Energy and Power Engineering, 2013, 5, 1463-1467
doi:10.4236/epe.2013.54B277 Published Online July 2013 (http://www.scirp.org/journal/epe)
Some Properties of the Electromagnetic Radiation
Emitted by the HV Transformers
Nikolay V. Kinsht, Natalia N. Petrun’ko, Peter V. Katz
Tech. Diagnosis Lab. Institute of Automation & Control Processes FEB RAS, Vladivostok, Russia
Received April, 2013
ABSTRACT
One important way to control the technical condition of the high-voltage power equipment element is the monitoring
and interpretation of electromagnetic radiation. Analysis of the spectra of emitted high-voltage equipment EMR pro-
vides information on which determines the intensity discharge processes occurring in insulation and other structural
elements of high voltage equipment under the operating voltage without interfering with the process.
Keywords: Partial Discharge; Power Transformer; Emitted Electromagnetic Radiation; Technical Condition
1. Partial Discharges in the HV Devices
Insulation
Development of new methods of diagnostics is today one
of the most science-intensive areas of research and as-
sessment of the technical condition and residual resource
of high-voltage transformer equipment (TE) in the early
stages of degradation of its technical conditions (TC).
One of the reasons for the lack of effectiveness of the
methods of TE early diagnostics is the lag in the funda-
mental component of comprehension of electrophysical
processes occurring in the real objects.
Technical conditions of the TE are largely determined
by the condition of its insulation. Partial electrical dis-
charges (PD) arise as during normal operation of the
equipment, as when the degradation of the insulation and
other structural elements. PD is the primary electro-
physical process, and they characterize the equipment
quality.
According the International Electrotechnical Commis-
sion (IEC) International Standard 60270 - 2000 [1] the
principles of the method, which is applied to the present
time and based on the concept of so-called “apparent
charge” well known. They are formulated over 70 years
ago. The simplest mathematical model based on "appar-
ent charge" is constructed on the analysis of the dis-
charge in the single inclusion in the insulation. However,
the practice of operation of the responsible power
equipment showed that the means of control of the TCS
of the equipment, based on the concept of apparent
charge, could not prevent large-scale accidents.
Our researches [2] has shown that this concept is
rather poor, it does not take into account a number of
essential factors: the geometric arrangement of inclusion
in the space of insulation, the variation of the ignition
voltage of the PD, the uncertainties in the voltage of ex-
tinction of the PD, the duration of the burning of the PD,
insulation tgδ. These factors are the parameters of a
common electrophysical processes; in the aggregate
variation of these factors is the uncertainty in several
orders of magnitude. In addition, this concept is funda-
mentally not taking into account the collective effects of
the large number PD. Some ways time-frequency repre-
sentation of a set PD as, for example, it is represented in
Figure 1 are known. However such analysis is insuffi-
cient for practice.
2. EMR as a Diagnostic Sign
PD are caused own electromagnetic radiation (EMR) in a
wide range of frequencies, which can be clearly detected
Figure 1. Example of a phase-numbers representation of
‘apparent’ charges.

N. V. KINSHT ET AL.
1464
in operation mode, and recorded the EMR spectral char-
acteristics with the radio-electronic equipment [3-5]. The
results of the simulation of the radioactive properties of
the entry are confirmed by research on real objects, have
allowed to formulate a methodology for the monitoring
of the spectra of EMR; protected by several Russian pat-
ents.
No recommendations are given in the Great Britain
standard for either measuring methods or bandwidth/
frequencies of instruments to be used in such investiga-
tions, as these methods or instruments, in general, do not
directly quantify the apparent charge of PD current
pulses [6]
The idea of registration of EMR radiated by the HV
equipment, in it, is elementary, on this basis were built
sensors of the PD, located directly on the equipment.
The problem consists only in the fact, what kind of
measuring equipment is used for the registration EMR
and how the results are interpreted. At the present time
among the experts there is no single opinion on the prop-
erties of the EMR emitted TE. So, according to some
data, the study of the TE’s EMR does not make sense to
produce in the range above 25 - 40 MHz. According to
some researchers, EMR entry is advisable to register in
the range of 1 - 3 GHz, and the recommendations of the
well-known company Doble, proposing to use in practice
analyzer of the PD, are in the monitoring of the PD in the
range of 250 - 1000 MHz [7].
On the experimental data of the authors the informa-
tion component of EMR TEs is not evenly distributed
from a few megahertz to a few gigahertzes. Since the
time of the PD current is estimated as 10-8 …10-9 s, and,
therefore, the upper part of the range of frequencies of
natural TE’s EMR extends down to the centimeter waves.
The question consists only in search of information fre-
quency bands EMR and effective selection of the signals
possessing the greatest informative and noise immunity
from the point of view of diagnostics of a technical con-
dition TE.
Study of this issue is of great importance in connection
with the appointment of the technical parameters of the
measuring set, controlling the operation of the TE and
the use of appropriate algorithms of information proc-
essing and decision-making. From the point of view of
the theory of random processes posterior integral imple-
mentation of the EMR, as well as their TE’s spectra are
non-stationary random processes, which is modulated of
the frequency operating voltage. It is evident, however,
that in them there is a clearly expressed the informational
component; these spectra are specific to individual ele-
ments of TE and are similar to the same type of EA.
In Figure 2 are given fragments of 3 the same type
single-phase power transformers 167000 kW 500/220 kV
Figure 2. Example of a spectra EMR of the same type
transformers.
in the range 40 - 170 MHz in the Zeya hydroelectric
power station. Stable explicit spectral lines correspond to
the TV channels.
3. Time-frequency Characteristics of the
EMR Emitted from TE
Spectrum observed visually on the screen of a spectrum
analyzer, is a very dynamic process. Single partial dis-
charges in the insulation TE excite the electromagnetic
oscillations in a wide frequency range inside the trans-
former tank.
Transformer at high frequencies may be represent as a
difficult equivalent circuit, the elements of its construc-
tion-winding, magnetic core, insulation in the aggregate
have a resonance properties of high quality, the fre-
quency response has a lot of resonances (natural fre-
quencies). The energy of the PD excites the transformer
rather lengthy intrinsic high-frequency electromagnetic
oscillations, and emits them in the frequency ranges
relevant to natural frequencies. Integrally this informa-
tion can be represented in the graphical form as a three
dimensional form as showed on Figure 3.
In the simplified approach, the spectra are measured
their maximum, minimum and average value of signal
amplitude (as on separate frequencies, so in a certain
range), as well as the root mean square deviation and the
energy spectrum. However, significantly more effective
appears to be more detailed time-frequency analysis.
Identification of frequency bands and amplitudes of these
oscillations is the main task of the study.

N. V. KINSHT ET AL. 1465
Figure 3. Example of amplitude spectra of EMR.
4. The Interpretation of the Experimental
Spectra
The subsequent presentation illustrates typical data from
the archives of the results of inspections of the power
transformer equipment on the 500/220 kV substation in
Siberia and the Far East Russia, which were made at the
Laboratory for Electrophysics and Electric Power IACP.
(Provided data correspond to the range of frequencies
40 … 170 MHz. Further concrete values of frequencies
and time points don’t represent special interest; they will
be identified only on the serial numbers appropriated at
processing of results).
Considering the maximum spectra, we note quite a
number of local maxima. These peaks may correspond to
both EMF initiated PD, and may correspond to other
sources of electromagnetic radiation, which is an
obstacle to the technology in its own analysis of the
equipment EMR. These other sources, on the one hand,
(in turn) may be the information signals of various kinds
(eg, radio and TV) and on the other hand, they may be
the result of industrial pollution of air. Clearly, the more
reliably able to identify local maxima of the spectrum
corresponding to the PD, the results would be more
correct assessment of the technical condition of the
equipment. Here there are two factors that must be taken
into account. So, first, for diagnostics it is necessary not
only to determine the characteristic frequencies of own
equipment EMR. Taking into account possible variations
in the parameters, the actual paths of radiation quality
factor, inaccurate calculations and natural measurement
errors need to associate each characteristic frequency
sub-band electromagnetic radiation to be monitored.
Second, it may be that even when correctly and reliably
identified above “useful” sub bands EMR noises
significantly exceed the level of “useful” signal. Thus the
observation of electromagnetic radiation in the sub-band
will give false information. Ultimately, the choice for
reliable “workers” subbands informative in terms of
monitoring partial discharges must be combined with a
time frequency analysis.
Let's show the stated principles and problems on ex-
amples.
Let's consider the maximum EMR spectrum registered
at the transformer 125000 kW 220/110 kV (Figure 4
corresponding given above 3-d spectrum). Let's imagine
that there are bases to look for information subband in
the range of frequencies with numbers [410 … 460]. If to
consider an amplitude range is more detailed, in this in-
terval there will be 2 maxima, with numbers about 430
and 441-442, respectively (Figure 5) which can be po-
tential considered as subband of the transformer EMR.
A fragment of the full 3d time-frequency spectrum
corresponding to the frequency range under consideration
makes it possible to clarify this issue (Figure 6).
Figure 4. Maximum EMR spectrum of the transformer
125000 kW 220/110 kV.
Figure 5. The fragment of the Figure 4 spectrum.

N. V. KINSHT ET AL.
1466
Thus, the time dependence of the signal at the
frequency of EMR № 441 shows that this signal is
typical for TV (Figure 7) and the time dependence of the
signal at the frequency of EMR № 429 (Figure 8) is
clearly associated with partial discharges.
Figure 6. The fragment of the full 3d time-frequency.
Figure 7. The signal at the frequency of EMR № 441.
Figure 8. The signal at the frequency № 429.
Figure 9. The fragment of the signal at the frequency of
EMR № 429.
Figure 10. The fragment of the 3d time-frequency spectrum
nearby frequency № 429 and frame № 1540.
Moreover, having analyzed the impulse which has
arisen in the range of time, corresponding 1500 … to
1580 frames, we can say it is a series of more than 10
partial discharges (Figure 9). Finally, the most obvious
is the registered pulse 3d representation EMR reflecting a
series of partial discharge, which is clearly seen its
localization in both time and frequency (Figure 10). The
degree of localization is due to the quality factor of this
natural frequency.
5. Conclusions
To study the elements of EMR near high-voltage
equipment is to identify criteria that characterize the
condition of the internal technical equipment. The most
informative is the analysis of time-frequency properties
of their own EMR spectra of the equipment.
The frequencies corresponding to the individual lines
of EMR related to partial discharges are the natural

N. V. KINSHT ET AL.
1467
frequencies of the transformer.
6. Acknowledgements
This work was supported by Russian Foundation for
Basic Research (grant N13-08-00924)
REFERENCES
[1] International IEC Standard 60270-2000.
[2] N. V. Kinsht and M. A. Katz, “Some Problems of the
Partial Discharge Burning Time,” TEEM, Vol. 7, No. 6,
2006, pp. 319-323.
[3] N. V. Kinsht and N. N. Petrun'ko, “The Experience of
Inspection of a Technical Condition of the HV Equipment
on Substations by a Method of Registration of the Own
Electromagnetic Radiations,” in Proceedings of
CMD2008, 2008, pp.738-740.
[4] N. V. Kinsht and N. N. Petrun'ko, “Research on Charac-
teristics of EMR of the High-Voltage Equipment during
the Process of its Diagnostic,” in Proceedings of
CMD2010, 2010, pp. 1149-1152.
[5] N. V. Kinsht and N. N. Petrun'ko, “Interpretation of
Monitoring Dataof Own Electromagnetic Radiation HV
Transformers” in Proceedings of CMD2012, 2012, pp
569-572.http://www.samgor.com/pdf/IEC60270.pdf.
[6] http://www.doble.com/litlib/files.html/download/395