Application of The Integrated AE and HFCT Sensors For Online Dry-Type Transformer Partial Discharge Monitoring. Case Study

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Kitimasak TIKAKOSOL, Tanes TANITTEERAPAN, and Narong MUNGKUNG

King Mongkut’s University of Technology, Thonburi, Thailand

doi:10.15199/48.2021.02.24

Application of the integrated AE and HFCT sensors for online


Dry-type Transformer Partial Discharge Monitoring. Case study
Abstract. This study proposes a measurement technique to identify and determine the occurrence of partial discharge in the insulation of Dry-type
transformers, as well as their criteria level. Concerning IEC 62478, the four AE (acoustic emission) sensors were used to determine the PD sources
and one HFCT (High-Frequency Current Transducer) technique was to identify the PD source and the criteria level. This process accurately
identified the PD source location occurring at any position in a transformer. The gap time was 3.2 ms. The level of partial discharge for this case
study is Major PD that means soon monitoring every 3 months. The results of this technique could prevent insulation failure of a transformer in time.

Streszczenie. Zaprezentowano technikę wykrywania wyładowań niezupełnych w izolacji suchego transformatora. Zgodnie z IEC 62478 użyto
czterech czujników emisji akustycznej AE oraz jednego czujnika prądów wysokiej częstotliwości HFCT. Umożliwia to lokalizację źródła wyładowania.
(Wykorzystanie scalonego czujnika AE I HFCT do online monitoringu wyładowań niezupełnych w suchym transformatorze)

Keywords: Partial discharge, High frequency current transducer, Dry-type transformer, Acoustic emission sensor.
Słowa kluczowe:wyładowania niezupełne, izolacja suchego transformatora, czujnik AE, czujnik HFCT.

Introduction detect, identify partial discharging problems, analyze the


A transformer insulation inspection and partial discharge location of partial discharges, and assessing the risk of the
diagnosis (PD) is an important parameter and technology to partial discharges.
reflect the state of theinsulation. Accurate PD signal Therefore, Cast Resin Transformers are widely used in
measurement is the key to bringing online detection industrial systems such as heavy industries, airports,
systems for transformer insulation [1]. Unusual devices hospitals, public transportation systems, and high-rise
produce nonstop PD pulses and allow physical phenomena buildings. Although Cast Resin transformers are safer than
such as electromagnetic radiation, noise and light. The oil type transformers, the diagnosis of the insulation of Cast
main methods for detecting PD using these physical Resin transformers is more difficult. The life of Cast Resin
phenomena include ultra-high frequency pulse (UHF), as transformers depends largely on the insulation conditions
well as methods for measuring signals and light waves. be it under the heat, electric field, mechanical force, and the
However, current pulse methods are at risk of various environment that leads to the electric charge transfer in the
disturbances. Electromagnetic signals are mainly used for transformer.
offline quantitative detection in environments with small This research proposes a method for measuring partial
disturbances. The methods of UHF may have the ability to discharge transfer on Cast Resin transformers, which is
prevent strong electromagnetic interference [2,3]. But the based on the IEC 62478 / Ed1 standard. These
PD UHF quantitative problems have not been thoroughly measurements will be put into actual use for Cast Resin
resolved. Besides, the ultrasonic method is sensitive to all transformers case study [7].
types of noise and is used primarily for evaluating the Experimental setup
quality of operating equipment. Also, the complexity of the The experimental setup is shown in Fig.1. Various
internal structure of the transformers requires that most methods have been explored for measuring PD-based
existing PD detection methods rely on external access to electrical and non-electrical phenomena. The methods that
capture signals, causing difficulty in locating internal faults can successfully identify and find PD are detecting signal
(Such as winding insulation defects) precisely. emission based on the fact that PD appears to be the origin
Therefore, the study of effective methods and of acoustic waves [8,9]. These acoustic waves spread
equipment for PD detection is important and useful. The through the internal structure of high voltage devices until it
analysis of partial discharges using signal and reaches the external surface. Detection and coarse location
electromagnetic wave detection methods is an effective of at least one source can be done by moving one or more
technique that can detect that the transformer insulation is external sensors to another location on the transformer
deteriorated without the need to stop the transformer housing. The accuracy of the position of the PD source may
operation. For decades, transformer users have tried to find be determined by the relative time of the acoustic signal in
ways to assess the insulation condition of transformers and each sensor. The electrical signal from PD is in the form of
identify possible problems. In the past, diagnostic tests a unipolar pulse, with an increase in time that can be as
have been greatly developed [4,5], causing some short as nanoseconds. The increase in pulse at the point of
discharging analysis to identify the problem position of origin depends on the type of discharge. The high
partial discharge more precisely. frequencies are reduced when the signal propagates
Periodic offline diagnostic tests has continued to play an through the transformer and the pulse shape is modified
important role in the industry. However, "continuous" or due to multiple reflections and the exciting resonant
"online" investigations are becoming increasingly popular. frequency of the transformer components. The acoustic PD
The offline work for the power system has been reduced for signal is compared to the electrical PD signal, which is
the last method. It can also increase the efficiency and based on the real PD signal from the transformer. The
reliability of transformers in power stations. It can reduce combined method is the electrical measurement that
maintenance costs and help increase the efficiency of improves confidence in the results of the PD test. The data
equipment and maintenance procedures, as well as helps storage device used to view the results on a computer is
manage the increased risk of aging transformers [6]. called "PowerPD" (PD-TP500A).
In this research, we present some methods for detecting
and diagnosing discharges in order to identify sensors that

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 2/2021 105


is the measurement of an 800 kVA 24kV Cast Resin
Transformer which is used in high-rise buildings. The
transformer has not been maintained for at least 3 years.
The techniques that will be used for measurement include
the Acoustic Emission Sensor and the High-Frequency
Current Transformer. The transformer measurements are
as follows.

Signal frequency measurement results using Acoustic


Emission Sensor.
In the measurement of transformers, 4 signal sensors
(AE1-AE4) will be installed, with 1 installation on each side
of the transformer, which is shown in Fig. 4, and Fig. 5,
showing the waveforms of each sensor measured with the
PowerPD program.

Fig.1. Experimental setup

The four AE sensors were mounted on the house of the


transformer and the high frequency current transducer
(HFCT) sensor was clamped to the ground wire of the
transformer [10,11]. The signals from the four acoustic
emission (AE) sensors and the HFCT sensor were
measured simultaneously and the readings were stored on
the computer. Then the dedicated software was used to
analyze the PD pattern, frequency domain, and time
domain. Acoustic emission sensor is a passive piezoelectric
sensor. Its frequency response is characterized by a peak
of 150kHz where it exhibits a resonance. The frequency
range is 100 to 450 kHz. HFCT has a split core ferrite to
allow retrospective fitting to earth straps without the need
for disconnection. The HFCT is constructed with an
aluminum body to provide RF shielding and improve
performance in noisy environments. The frequency range is
100kHz to 10MHz with transfer function at 4.5-5.0 V/A.
During the installation of AE and HFCT sensors, the AE
sensor No.1 (Green color) is installed at the top of the HV
side of the transformer. The AE sensor No.2 (Red color) is
installed next at the bottom of the transformer, determined
by the clockwise direction. The AE sensor No.3 (Yellow
color) is installed at the LV side of the transformer at the top Fig. 2. Partial discharge signal
position approximately opposite to the AE sensor No.1. The
AE sensor No.4 (Blue color) is installed approximately
opposite to the AE sensor No.2 at the bottom position. The
HFCT sensor on the other hand is installed at various
positions of the ground connection from the transformer that
is to be tested. The PD signal detects that the other AE
sensors are moved to the corresponding AE sensor which
detects the PD activity and installed in X-Axis and Y-Axis
fashion before confirming the exact location of the PD
source. If PD bursts were detected from both AE sensors
and HFCT sensor, it could be concluded that the PD
occurred within the transformer as seem in Fig. 2. The
severity criterion of partial discharge is called “Gap time”
and the unit is in milliseconds. The gap time is a space
between the end of one burst and the beginning of the next
burst, called burst interval as seen in Fig. 3. It gets
dangerously close to 2 ms, clearly indicating that failure is
imminent. Burst interval is a critical information in
determining the severity of PD.
. 3 The severity criteria of partial discharge; Gap time
Results of Case Study Results of electromagnetic wave frequency
In this research, the transformer partial discharges were measurement results using the HF current sensor
measured with real-life transformers while the transformers To measure the transformer, 1 HFCT sensor will be
were still working. By adopting non-conventional installed by installing it on the ground wire of the
measurement techniques, the technique of locating partial transformer [12,13]. This will be shown in Fig. 6 and Fig. 7
discharging on transformers was performed. In assessing showing the electromagnetic signal waveform that can be
the risks of partial discharges, the case study of the Cast measured with the PowerPD program.
Resin Transformer was 24 kV 800 kVA in/of 2004. This case

106 PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 2/2021


Fig. 4. The location of the signal sensor installation

e) The partial discharge signal from the AE sensor No.4 (Blue)

Fig. 5. The graph of signal frequency waves, which a) compares


the size of the signal waves of all 4 sensors, b-e) the signal
waveforms obtained from each sensor.

a) The comparison size of 4 signal waveforms,AE1-AE4

Fig. 6. The location of the installation of the HF current sensor

b) The partial discharge signal from the AE sensor No.1 (Green)

Fig. 7. The electromagnetic signal waveform

c) The partial discharge signal from the AE sensor No.2 (Red)

Fig. 8. The installation position of the horizontal axis signal sensor

d) The partial discharge signal from the AE sensor No.4 (Yellow)r)

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 2/2021 107


Results of locating the partial displacement in the X-
axis.
Due to the signal waveform measurement results shown
in Fig.5, it is found that the maximum signal size is from
sensor 2 (red), because the PD signal source is near the
red sensor. After that, all 4 sensors will be moved to the 2nd
sensor, arranged in the X-axis, and then compare the signal
wave size again to find the closest PD position. Fig. 8
shows the position of the installation of the horizontal axis
signal sensor and Fig. 9 shows the signal wave graph in
which the 4 signal sensors are installed.

d) The partial discharge signal from the AE sensor No.3 (Yellow)

a) The comparison size of 4 signal waveforms,AE1-AE4

e) The partial discharge signal from the AE sensor No.4 (Blue)


Fig. 9. The horizontal axis of the signal frequency waveform, a) The
comparison size of 4 signal waveforms,AE1-AE4, b-e) the signal
waveform from each sensor, AE1-AE4.

Results of finding the position of partial displacement


along the Y-axis.
Due to the signal waveform measurements shown in
Fig. 10, the maximum signal size is at sensor 3 (yellow),
where the signal source PD is near the yellow sensor. After
that, all 4 sensors will be moved to the 3rd sensor, arranged
in a Y-axis, then the signal waveform size will be compared
again. To find the closest PD position. Fig. 11 shows the
location of the vertical signal sensor and Fig. 11 shows a
graph of signal waves where all 4 signal sensors are
installed.
b) The partial discharge signal from the AE sensor No.1 (Green)

c) The partial discharge signal from the AE sensor No.2 (Red color) Fig. 10. The location of the vertical signal sensor installation.

108 PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 2/2021


a) The comparison size of 4 signal waveforms,AE1-AE4
e) The partial discharge signal from the AE sensor No.4 (Blue)

Fig. 11. The vertical waveform of the signal , which a) compares


the size of the sound waveforms in all 4 sensors, b) the sound
waveforms obtained from each sensor.

Results of the locating of the partial discharges.


Due to the signal waveform measurements shown in
Fig. 11, it is found that the maximum signal size is from
sensor 2 (red) because the PD signal source is near the
yellow sensor. After that, all 4 sensors will be moved to the
2nd and 3rd sensors, then, the signal wave size will be
compared again. To specify the position of the PD. Fig. 12
shows the location of the partial discharges on the
transformer and Fig. 13 shows a signal waveform in which
the 4 signal sensors are installed.

b) The partial discharge signal from the AE sensor No.1 (Green)

Fig. 12. The position of partial discharging on transformers.


c) The partial discharge signal from the AE sensor No.2 (Red)

d) The partial discharge signal from the AE sensor No.3 (Yellow) Fig. 13. The frequency waveform of sound waves in the position of
partial discharging.

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 2/2021 109


Assessment results of partial discharge type and risk
Due to the results of the electromagnetic wave
measurement shown in Fig. 7, it was found that the form of
the electromagnetic wave signal is characteristic of a partial
discharge on the insulation surface (Surface Discharge),
which is probably caused by dust and conductive (or semi-
conductive) particles on the insulation surface. However,
the risk level of PD according to PowerPD's reference data
is moderate risk level (Major PD) of 3.2 ms. Fig. 14
compares some of the discharging waveforms with the
standard format and Fig. 15 shows the result risk analysis
of PD.

b) The risk level of PD according to the PowerPD reference.

Fig. 15. The comparison of the risk level of PD, which a) is the
electromagnetic waveform curve from the PowerPD meter, b) is the
table of the risk level of PD according to the PowerPD reference.

For the case study, the transformer has been stopped,


and insulation surface of the transformer checked. Particles
were found as partial discharge marks on the high-voltage
coils as shown in Fig. 16 where the trace positions match
the measurement and analysis results.
a) The electromagnetic waveform

Discharge Marks

PD

Fig. 16 The evident of partial discharge on the insulation surface of


b) The standard waveform graph of PD. high voltage coils.
Fig. 14. The comparison of partial discharges waveforms with the
standard model, which a) is the electromagnetic signal waveform As the results show, this transformer was tested with the
from the PowerPD meter. b) is the standard signal waveform of PD. combined AE and HFCT detection technique and it was
detected that there is partial discharge within the surface
and inside the insulation of the coil of the transformer. This
detection led to immediate shut down of the equipment and
correction was carried out and the transformer was put to
normal use. This transformer was tested at a hotel in
Thailand with the combined AE and HFCT detection
technique and it was detected that there is PD within the
Surface of insulation between the terminator of phase B and
HV tap bar [14-17]. The treeing can be clearly seen within
the coil of the transformer which could have led to fatal
consequences as seen in Fig. 15. The gap time of the
Partial discharge source was 3.2 ms that is “major PD ”
state. If the trend is rising, consider removal from service in
near future in monitoring every 3 months.

Conclusions and discussions


From the online PD measurements on dry-type
transformers using the PD-TP500A together with the
emission sensor (refer to IEC62478) and the HFCT type
mounting, the PD source can be detected and searched. By
a) the electromagnetic signal waveform from the PowerPD meter using the AE sensor to measure the PD, it is possible to find

110 PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 2/2021


the PD position that occurs within the dry type transformer. [8] Harold, R.T., 1985, “Acoustical Technology Applications in
The equipment used for measuring PD (PDTP500A, AE, Electrical Insulation and Dielectrics”, IEEE Trans. Electr. Insul.,
and HFCT sensors) can be installed on the measuring Vol. 20, No. EI-1, Feb. 1985, pp. 1-3.
[9] Varlow, B.R., Auckland, D.W. & Smith, C.D., 1999, “Acoustic
device during operation and there is no need to turn the
Emission Analysis of High Voltage Insulation”, IEE Proc.-Sci.
equipment off and separate it from the electrical system. Meas. Technol., Vol. 146, No. 5.
This set of devices can detect and analyze problems [10] IEC Committee, 2000, IEC 60270 High-voltage test techniques
occurring within the transformers at initial conditions for – Partial discharge measurements, IEC The International
solutions. More importantly, it can be determined whether Electrotechnical Commission, December 2000, Geneva,
the problem is caused by PD, mechanical problems, arcs, Switzerland.
or loose parts within the transformer [18,19]. This PD [11]Transformer Committee, 2007, IEEE Std C57.127-2007 IEEE
diagnosis method can accurately detect and locate the PD Guide for the Detection and Location of Acoustic Emissions
from Partial Discharges in Oil-Immersed Power Transformers
position occurring within the transformer which will enable
and Reactors, The Institute of Electrical and Electronics
timely and efficient maintenance. This will help prevent the Engineers, 31 August 2007, New York, USA.
complete erosion of the transformers and reduce the costs [12]Urairat Fuangsoongnern, Winai Plueksawan, & Promsak
resulting fromserious repairs of damaged transformers. Apiratikul, 2008, “Partial Discharge Analysis for Power
Online partial discharge measurement is a method to Distribution Transformer Model”, The Proceeding of 46th
prevent transformer damage caused by t explosion and it Kasetsart University Annual Conference, 29 January – 1
provides guidelines for determining transformer February 2008, pp. 72-80.
maintenance for safer use [20,21]. [13]Plueksawan, W., Apiratikul, P. & Fuangsoongnern, U., 2008,
“Partial Discharge Analysis for Power Distribution Transformer
To help solve the problem of electrical instability caused
Model”, EECON-31. 31st Electrical Engineering Conference,
by power outages due to the weather in Thailand with 29-31October 2008, pp. 299-302.
heavy rain and lightning, leading to the defective electrical [14]Santosh Kumar, A., Gupta, R.P., Udayakumar, K. &
insulation of transformers caused by partial discharges in Venkatasami, A., 2008, “Online Partial Discharge Detection
the transformer system, this therefore applied AE and and Location Techniques for Condition Monitoring of Power
HFCT device for the measurement and positioning of partial Transformers: A Review”, IEEE International Conference on
discharges in transformers. It has been found that it can be Condition Monitoring and Diagnosis, 21-24 April 2008, Beijing,
used to measure and locate partial discharges. The data China.
[15] R.K.K. Gupta, and R.A. Jaswal, “Partial Discharge Test in
can also be used to plan for transformer maintenance as
Power Transformer Against Winding Insulation: A Review,”
per the plan. International Journal of Applied Engineering Research, Vol. 9,
Number 9, pp. 1087 –1090, 2014.
Authors: Mr. Kitimasak Tikakosol, E- [16] Li JISHENG, Luo YONGFEN1, Li JUNHAO, Li YANMING,
mail:kittimasak.tikakosol@mail.kmutt.ac.th, Assoc.prof. Dr. Tanes “The study of phased-ultrasonic receiving-planar array
Tanitteerapan, E-mail:tanes_kmutt@yahoo.com, assoc.prof.Dr. transducer for PD location in power transformer,” PRZEGLĄD
Narong Mungkung, E-mail:narong_kmutt@hotmail.com. ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097,
Department of electrical technology education, King Mongkut’s R. 87 NR 3/2011, pp.324-328.
University of Technology Thonburi, Thailand. 126 Prachauthit road, [17] Xiaoli ZHOU, Yong QIAN2, Muqing LIU, “The Application of
bangmod, Thungkru, Bangkok, 10140, Thailand. Partial Discharge Detection for the condition assessment of
XLPE Power Cables,” PRZEGLĄD ELEKTROTECHNICZNY
REFERENCES (Electrical Review), ISSN 0033-2097, R. 88 NR 6/2012,pp.313-
[1] Rutgers, W.R.; Fu, Y.H. UHF PD-Detection in a power 316.
transformer. In Proceedings of 10th International Symposium [18]Azevedo, C.H.B; Marques, A.P. and Ribeiro, C.J., Methodology
on High Voltage Engineering, Montreal, Canada, 1997; pp. for the detection of partial discharges in power transformers
219–222. using the acoustic method,” in Proc. EUROCON 2009
[2] Judd, M.D.; Li, Y.; Hunter, I. Partial discharge monitoring for Conference, Saint Petersburg, Russia, May, 2009.
power transformers using uhf sensors part I: sensors and [19]A. P. Engelbrecht, Computational Intelligence – An Introduction,
signal interpretation. IEEE Electr. Insul. Mag. 2005, 21, 5–14. 2ª ed.,John Wiley & Sons, 2007.
[3] Aschenbrenner, D.; Kranz, H.G. On line PD measurements and [20] H.P. Romualdo, P.H.S. Palhares, L.C. Brito, C.J. Ribeiro, A.P.
diagnosis on power transformers. IEEE Trans. Dielectr. Electr. Marques; C.H.B. Azevedo, J.A.L. Santos, “Identification,
Insul. 2005, 12, 206–222. Measurement and Three-dimensional Localization of partial
[4] Kreuger, F.H., 1964, Partial Discharge Detection in High discharges in power transformers based on Acoustic Emission
Voltage Equipment, Butterworths, English. Tests," In: 5th Brazilian Symposium on Electrical Systems -
[5] Boggs, S.A., 1990, Partial Discharge-Part III: Cavity Induced SBSE 2014, Foz do Iguaçu - PR, Brazil, v. 1, p. 1-6, 2014.
PD in Solid Dielectrics, IEEE Electr. Insul. Mag., Vol. 6, No. 1, [21] A.P. Marques, C.J. Ribeiro, C.H.B. Azevedo, L. C. Brito, J.A.L.
Jan./Feb., pp. 11-20. Santos; M.A.S. Carvalho, e R.B. Nascimento, “Analysis of the
[6] Ogihara, H., 1964, “Detection and location of Coronas in oil implementation of an intelligent classification system of partial
Immersed Transformer with Corona Detector” Electr. Eng. Jpn., discharges in power transformers," In: 23rd National Seminar
Vol. 84, 1964, pp. 12-22. on Electricity Production and Transmission (SNPTEE 2015)],
[7] Kuffel, J., 2012, IEC 62478/Ed1: High Voltage Test Techniques Foz do Iguaçu - PR, Brazil, 2015.
- Measurement of Partial Discharges by Electromagnetic and
Acoustic Methods" Proposed Horizontal Standard”, IEC The
International Electrotechnical Commission, 17 August 2012,
Geneva, Switzerland.

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 2/2021 111

You might also like