Advanced Insulation Diagnostic by Dielectric Spectroscopy (Part II)


Electrical Insulation Test Methods

Capacitance, Dissipation or Power Factor Measurement at Mains Frequency

The dissipation or power factor at one single frequency point has been used for decades to classify insulation materials. Standards give various limits for power and dissipation factor. For example [6] states that in case of new oil-filled transformers and reactors, the power factor should not exceed 0.005. It further recommends for most older transformers a power factor of < 0.005, power factors between 0.005 and 0.01 may be acceptable; power factors > 0.01 should be investigated.

The first instrument to measure an unknown capacitance and its dissipation factor was the Schering Bridge. This is basically a four-arm alternating-current (AC) bridge circuit whose measurement depends on balancing the loads on its arms. For easier balancing of the bridge a high measurement voltage of some kV was required and for practical reasons the frequency was mostly limited to power frequency. These historical conjunctures found its way into standards and field test practices, where a test voltage of typically 10 kV and a limited frequency range close to power frequency are used.

Application for Power Transformers. Figure 7 depicts the dissipation factor for a new, a moderately aged and a heavily aged transformer at similar temperatures around 25°C. As it can be seen, the dissipation factor at power frequency (50-60 Hz) will reflect the cellulose insulation for the new transformer and the condition of the oil for the moderately aged and the aged transformer. A dissipation factor measurement at 0.1 Hz reflects the oil for the new and the moderately aged transformer while it shows the cellulose for the heavily aged transformer. Conclusively, dissipation factor measurements at one single frequency enable only for a limited assessment of the insulation condition, a discrimination of the various effects is impossible.

he so ton hao o 3 mba
Figure 7:Dissipation factor of three transformers in different aging conditions

Voltage Sweep – Tip Up Test. Here the test voltage is increased up to several kV and the dissipation factor recorded. According to the common interpretation, the magnitude of change indicates aging of the insulation. Looking at Figure 7, the voltage sweep will stress different materials depending on the condition of the dielectric. For the depicted new transformer, the voltage sweep at power frequency will test the pressboard; resulting in a low voltage dependence of the dissipation factor. For the transformer in “moderate” condition, a mixture of oil and solid insulation will be stressed; the voltage dependence will be more distinct. Finally, the “aged” transformer will show the strongest voltage dependence of dissipation factor since here the oil area is exposed to the electric field.

Another issue related to voltage sweeps is the dimension of the tested insulation. It is not the voltage itself that causes the effects, but the field strength. The field strength depends on the geometry of the tested insulation (gap between conductors), which is unknown to the tester. For a narrow gap, the stress will be high but low for a large gap. Therefore it would be more meaningful to make a field strength sweep. Since the geometric conditions of the insulation are almost always unavailable, the voltage sweep is of limited benefit for transformer insulations. For bushings, on the other hand, a high voltage sweep can indicate partial breakdowns and a defective connection to the outer capacitive layer.

Dielectric Absorption Ratio and Polarization Index

The dielectric absorption ratio DAR relates the DC resistance measured at 60 s to that at 15 s; the polarization index is the ratio at 10 min to that of 1 min. Figure 8 (left) depicts the dielectric response of two new transformers in frequency domain. For transformer B, the polarization index mirrors the insulation geometry whereas for transformer A it shows the influence of the oil. The polarization index of transformer B is worse than that of A. On the other hand, the moisture content of both transformers was identical with 0.4 %. The differences in the dielectric response arise from different oil qualities, not from the insulation dryness.

ti so hap thu dien moi DAR
Figure 8:
Dielectric absorption ratio DAR and polarization index PI with the dielectric response of two new transformers in frequency domain (left) and with the influence of an oil exchange in time domain (right)

A second test led to similar results. For the large insulation model as described at Figure 6 and [3], the original oil having a conductivity of 1.6 pS/m was exchanged by aged oil with

16.5 pS/m. The moisture content of the cellulose insulation remained unchanged at 1.1 %. Figure 8 (right) shows the polarization and depolarization currents in both conditions and also the dielectric absorption ratio and the polarization index. Due to the oil exchange, the DAR increased from 1.25 to 4.43 and the PI from 3.69 to 6.34 although the moisture content remained identical. The increased oil conductivity caused this dramatic change. As a higher DAR and PI are considered to be better, these parameters actually call the transformer filled with bad oil better than one with new oil.

Both values, PI and DAR, are defined for the narrow time range which is dominated by oil conductivity and are unable to discriminate between the influence of the different materials and the interfacial polarization effect, which is similar to conventional dissipation factor tests.

Methods Applying a Frequency Sweep

Frequency Sweep Close to Power Frequency. Newer diagnostic methods involve a frequency sweep around power frequency for gathering more information about the insulation condition. Figure 9 (left) shows the measurement circuit of a recent dissipation factor bridge. In contrast to the traditional Schering Bridge the currents are measured directly, no balancing path is used and the bridge is frequency independent. The diagram of Figure 9 (right) depicts the dissipation factor of the HV-LV capacitance CHL, the HV-tank capacitance CH and the LV- tank capacitance CL of a new transformer. On the background of the interpretation scheme of Figure 3 this measurement mainly reflects the pressboard. Towards the low frequencies also some influence of the oil becomes visible.

mach do
Figure 9:
Measurement circuit of a recent dissipation factor bridge (left) and dissipation factor of CHL, CH and CL of a new power transformer (right)

Such combination of frequency sweep with high measurement voltage unites the advantages of frequency sweep and measurement voltage; however the narrow frequency range limits the validity.

Dielectric Spectroscopy – Dielectric Response Measurement

Dielectric spectroscopy conducts the properties of insulation systems across a wide frequency range of e.g. 1000 Hz to 0.0001 Hz. This frequency span over 7 decades enables for discrimination between different effects and finally allows for the determination of moisture in the solid insulation. Measurements in time and frequency domain have been used for one decade. Today, a new device combines the advantages of both measurement domains in order to have a very fast measurement particularly of the low frequencies, Figure 10. An advanced analysis algorithm not only calculates the moisture content but also compensates for other conductive aging by-products [7].

phep do mien thoi gian va tan so
Figure 10:
Combination of time and frequency domain measurements (left) and required time duration and acquired frequency range for the different measurement techniques (right)

According to the current state of science, dielectric spectroscopy allows for the most comprehensive diagnostic and evaluation of insulation systems.


 Maik Koch leads the product management for HV testing equipment of Omicron Energy, Austria. He graduated as a Doctor of Philosophy at the University of Stuttgart in Germany in 2008. His fields of research are ageing and moisture determination in power transformers using chemical and dielectric analysis methods. He collaborates in working groups of Cigrè and IEEE.

Michael Krueger is head of the OMICRON competence center primary testing. He studied electrical engineering at the University of Aachen (RWTH) and the University of Kaiserslautern (Germany) and graduated in 1976 (Dipl.-Ing.). In 1990 he received the Dr. techn. from the University of Vienna. Michael Krueger has 30 years experience in high voltage engineering and insulation diagnosis. He is member of VDE, Cigré and IEEE.

Markus Puetter is product manager for primary testing solutions at OMICRON energy, for the CPC100 product family and the dissipation / power factor measuring system MI600. He has 10 years experience in developing primary testing solutions for high voltage equipment. He studied electrical engineering at the University of Paderborn / Soest and graduated in 1976 (Dipl.-Ing.).


[1] “Company Provides Transformer Monitoring for Aging Equipment” Transmission and Distribution World 2007, Prism Business Media, at March 5, 2007

[2] Gernandt, G. Balzer, C. Neumann: „Auswertung von Störungen und Gas-in- Ölanalysen bei Hochspannungs-Transformatoren“, ETG Fachbericht 104, Fachtagung, Diagnostik elektrischer Betriebsmittel“, Kassel 2006

[3] M. Koch, S. Tenbohlen, M. Krüger and A. Kraetge: “A Comparative Test and Consequent Improvements on Dielectric Response Methods” Proceedings of the XVth International Symposium on High Voltage Engineering, ISH, Ljubljana, Slovenia, 2007

[4] C. G. Garton: “Dielectric Loss in Thin Films of Insulating Liquids”, Proceeding Institution Electrical Engineering, Vol.88 (1941) p.103-120

[5] Koch: “Reliable Moisture Determination in Power Transformers”, PhD thesis, Institute of Energy Transmission and High Voltage Engineering, University of Stuttgart, Sierke Verlag Göttingen, Germany, 2008.

[6] IEEE Standard C57.152: “IEEE Guide for Diagnostic FieldTesting of Electric Power Apparatus-Part 1: Fluid Filled Power Transformers, Regulators, and Reactors”, IEEE Std 62-1995

[7] Koch, M. Krüger: “Moisture Determination by Improved On-Site Diagnostics”, TechCon Asia Pacific, Sydney 2008


This entry was posted in . Bookmark the permalink.
024 6683 0230

Send us
an Email