26th Nordic Insulation Symposium on Materials, Components and Diagnostics (NORD-IS 19), Tampere, Finland, 2019

Temperature Influence on Dielectric Response of

Rotating Machine Insulation and Its Correction
J. Cheng1, P. Werelius2, N. Taylor1
1
(School of Electrical Engineering and Computer Science,
KTH – Royal Institute of Technology, Stockholm, Sweden)
2
(Megger Sweden, Stockholm, Sweden)
[email protected]

Abstract is an important factor that might change the results

dramatically. Comparison of results over time needs to be
Insulation condition is an essential aspect for the performed without strong influence from extraneous
operational reliability of high voltage rotating machines in parameters such as temperature. This can be done by
power plants and industrial applications. Insulation resistance controlling such parameters, or by compensating for their
(IR) and line-frequency dissipation-factor / power-factor (tanδ) effect. Figure 1 is an example of trending the insulation
measurement are often performed for the assessment of stator resistance of a motor over time (two measurements per year
insulation condition. These measured values need to be performed in summer and winter respectively) with and
normalized to a reference temperature (e.g. 40 °C) for without temperature compensation [2]. It indicates that without
comparison and trending and this is traditionally achieved by proper temperature correction, even a small fluctuation of
multiplying the results with a certain factor. However, this temperature causes difficulty in detecting the insulation
correction could be subject to error for an individual device condition change. It happens frequently that people forget to
since the correction factors recommended by various standards record temperature during field insulation measurement. And
are average values of a certain number of machines at different fails to do so one is not able to be aware of the deterioration of
conditions. In addition to that, insulation condition also has stator insulation even if it is close to failure [3].
some influence on the temperature dependent property.
With the introduction of Dielectric Frequency Response, DFR
and Polarization/Depolarization Current, PDC as more
advanced insulation diagnostic methods, with proper
modelling, temperature correction can be done based on the
insulation condition of an individual device and thus accuracy
is considerably improved. In this paper, the background of
DFR and its superiority in temperature correction are
introduced. After that, the numerical Fourier and Inverse
Fourier Transformation algorithm is applied to correct the time
domain measurement (IR and PDC).

1. Introduction
High voltage motors and generators are key components in
power generation stations as well as other applications such as
oil & gas industry, steel company, pulp and paper industry.
Any unexpected failure of those high voltage rotating machines
may cause extensive damage to other apparatus, long time
interruption of power supply and significant financial losses.
Figure 1. Insulation resistance graph with and without
Effective maintenance strategy can reduce the risk of temperature correction
unexpected failure. According to a CIGRE survey [1], stator
insulation is the component that has the highest failure rate for
hydrogenerators. In general, the condition of stator insulation is The commonly used temperature correction method has its
poorly known, and appropriate electrical testing is required to limitations in aspect of accuracy. A more advanced correction
detect potential deterioration and avoid catastrophic failure. algorithm for both IR and tanδ results is analyzed in this paper.
However, due to the widely varying stator insulation materials
Insulation resistance, IR (including the polarization index)
and system designs, this paper puts more emphasis on the
and dissipation factor / power factor / tanδ test are the most
methodology of how to calculate the correction factor rather
widely used testing techniques all over the world. Temperature
than giving correction factors for different insulation materials.

145
 26th Nordic Insulation Symposium on Materials, Components and Diagnostics (NORD-IS 19), Tampere, Finland, 2019

By following the proposed procedures given by this paper, The measurement principle and setup of DFR is very close
more reliable temperature correction is possible for various to dissipation factor/power factor with the difference that
insulation materials. insulation property is measured in much wider frequency
range, typically from 1 mHz to 1 kHz rather than just at power
frequency as shown in Figure 2. It is obvious that frequency
2. Typical temperature correction method sweep contains more information than just a single frequency
measurement.
It is common practice to do temperature correction by
Another benefit of DFR is its easy-recognized temperature
multiplying the testing result by a factor recommended by
characteristics. It has been recorded by various literatures that
relevant standards. Table 1 is an example of commonly used
the increase / decrease of temperature makes the curve shift
temperature correction factors for insulation resistance
towards higher/lower frequencies while the curve shape
measurement on rotating machines [3]. Such correction method
remains unchanged, as shown in Figure 2 [7].
can also be applied to tanδ measurements. However,
temperature correction factors in such table are usually average
values of different materials and describe a material under
“normal” conditions. In reality the condition of an insulation
material is unknown and therefore the correction might be
subject to error for an individual device [4].

Table 1 Approximate temperature correction factors for motor

insulation resistance, from [3]
Temperature (°C) Correction Factor (Kt)
20 0.25
30 0.50
40 1.00
Figure 2. Typical curves of tan  vs. frequency. The arrow
50 2.00 shows shifting of curves towards higher frequency at higher
60 4.00 temperatures

The newer edition of the relevant standard C57.12.90-

2010 [5] has the following note: “Experience has shown 3.2 Individual temperature correction in frequency
that the variation in power factor with temperature is
substantial and erratic so that no single correction curve domain (AC)
will fit all cases.” The IR of a Class A rotating equipment Instead of using average values, the individual
at 50 °C differs by more than an order of magnitude as temperature correction, ITC algorithm is built on the basis
compared to that of 20 °C. Therefore, if the given device of actual temperature behavior of the insulation material
does not conform to the “average” value then the applied and takes its condition into account [6]. Therefore it is able
correction factor of 10.45 might lead to a significant error to provide temperature correction for an individual device
in the estimation of the IR value at 20 °C. Therefore, an in a more reliable way.
individual approach for accurate temperature correction is of
great benefit. Taking the simplest dielectric property Debye relaxation as
an example, its temperature dependent complex permittivity
can be written as:
3. Dielectric frequency response and its
temperature characteristics ∆𝜀
𝜀 = 𝜀∞ + (1)
1 + 𝑗𝜔𝜏𝑒 −𝐸𝑎/(𝑘𝐵∙𝑇)
3.1 Dielectric frequency response
With the development of modern electronics, newer
technology enhances the performance of measurement circuit. where Ea is the activation energy of the insulation material (J),
It enables the tester to be able to measure the minuscule kB is the Boltzmann constant (1.38  10-23 J/K) and T is the
amounts of current that passes through the insulation system of temperature (K) of the object.
a stator. Dielectric Frequency Response, DFR is a new testing The activation energy can be determined by horizontally
technique of particular interest as it can yield greater insights shifting complex relative permittivity curves obtained from
into the insulation condition assessment. various temperatures until a roughly continuous curve is

146
 26th Nordic Insulation Symposium on Materials, Components and Diagnostics (NORD-IS 19), Tampere, Finland, 2019

constructed at a chosen reference temperature [8]. This curve is

the so called “Master Curve”.
4. Polarization and depolarization current and
its temperature characteristics
4.1 Individual temperature correction in time
domain (DC)
In time domain, insulation testing is performed by
applying a DC voltage to the test object and recording the
current with respect to time. The accurate correction of time
domain data is difficult because that the magnitude of
current/resistance curve also varies with temperature.
Fortunately, it is possible to transform the time domain data
into frequency domain by Fourier Transformation. Suppose
that the current vs. time curve has been obtained with
object temperature 60 °C, the correction to 40 °C can be
achieved by following procedures:
 Convert the time domain data obtained at 60 °C (T1)
into frequency domain (F1) using the Fourier
Transformation
Figure 3. Formation of the Master Curve from a set of DFR  Apply individual temperature correction algorithm to
measurements on one object at at different temperatures the frequency domain data (F1) so that frequency
domain data at 40 °C (F2) is obtained
 Transform the frequency domain data (F2) back into
After that, the shifted distance in logarithm scale is plotted the time domain (T2) using the inverse Fourier
against the multiplicative inverse of temperature. The slope of Transformation
the straight line (absolute value) in Figure 4 is equal to the
activation energy. 4.2 Fourier and inverse Fourier Transformation
Based on [9] and neglecting some components that
approaches zero when either time or frequency goes to zero
and infinity, the time domain data and frequency domain
data can be transformed into each other by using following
equations:
∞
′′ (ω)
1
C = ∫ ipol (t) sin(𝜔𝑡)𝑑𝑡 (3)
U0
0
2 ∞
ipol (t) = U0 ∫0 C ′′ (ω) sin(ωt)dω (4)
π

where ω is the angular frequency, C ′′ is the loss of

dielectric material, ipol (t) is the insulation leakage current
during charging.
Figure 4. Plotting of shifted distance vs. reciprocal for Since it is nearly impossible to obtain the analytical
temperature solution of the integration above, some approximation
Knowing the activation energy, the correction of measured methods especially the Hamon approximation [10] are
DFR curve from temperature T1 to temperature T2 could be often used to simplify the transformations. Some error
achieved by shifting the curve in logarithm scale by distance L, might be introduced and in addition to that, it often causes
which is equal to: hesitation when giving a value to some unknown
parameters. The numerical solution which can be expressed
by equation (5) and (6) is a very good candidate for high
𝐸𝑎 1 1 quality transformation.
𝐿 = ln [− ( − )] (2) ∞
𝑘𝐵 𝑇2 𝑇1 1 tn +∆t
′′ (ω)
C = ∑ ipol (t n ) ∫ sin(ωt)dt (5)
U0 tn
n=0

147
 26th Nordic Insulation Symposium on Materials, Components and Diagnostics (NORD-IS 19), Tampere, Finland, 2019

∞
ω𝑛 +∆ω increased voltage levels, significant non-linear behavior is
2
ipol (t) = U0 ∑ C ′′ (ω𝑛 ) ∫ sin(ωt)dω (6) present.
π ω𝑛
n=0
Therefore, to reach the balance between measurement
accuracy and maintaining linear property, test voltage
The minimum time interval (Δt) of two adjacent PDC between 1 kV and 2 kV is recommended for measurement
that needs temperature correction afterwards. Fourier
measurement points is 1 second due to the instrument
Transformation of the time domain result measured at 2 kV
limitation and the first point is assumed to be captured
DC is shown in Figure 8. The comparison of transformed
at the first second since start of measurement. The
data and actual measurement data suggests that accuracy is
missing data between time zero and the first second
satisfactory.
doesn’t have much influence on the frequency range of
interest. So expression (5) could be re-written to
equation (7):
∞
′′ (ω)
1 ipol (t n )
C = ∑ [cos(𝜔𝑡𝑛 ) − cos(𝜔𝑡𝑛+1 )] (7)
U0 𝜔
n=0

where 𝑡𝑛 = 𝑛 + 1. Selection of angular frequency 𝜔 is

set according to the total measurement time 𝑡𝑝𝑜𝑙
following equation (8) and (9):
𝜔 = 𝑛𝜋/𝑡𝑝𝑜𝑙 n=2, 3, 4…. (8)
2𝑛+1
𝜔= 𝜋/𝑡𝑝𝑜𝑙 n=2, 3, 4…… (9)
2

The purpose of such selection is to eliminate the

fluctuation introduced by using a fixed current between
two time points.

A 110/10 kV transformer is measured in frequency Figure 5. Transformation from time domain to frequency
domain (DFR) at 140 V RMS and time domain (PDC) at domain. The red curves are reference curves obtained from
100 V respectively. Then results are transformed into the actual measurement. Blue and black curves are transformed
other domain respectively and compared with the actual curves from time domain data.
measured data. It can be observed from Figure 5 and Figure
6 that good agreement has been achieved for both
transformations. To ensure accurate transformation, the
time domain measurement should last for at least 1000 s
and the frequency domain measurement should go down to
0.1 mHz. Although the transformation is only valid for a
certain frequency and time range, this range is of particular
interest for condition diagnosis.

5. Non-linear system
Linear insulation material exhibits identical dielectric
property at different voltages. But almost all insulation
materials or systems are more or less non-linear. For HV
rotating machines coils, the stress grading material at the
slot end makes the stator insulation system non-linear. The
nonlinear property has been studied thoroughly and it Figure 6. Transformation from frequency domain to time
causes difficulty in results interpretation and diagnosis domain. The red curve is the reference curve obtained from
[11]. actual measurement. Blue curve is the transformed curve
from frequency domain data.

Figure 7 is the DFR of an aged motor winding coil

measured at different voltages. In lower voltage limit the
non-linear property of the system is not obvious. With

148
 26th Nordic Insulation Symposium on Materials, Components and Diagnostics (NORD-IS 19), Tampere, Finland, 2019

References
[1] CIGRE Working Group A1.10, “Survey of Hydrogenerator Failures”,
2009, pp. 5-9.
[2] Application Guide, A Guide to Diagnostic Insulation Testing Above 1 kV,
Megger Ltd., Dover, 2017, pp. 25-26.
[3] IEEE Standard 43 - 2000, “IEEE Recommended Practice for Testing
Insulation Resistance of Rotating Machinery”, Electric Machinery
Committee of IEEE Power Engineering Society, 2006.
[4] S. Zurek, A.E. Rasheed and M. Ohlen, “Individual Temperature Correction
(ITC) for Insulation Resistance Measurements”, 13th International
Electrical Insulation Conference, 2017.
[5] R.K. Tyagi, S. Victor and N.S. Sodha, “Application of Temperature
Correction Factors for Dissipation Factor Measurements for Power
Transformers – A Case Study”, Doble Client Conference, 2006.
[6] D. Robalino, P. Werelius, M. Ohlen and J. Cheng, “Dielectric Frequency
Response Measurements and Dissipation Factor Temperature
Dependence”, The International Symposium on Electrical Insulation
Conference, 2012.
[7] R. Neimanis, T. Saha, and R. Eriksson, “Determination of moisture content
in mass impregnated cable insulation using low frequency dielectric
spectroscopy”, IEEE Power Engineering Society Summer Meeting, 2000.
[8] D. Linhjell, L. Lundgaard and U. Gäfvert, “Dielectric Response of Mineral
Figure 7. DFR of the aged motor coil at various AC Oil Impregnated Cellulose and the Impact of Aging”, IEEE Transactions
voltages on Dielectrics and Electrical Insulation , Volume: 14 Issue:1, 2007.
[9] A. Helgesson, “Dielectric Properties of Machine Insulation studied with
Dielectric Response”, Licentiate thesis, KTH, Stockholm, 1997.
6. Conclusion [10] A. K. Jonscher, “Dielectric Relaxation in Solids”, Chelsea Dielectrics
Press, London, 1983.
Temperature correction can be achieved accurately in [11] N. Taylor, “Dielectric Response and Partial Discharge Measurement on
frequency domain by shifting the dielectric frequency response Stator Insulation at Varied Low Frequency”, PhD thesis, KTH, Stockholm,
2010.
of an individual insulation material with certain distance. Good
accuracy of numerical Fourier and inverse Fourier
Transformation makes it possible to transform the time domain
measurement into frequency domain and vice versa for the
particular insulation material studied. Therefore, the time
domain temperature correction can be achieved by
transforming it into the frequency domain for ITC and then
transforming back. In this way, the change of insulation
condition over time can be identified more reliably and without
confusion.

Figure 8. Fourier transformation of time domain

measurement on the aged motor coil

149