real capacity is 2.4056E-9 and the imaginary capacity is 1.3344E-11. Values for F-E are power factor 0.0044075, the loss factor is 0.0044075, the real capacity is 4.8008E-9 and the imaginary capacity is 2.1159E-11. Loss factor, power factor, imaginary and real capacity changes were analyzed by frequency. In addition, the effects of moisture, temperature and aging on dissipation factor and power factor were evaluated with auxiliary parameters. The importance of frequency magnitude in evaluating the dissipation factor is emphasized. At the same time, the relationship between dissipation factor and power factor with the help of equivalent circuit and vector diagram is mentioned. With the dissipation factor tan δ, the conductivity of oil in the transformers, the moisture and the aging of the insulating paper can be determined. In transformers with power factor cos θ, the core movement between the transformer tank and the core, and the change of the distance between the windings can be determined. The behavior of these two cases at low, medium and high frequencies should be known. It is very important to understand these two cases in order to very good know the FDS method. In the studies done, it is possible to know exact cause of the defect in the insulation material(environmental conditions or insulation deterioration) by FDS. The results of the experiment showed that the effect of temperature and aging on the insulating material affected FDS results at low frequencies. In studies performed at low frequencies, dissipation factor, power factor, imaginary capacity and real capacity values increased with the measurement temperature and aging times. In addition, as the oil content of the transformer insulation decreases, the values of imaginary capacity, real capacity and dissipation factor decrease. By applying different voltages (50 V, 75 V, 100 V, 140 V), imaginary, real capacity and dissipation factor changes were investigated. It has been observed that the capacity curves have almost the same curves under different voltages. In the dissipation factor, the peak values of the curves were at higher frequencies as the voltage increased. Under the same connection point and equal voltages (140 V), different winding locations are measured and plotted. It can be said that the dissipation factor, capacity and power factor curves for CL and CH are similar. The dissipation factor and power factor reached its peak at 0.1 Hz. Dissipation factors, capacity and power factor curves for C1H1, C1H2, C1H3 and C2H1, C2H2, C2H3 curves are similar. However, the dissipation factor, capacity and power factor curves for CHL differ slightly from other conditions. In CHL, the loss factor and power factor peak at approximately 0.1 Hz and the capacity at 0.001 Hz. Power factor, capacity and loss factor curves in all connection forms (CH-B, DG-CH, E-DG and FE) and all winding locations (CHL, CH, CL, C1H1, C1H2, C1H3, C2H1, C2H2 and C2H3) were found to be very close to each other in the high frequency range. However, these values differed at low frequencies. So we can say that low frequencies that are decisive for us. The dissipation factor of CHL for a frequency of 0.1 Hz in CH-B mode is approximately 2.5 times greater than that of CH and approximately 7 times greater than that of CL. On the frequency of 60 Hz, the dissipation factor of CHL is approximately 1.5 times higher than that of CH and approximately 2 times higher than that of CL. At the frequency of 1000 Hz, the rates are quite variable. The dissipation factor of the DG-CH connection for the frequency of 0.1 Hz in CH-L is about 2 times the dissipation factor of the E-DG connection and about 9 times the dissipation factor of the F-E connection. At 60 Hz, the dissipation factor values are very close to each other, but we see that the biggest E-DG connection and the smallest DG-CH connection. While the loss factor has the highest value at a frequency of 0.05 Hz, it is smaller at lower frequencies. While the dissipation factor of CL and CH are close to each other, the dissipation factor of CHL are about 1.5 times smaller than that of CL and CH. The dissipation factor of the DG-CH connection at 0.1 Hz is about 2 times the dissipation factor of the E-DG connection and about 5 times the dissipation factor of the F-E connection. Although real capacities are close to each other, the real capacity of the DG-CH connection is the lowest, while the real capacity of the F-E connection is the highest. In the imaginary capacities, the DG-CH connection is the highest and the F-E connection has the lowest values. Although the values at 60 Hz were very close to each other, the highest dissipation factor value was found in E-DG and the lowest dissipation factor value was observed in F-E connection. The highest real capacity value was found in the F-E connection, the lowest DG-CH connection. The highest imaginary capacity value is found in the DG-CH connection and the lowest in the F-E connection. At 1000 Hz, the dissipation factors of the DG-CH and F-E connections are almost equal, and the dissipation factor of the E-DG connection is approximately 5/4 about of the other two connections. In the highest real capacity value F-E connection, the lowest real capacity value was seen in the DG-CH connection. The highest imaginary capacity value is seen in the F-E connection and the lowest imaginary capacity value is seen in the DG-CH connection. Today, transmission and production at low frequencies are preferred because of the low losses. In particular, a more detailed analysis of the low frequency experiments is necessary since the diagnostic tests are performed at low frequencies such as 0.1 Hz. Changes in capacity and dissipation factor inform us about the type of error. For example, short circuit turns and changing of magnetizing are seen at the low frequency response while the medium frequency response is sensitive to hoop buckling and axial movement faults in the transformers. The high frequency response is sensitive to the part of the properties of windings and grounding problems. For this reason, the capacitance and dissipation analysis according to frequency are very important for detecting faults in transformers. Frequency domain spectroscopy (FDS) is less noisy than other insulation diagnostic methods (RVM, PDC) and is independent of the shape of the transformer. It is also little affected method by external factors. Therefore, it can be said that FDS is the most effective diagnosis method.