Dielectric Frequency Response Measurements and

Dissipation Factor Temperature Dependence

Peter Werelius Jialu Cheng
Megger Megger
Täby, Sweden Täby, Sweden
Peter.werelius@megger.com Jialu.cheng@megger.com

Matz Ohlen Diego M. Robalino

Megger Megger
Täby, Sweden Dallas, USA
Matz.ohlen@megger.com Diego.robalino@megger.com

Abstract----- The condition of the insulation is an essential Transformers are by far, one of the most mission critical
aspect for the operational reliability of electrical power components in the electrical grid. The need for reliable
transformers, generators, cables and other high voltage diagnostic methods drives the world’s leading experts to
equipment. Transformers with high moisture content can not
evaluate new technologies that improve reliability and
without risk sustain higher loads. Bushings and cables with high
optimize the use of the power network.
moisture content at high temperature can explode due to
‘‘thermal runaway’’. The condition of the insulation is an essential aspect for the
Typically, dissipation factor (DF) or power factor (PF) test at operational reliability of electrical power transformers,
power frequency 50/60 Hz is carried out in the field following generators, cables and other high voltage equipment.
well known procedures. DF measured values are then normalized Transformers with high moisture content can not without risk
to 20ºC for comparisons with guidelines and trending. However, sustain overloading conditions. Bushings and cables with high
the temperature correction factor for the normalization of the dissipation factor (DF) / power factor (PF) at high temperature
field measurement is questioned because cellulose with different can explode due to ‘‘thermal runaway’’.
moisture contents as well as oil with different conductivity will
On the other hand it is also very important to identify
have different correction factors. It has been proved that good
‘‘good’’ units among the aging population of equipment.
insulation has less temperature dependent response than the bad
insulation. Adding just a few ‘‘extra’’ service years to the expected
DFR modeling, accurately taking the temperature effect into service life of an electrical device implies substantial cost
account, can be used to model the temperature dependence of DF savings for the power industry.
over a wide temperature range based on measurements over a The widely accepted and most common insulation
frequency range. Naturally, from such modeling, also diagnostic test is carried out measuring capacitance and DF/PF
temperature correction factors for correcting a measurement at line frequency (50/60 Hz). This test is performed whenever
values obtained at one temperature, e.g. 32C, to a reference there is a need for investigating insulation properties. Analysis
temperature, e.g. 20C, can be calculated.
is based on historical values (trending analysis) and comparing
This paper will provide a background of DFR and its
against factory values. Since insulation properties are
modeling specifically applied to model the effect of temperature
and calculating temperature correction factors based on actual temperature dependent, temperature compensation has to be
DFR response. The theoretical basis is backed up with several used for measurements not performed at 20° C, this is
case studies of measurement on samples as well as on real normally achieved using temperature correction table factors
objects, e.g. bushing and transformers at multiple temperatures. for certain classes of devices [1].
However, the standard table values for temperature
Keywords- Dielectric frequency response; DFR; frequency correction are only average values and therefore subject to
domain spectroscopy; FDS; power factor; dissipation factor;
some error [2, 3]. The DFR of dry insulation is less
temperature correction; temperature dependence
temperature dependent while wet insulation is very much
I. INTRODUCTION affected by the temperature. The insulation of the power
With an aging power component population, today’s transformers is even more complicated. It consists of mineral
electrical utility industry faces a tough challenge as failures, oil and cellulose which have different temperature
consequent repair and revenue loss may inflict major costs. characteristics. Therefore, the correction factors are usually
different for different transformers depending on the design

978-1-4673-0487-0/12/$31.00 ©2012 IEEE 296

and individual aging. curve fitting process, the activation energy is calculated. An
example of finding the activation energy is shown in Fig. 2.
The frequency-temperature dependence curves for three
II. TEMPERATURE DEPENDENCE different activation energies are shown in Fig. 3. Equation (1)
To be able to find the correct temperature correction factor, indicates that the increase of temperature has the same effect
the temperature dependence of the insulation system must be as the increase of frequency. To what extent they are related to
investigated. The susceptibility of the insulation material can each other is represented by the activation energy. The curves
be expressed as a function of frequency and temperature [4]: are examples showing that for a certain increase of
temperature, the material with larger activation energy needs
⎛ ω ⎞ to be measured at higher frequency in order to obtain the same
X ( ω , T ) = A ( T )i F ⎜
⎜ ω (T ) ⎟⎟ (1) permittivity.
⎝ c ⎠

A(T) is a temperature dependent amplitude factor, F(x) a

spectral function and ωc ( T ) a characteristic frequency. A(T)
is constant for cellulose. It means that the shape of the
spectrum remains unchanged at different temperatures. Some
insulation materials change the shape of their spectrum when
temperature changes.
Most insulation materials are temperature dependent and in
general, the increase/decrease of temperature makes the
dielectric spectrum shift towards higher/lower frequencies.
However, the shape is usually not changed. In the special case
of an ideal Debye function, the complex permittivity can be
written as:
Δε Figure 1. Example of making the master curve [5]
ε = ε∞ + (2)
1 + jωτ exp ( −Wa / kT )
where Wa is the activation energy, k is the Boltzmann
constant; and, τ is the relaxation time. From (2) it is clear that
the permittivity is a function of −Wa / kT in logarithmic
scale. The dielectric response moves to higher frequency with
temperature increase, or conversely, to higher temperature as
frequency increases. One can obtain the same effect as
increasing the frequency or increasing the temperature. The
activation energy of the Kraft paper is about 1-1.05 eV [5]
while mineral oil has activation energy of 0.4-0.5 eV [6].
Increasing temperature has the same effect on the dielectric
response as decreasing frequency.
The general shape of the curve is often preserved if the data
is plotted on a log-log scale. This feature allows the formation
of a ‘‘Master Curve’’ by shifting the curves along the
frequency axis until they form a continuous curve at a chosen
reference temperature [5]. How to obtain the master curve is
shown in Fig. 1. The dielectric frequency response of the Figure 2. Example of plotting of frequency shifts to find the activation
energy from the slope [5]
cellulose is measured at various temperatures. Then the curves
are shifted until all data points fall on the same curve: the
master curve. The actual fitting of the master curve is done
numerically by an algorithm that allows us to take into
account both the real and imaginary part of the data. In the

297
Figure 3. Relationship between power factor values at different frequencies
taken at different temperatures.

III. DIELECTRIC FREQUENCY RESPONSE MEASUREMENTS Figure 5. Dissipation factor as function of frequency for dry Kraft paper.

A field instrument for DFR measurements of transformers,

bushings and cables was introduced in 1995 [7]. Since then
numerous research groups have evaluated DFR technology
used for field measurement. Several international
projects/reports define dielectric response measurements
together with insulation modeling as the preferred method for
measuring moisture content of the cellulose insulation in
power transformers [8], [9], and [10].
The results are normally presented as capacitance and/or
dissipation factor/power factor versus frequency. DFR results
obtained from a variety of transformers under different
conditions are shown in Fig. 4.
Samples of Kraft paper with various moisture contents
were measured at different temperatures [11]. Results for dry
paper, moisture content <0.5% are shown in Fig. 5.

Figure 6. Dissipation factor (50 Hz) as function of temperature for dry Kraft
paper.

The combined type insulation system of power

transformers consisting of paper/pressboard and oil requires
deeper investigation. In this complex insulation system, the
temperature correction factors are a combination of
temperature dependence of cellulose and temperature
dependence of oil. Also, the amount of cellulose and oil need
to be taken into account.
DFR measurement combined with mathematical modeling
Figure 4. DFR measurement examples. our different transformers with of the dielectric response using the so-called XY model is
moisture content ranging from 0.3 to 3.4%
capable of evaluating the temperature dependence of the
The dissipation factor at 50 Hz as a function of temperature insulation system. The insulation structure of the power
for dry Kraft paper measured as a function of frequency at transformer is represented by the relative amount of spacers
several temperatures is presented in Fig. 6. This analysis (sticks) and barriers in the cooling duct as shown in Fig. 7.
clearly demonstrates that the typical correction factors are not Parameter ‘‘X’’ is defined as the ratio of the sum of all barriers
correct for an individual transformer or bushing. in the duct, lumped together, and divided by the duct width.
The spacer coverage ‘‘Y’’ is defined as the total width of all
the spacers divided by the total length of the periphery of the

298
duct. example, the oil temperature dependence dominates but the
influence of oil and cellulose varies case by case.
Y 1-Y

1-X
∗
ε spacers ε∗oil
∗
ε barriers X

Figure 7. XY model of the transformer [6]

*
The permittivity of oil, ε oil , spacers, ε spacers
*
and barriers,

ε *
barriers , are complex functions of both frequency and
temperature. The equivalent permittivity of the XY model is
given by:
Figure 9. Dissipation factor (50 Hz) as function of temperature for a 20MVA
distribution transformer
Y 1−Y
ε (ω , T ) = +
1− X X 1− X X (3)
+ +
ε spacer ε barrier ε oil ε barrier

Using DFR for moisture determination is based on the

comparison of the transformers dielectric response to a
modeled dielectric response. An analyzing algorithm
rearranges the modeled dielectric response and delivers a new
modeled curve that reflects the measured transformer. Results
are presented as moisture content and oil conductivity for the
transformer.
An example is shown in Fig. 8. The transformer is a 20
MVA distribution transformer. Moisture content of cellulose
and oil conductivity were obtained by varying the parameters
in the model curve to the measured dielectric frequency
response at a certain insulation temperature.
Figure 10. Relationship between dissipation factor values at different
temperatures for materials with different activation energy.

Moisture analysis (moisture in paper insulation) performed

on the same specimen at different temperatures may show
slight deviation. Combining the modeling results and
converting to temperature dependence, temperature curves can
be obtained. Also, for any insulation system containing two
different temperature dependent materials, the applied
conversion considering activation energy of the system
provides results matching very close to the actual measured
dissipation factor values (tan delta) at different temperatures.

Figure 8. MODS® moisture analysis for a 20 MVA transformer IV. SUMMARY AND CONCLUSIONS
The dielectric frequency response in terms of capacitance
Based on the model curve, the dissipation factor as a
and dissipation factor of an electric insulation material or an
function of temperature for that specific 20 MVA transformer
insulation system depends not only on insulation condition but
is shown in Fig. 9. Figure 10 compares the dissipation factor at
also on temperature. When restricted to power frequency
power frequency as function of temperature for oil, cellulose
measurements the effect of temperature is unknown.
and for a combined oil and paper insulation. In this specific

299
 Fortunately, most insulation materials have a simple [3] R.K.Tyagi, S. Victor, N.S.Sodha, “Application of Temperature
Correction Factors for dissipation factor Measurements for Power
relation between capacitance and dissipation factor measured Transformers – A case study”, Doble Client Conference, Vadodara,
as a function of temperature and as a function of frequency India 2006
that can be effectively used when predicting the temperature [4] A.K.Jonscher, “Dielectric Relaxiation in Solids”,
effect on the insulation system by measuring the dielectric [5] D. Linhjell, L. Lundgaard, U.Gäfvert, “Dielectric Response of Mineral
Oil Impregnated Cellulose and the Impact of Aging”, IEEE Transactions
response of the insulation system as a function of frequency. on Dielectrics and Electrical Insulation, Volume: 14 Issue:1, 2007
This relation become a bit more complex, but still manageble, [6] Neimanis, R.; Arvidsson, L.; Werelius, P., “Dielectric Spectroscopy
for an insulation system consisting of e.g. impregnated Characteristics of Aged Transformer Oils”, Electrical Insulation
Conference and Electrical Manufacturing & Coil Winding Technology
cellulose and insulating liquids as it is the case in power Conference, 2003.
transformers. [7] P. Werelius et al, “Diagnosis of Medium Voltage XLPE Cables by High
The temperature-frequency model which takes into account Voltage Dielectric Spectroscopy”, paper presented at ICSD 1998.
the effect of temperature is essential because it makes the [8] U. Gäfvert et al, “Dielectric Spectroscopy in Time and Frequency
Domain Applied to Diagnostics of Power Transformers”, 6th
dielectric frequency response measurement an accurate tool International Conference on Properties and Applications of Dielectric
for estimating the moisture in the cellulose of an insulation Materials, June 21-26, 2000, Xi'an, China.
system. [9] S.M. Gubanski et al, "Dielectric Response Methods for Diagnostics of
Power Transformers”, Electra, No. 202, June 2002, pp 23-34¸also in
The application of dielectric frequency response can also be CIGRE Technical Brochure, No. 254, Paris 2004
used to estimate dissipation factors at other temperatures, e.g. [10] S.M. Gubanski et al, “Reliable Diagnostics of HV Transformer
a reference temperature of e.g. 20° C or at a higher operating Insulation for Safety Assurance of Power Transmission System.
temperature to evaluate the possible risk of thermal runaway. REDIATOOL - a European Research Project”, paper D1-207 CIGRE
2006
V. REFERENCES [11] R. Neimanis, T. Saha, and R. Eriksson, “Determination of moisture
content in mass impregnated cable insulation using low frequency
[1] IEEE Guide for Diagnostic Field Testing of Electric Power Apparatus; dielectricspectroscopy,” IEEE Power Engineering Society Summer
Part 1: Oil Filled Power Transformers, Regulators, and Reactors”, IEEE Meeting, 2000, vol. 1, pp. 463–468 vol. 1, 2000.
62-1995
[2] “Swedish Bushings Plant Sees Growth in RIP Designs”, INMR
Quarterly, Issue 68, 2005

300