EVENTS

TESTS

ABSTRACT

Individual
Dielectric testing techniques in the
time and frequency domains are in-
creasingly being used by transformer
manufacturers, power utilities and
researchers for transformer oil-paper

temperature
insulation systems condition assess-
ment. Since 1997, when the first por-
table device designed to carry out
dielectric response tests in the fre-

compensation
quency domain in the field was put
on the market, the technology has
evolved and new features have been
incorporated. One of these features
is becoming a “must have” tool for
power transformer dielectric condi-
tion assessment: individual tempera-
ture compensation. Benefits of dielectric response measurements
1. Introduction power network. Transformers are without
KEYWORDS a doubt one of the most critical compo-
A healthy insulation condition of an elec- nents in the system and, for this reason,
dielectric response, temperature cor- trical apparatus is essential for the ope- a great amount of research has been per-
rection, power factor, dissipation factor rational reliability of the entire electrical formed to better comprehend the values

42 TRANSFORMERS MAGAZINE | Volume 2, Issue 3

 Diego ROBALINO

„ Dielectric response measurement in the fre-

quency domain provides a tool to better inter-
pret the thermal behaviour of the dielectric
characteristics of oil-paper insulation systems

fore, a power factor test should be compa- a voltage high enough to allow easy mea-
rable if performed at the same frequen- surement under substation interference
cies. But frequency is not the only factor conditions, but not too high as to stress
affecting the power factor value. the system. Test voltages in the field test
instrument range from below 100 V to as
When repeating the power factor test on high as 12 kV. Field tests are usually per-
the same specimen, if moisture, oil condi- formed at rated voltage or a maximum of
tion and aging have not been altered, but 10 kV. The AC signal is typically applied at
the temperature of the system changes, the two different frequencies which are very
power factor will change as well, and the close to the power frequency. The use of
values at two different temperatures will two different frequencies is called fre-
not be comparable. In order to be compa- quency variation suppression mode, and
rable, the PF values must be normalized to instead of running one single test at power
a 20 °C reference. Therefore, temperature frequency (50 or 60 Hz), the test is carried
has a significant effect on the resulting va- out at two frequency values close to the
lue of the power factor and this fact should reference line frequency. The AC capaci-
be taken into account and improve the tance test is part of the PF test because the
existing methods used to compensate po- capacitance value and its associated char-
wer factor measurements for temperature ging current are required to calculate the
variation. PF value of that specific capacitive (insula-
ting) system later.
This article provides a background on
dielectric response measurements in the PF testing of transformers is carried out to
Frequency Domain (DFR), also known as assess the level of contamination of the in-
Frequency Domain Spectroscopy (FDS), sulation system and reference limits have
and the advantages of using this testing been set in international standards [1] [2]
technique in liquid–filled power trans- to determine the possibility of dielectric
formers to assess their insulation system degradation/contamination or mechani-
condition, as well as to obtain the “unique” cal damage of the insulation material. Fol-
individual thermal response of the capaci- lowing the limits set by the different inter-
obtained in routine dielectric tests such tive system analyzed. national references [3] [4], PF is a trigger
as power factor / dissipation factor (tan to announce potential accelerated aging
delta). Power factor is widely used among or degradation of the insulation system.
operators and manufacturers to set a refe-
rence value for the losses encountered in 2. The routine dissipation It is necessary for this test to record the in-
the insulation material. factor / power factor test at sulation system temperature to later nor-
power frequencies malize measured values to the 20 °C refe-
Typically, Dissipation Factor (DF) or Po- rence. The dissipation factor of insulation
wer Factor (PF) tests are carried out at The PF and capacitance test is one of the can be more or less sensitive to the effect
power frequency of 50/60 Hz. A voltage is most effective methods of assessing the of temperature depending on the con-
applied to one electrode of the capacitive overall condition of a transformer. An AC dition of the bulk insulation system. So
system and the total resulting current is signal is applied to the insulation system at far, the method used to normalize p­ower

„
measured. From there, the angle between
the total current vector and the applied
voltage vector is obtained and the cosine
function given in percentage values is the
power factor value of the tested capaci-
The dissipation factor of insulation can be
tance. In this test, the power factor value more or less sensitive to the effect of tempera­
is a function of frequency. The capacitive
reactance of the object is directly related to
ture depending on the condition of the bulk
the excitation signal frequency and there- i­nsulation system
w w w . t ra n sfo r m e r s - m a g a z i n e . co m 43
„
TESTS

Experience has shown that the variation in Where Ea is the activation energy, a term
best regarded as an experimentally deter-
power factor with temperature is substantial mined parameter that indicates the sensi-
and erratic so that no single correction curve tivity of the reaction rate to temperature, k
is the Boltzmann constant (8.6173324(78)
will fit all cases x10-5 eVK-1); and, τ is the relaxation time.
From (2) it is clear that the permittivity is a
function of in the logarithmic scale.

factor values obtained at temperatures dif- 3. Temperature dependence The activation energy of oil-impregnated
ferent from 20 °C has been to apply cor- cellulose is about 0.9-1.1 eV, while mineral
rection factors. Correction factors may be To be able to determine the correct tem- oil has activation energy of 0.4-0.5 eV. The
available from equipment manufacturers perature correction factor, the tempera- general shape of the curve is often preser-
and test equipment manufacturers and ture dependence of the insulation system ved if the data is plotted on a log-log scale.
are only based on nameplate data. Generic must be investigated. The susceptibility of
correction factors were available in IEEE the insulation material can be expressed The correction factor curves for different
standard C57.12.90-2006, section 10.10.5, as a function of frequency and tempera- temperatures and three different activati-
but were removed in C57.12.90-2010 [5] ture [6]: on energies are shown in Fig. 1.
with the following note on page 48:
Equation (1) indicates that the increase
“NOTE 3.b) Experience has shown that (1) of temperature has the same effect as the
the variation in power factor with tem­ increase of frequency. To what extent they
pera­ture is substantial and erratic so that are related to each other is represented by
no single correction curve will fit all cases.” the activation energy as presented in (2).
Where A(T) is a temperature dependent The curves are examples showing that for
A PF test at power frequency by itself is amplitude factor, F(x) a spectral function a certain increase of temperature, the ma-
capable of detecting moisture and con- and ωc(T) a characteristic frequency. A(T) terial with larger activation energy needs
tamination in a transformer; however, it is constant for cellulose. It means that the to be measured at higher frequency in or-
cannot differentiate whether the source of shape of the spectrum remains unchanged der to obtain the same permittivity.
power factor values beyond recommen- at different temperatures. The dielectric
ded limits or in an unexpected accelera- response moves to higher frequency with
ted growth from historical values corre- temperature increase, or conversely, to 4. The influence of moisture
spond to moisture in the solid insulation higher temperature as frequency incre-
or contamination on the liquid insulation.
in the cellulose
ases. One can obtain the same effect by
Further analysis is performed to investiga- increasing the frequency or increasing Dielectric response measurements to-
te the cause of values beyond the establis- the temperature. However, the shape is gether with insulation mathematical mo-
hed limits and field users should perform usually not changed. In the special case of delling using the so called XY-model [3]
other tests including physicochemical an ideal Debye function, the complex per- are today one of the preferred methods for
analysis of oil, DGA and of course DFR. mittivity can be written as: measuring moisture content of the cellu-
lose insulation in power transformers. The
The recommended limits for new and ser- results are normally presented as capaci-
vice aged power transformer insulation tance and/or dissipation factor / power
power factor at 20 °C are detailed in Table (2) factor versus a wide frequency spectrum
18 of IEEE C57.152-2013 [1], and are pre- that typically ranges from 1 kHz down to
sented in Table 1: 1 mHz.

New power transformers are expected

„
to have no more than 0.5 % moisture
Table 1. Power factor recommended limits for power transformers as per IEEE C57.152-2013

Insulating liquid

kV rating Nominal/New PF Serviceability
limit aged limit
The dielectric re­
sponse moves to high­
er frequency with tem­-
Mineral oil < 230 kV 0.5 % 1.0 %

Mineral oil ≥ 230 kV 0.4 % 1.0 %

perature in­crease, or
Natural ester All 1.0 % 1.0 % conversely, to higher
“The numbers shown here for natural esters are only provisional as there are no correction
temperature as freq­u­
curves established by the industry yet.” ency increases
44 TRANSFORMERS MAGAZINE | Volume 2, Issue 3
„ Dielectric respon- 100
se measurements to-
­ge­ther with insulation 10
modelling are today

Correction factor
one of the preferred 1
methods for measuring
moisture content of 0.1
activation energy 0.5 eV

the cellulose insulation activation energy 0.7 eV

activation energy 0.9 eV
in power transformers 0.01
0 10 20 30 40 50
Temperature [°C]

concentration in cellulose. Throughout

service life, normal aging process and Figure 1. Correction factor curves for different activation energy values
the factors stressing the insulation in the
transformer will increase the moisture
concentration in the cellulose. Operators
worldwide will prefer not to exceed 3.5 %
moisture in cellulose, as it severely increa- 1
ses the aging process and the risk of failure. at 5 °C
at 20 °C
An example of a combined mineral oil and at 35 °C
Power factor [%]

paper insulation inter-winding capaci- 0.1 at 50 °C

tance is shown in Fig. 2. The moisture con-
centration of the sample is 1 % and mea-
surements are taken at 5, 20, 35 and 50 °C.
0.01
Taking only the 60 Hz values and coll-
ecting those in a table, one can see that
the PF at 60 Hz may only increase or
0.001
decrease depending on the temperature
0.001 0.01 0.1 1 10 100 1000
condition of the insulation, and, in this
Frequency [Hz]
case, where the specimen is with only 1 %
moisture concentration, the correction
factor along the thermal spectrum of 5 °C Figure 2. Dielectric response in the frequency domain for a sample with 1 % moisture concent-
and 50 °C is very close to unity. The ob- ration in the cellulose at 5, 20, 35 and 50 °C
tained values can be summarized in Fig. 3

„
and Fig. 4.

The temperature- 0.3

f re q u e n c y m o d e l 0.25

makes possible the

Power factor [%]

0.2
normalization of di­ 0.15
electric parameters 0.1
such as power factor 0.05
/ dissipation factor to
the selected reference 0
0 10 20 30 40 50 60
temperature (20 °C) Temperature [°C] TX1 mc=1 %
at the selected fre-
quency (50 or 60 Hz) Figure 3. Thermal behaviour of the PF for the sample with 1 % moisture concentration in the cellulose

w w w . t ra n sfo r m e r s - m a g a z i n e . co m 45
 „
TESTS

The temperature
1.4 correction factor is
1.2 not a single curve for
all conditions
Correction factor

0.8

0.6
Important to highlight here is that all the
0.4
dielectric responses shown in Fig. 2 deve-
0.2 lop only one thermal response, as shown
in Fig. 3.
0
0 10 20 30 40 50 60
The same sample tested at high moisture
Temperature [°C] TX1 mc=1 % concentration (3.5 %) in the same ther-
mal spectrum provides a different dielec-
Figure 4. Correction factors for PF based on thermal behaviour for the sample with 1 % moisture tric response (Fig. 5).
concentration
Following the same procedure as that with
a low moisture concentration, the power
factor values measured at 60 Hz are pre-
sented separately in Fig. 6 and Fig. 7.
1
at 5 °C
at 20 °C
As it can clearly be seen, the temperature
at 35 °C
correction factor is not a single curve
for all conditions. Aging, moisture and
Power factor [%]

0.1 at 50 °C
contamination will have an effect on the
dielectric response and one main way to
obtain an accurate normalization of po-
0.01 wer factor values is by using DFR’s feature
- individual temperature compensation.
The dielectric response of a fluid-filled
transformer is unique, as is its dielectric
0.001
thermal behaviour. Incorrect correction
0.001 0.01 0.1 1 10 100 1000 of power factor values may lead to erratic
Frequency [Hz] decisions overestimating its value or un-
derestimating the risk of failure.
Figure 5. Dielectric response in the frequency domain for a sample with 3.5 % moisture concen-
tration in the cellulose at 5, 20, 35 and 50 °C
Conclusion
The dielectric frequency response me-
thod is nowadays widely used to estimate
2.5 the moisture concentration of the solid
insulation in fluid-filled transformers
in terms of capacitance and dissipation
2
factor as a function of frequency. The
Power Factor [%]

measured power factor at one specific

1.5
frequency of an insulation system de-
pends not only on temperature, but also“
1 on the individual condition of the insu-
lation system under test. When the test is
0.5 restricted to power frequency measure-
ments only, the accurate effect of tempe-
0 rature is unknown.
0 10 20 30 40 50 60
Temperature [°C] TX1 mc=3.5 % The dielectric frequency response taken
at only one voltage level and constant
temperature allows conversion from the
Figure 6. Thermal behaviour of the PF for the sample with 3.5 % moisture concentration frequency domain into the time domain

46 TRANSFORMERS MAGAZINE | Volume 2, Issue 3

 Bibliography
4
3.5 [1] IEEE C57.152 – 2013 Guide for Diag-
3 nostic Field Testing of Fluid-Filled Power
Correction factor

Transformers, Regulators, and Reactors

2.5
2 [2] NETA ATS – 2013 Standard for Ac-
ceptance Testing Specifications for Elec-
1.5
trical Power Equipment and Systems
1
[3] CIGRE TB 445 - Guide for Transfor-
0.5
mer Maintenance
0
0 10 20 30 40 50 60 [4] IEEE Std C57.106-2006 Guide for Ac-
ceptance and Maintenance of Insulating
Temperature [°C] TX1 mc=3.5 % Oil in Equipment
[5] IEEE Std C57.12.90-2010 - IEEE Stan-
Figure 7. Correction factors for PF based on thermal behaviour for the sample with 3.5 %
moisture concentration dard Test Code for Liquid-Immersed Dis-
tribution, Power, and Regulating Trans-
formers
and, as it has been demonstrated in this rence temperature (20 °C) at the selected [6] A.K. Jonscher, Dielectric Relaxation
article, from frequency domain to the frequency (50 or 60 Hz). An extended ap- in Solids, Chelsea Dielectrics Press, July
temperature domain. The temperature- plication of the mathematical approach 1983
frequency model is essential because it allows normalizing power factor values
makes possible the normalization of die- to any temperature between 5 °C and 60 [7] E. Kuffel, W.S. Zaengl, J. Kuffel, High
lectric parameters such as power factor °C at any frequency value within the fre- Voltage Engineering Fundamentals, 2nd
/ dissipation factor to the selected refe- quency measurement range. edition, Butterworth-Heinemann, 2000

Author
Diego Robalino
is currently a Seni­or
Applications Eng­
in­e er at Megger,
where he special­
izes in the diagno­
sis of com­plex el­-
ec­trical testing pro-
­cedures. He obtained his Ph.D. in el-
ectrical engineering from Tennessee
Technological University while doing
research in power system optimizati-
on with a focus on aging equipment.
Robalino has been engaged in electri-
cal engineering for 20 years, with ma-
nagement responsibilities in the power
systems, oil and gas, and research are-
nas managing the design, construction
and commissioning of electrical and
electro-mechanical projects. He is a se-
nior member of the IEEE and member
of the IEEE transformers main com-
mittee. Other memberships include
ASTM and PMI. His multidisciplinary
engineering interests had led him to
author and co-author papers and artic-
les, as well as present at major internati-
onal energy industry conferences.

w w w . t ra n sfo r m e r s - m a g a z i n e . co m 47