2018 10th International Conference on Information Technology and Electrical Engineering (ICITEE)

Investigation of 150/66 kV 100 MVA Transformer

Internal Failure
Brian Bramantyo S.D.A Harsono, Aristo Adi Kusuma, Buyung Sofiarto Munir, Nur Widi Priambodo
Transmission and Distribution Department
PLN Research Institute
Jakarta, Indonesia
[email protected]

Abstract—An investigation was performed upon the failure of overvoltage problem which might cause the breakdown of the
a 150/66 kV 100 MVA power transformer and nearby 66 kV transformer itself. However, further investigation is needed to
arrester during heavy rain. To verify the possible cause of the verify the correlation of lightning activity toward transformer
failure, an evaluation was conducted by performing visual breakdown.
inspection on the transformer and the arrester after breakdown,
analyzing related historical assessment and maintenance data, II. METHODOLGY
evaluating recorded lightning activity data from lightning
detection system and generating relevant simulation using
The investigation and further analysis was performed upon
transformer nameplate and actual configuration. Majority of the collected data which consist of field inspection after
historical assessment and maintenance data results showed no breakdown, historical assessment and maintenance data of the
indication of transformer early breakdown, but the sweep transformer, historical fault data of the transformer, digital fault
frequency response analysis (SFRA) test result showed severe recorder (DFR) data during failure and also recorded lightning
deformation on primary and secondary winding. According to the activity data from lightning detection system. Several
simulation using 20 kA 1.2/50 s lightning strike on 66 kV simulation scenarios were generated using Electromagnetic
transmission line connected to the transformer, overvoltage on the Transient Program to verify the possible cause of the
transformer reached 1.69 p.u. for primary winding and 8.47 p.u
transformer failure. The transformer parameters used in
for secondary winding. However, according to recorded lightning
data during time of failure, there’s no nearby lightning strike
simulation was adjusted and configured accordingly to
around the transformer or the transmission lines connected to it. transformer name plate and actual field configuration in order
Moreover, lightning overvoltage would initially shatter the to provide reliable result.
transformer’s bushing rather than directly deforming the
transformer tank. The simulation result for short circuit between III. FIELD DATA EVALUATION
primary winding phase R to ground showed identical current and A. Visual Inspection of Transformer and Arrester
voltage waveform with digital fault recorder data during
breakdown. According to data and simulation result analysis, According to visual inspection documentation of the
lightning activity has no contribution toward the failure of the transformer, both primary and secondary bushing were missing
150/66 kV 100 MVA transformer. The possible cause of (Fig. 1). All of the bushing were shattered during the breakdown
transformer breakdown was internal failure in form of arcing which might indicate the increase of internal oil pressure. From
between primary winding phase R to ground. bushing turret, there’s visible white smoke which indicated the
presence of hot spot inside the tank. For safety reason, climb up
Keywords—Transformer, SFRA, ligshtning data, internal visual inspection was cancelled.
failure, arcing

I. INTRODUCTION
As one of important apparatus in power system, routine
assessment data analysis and evaluation on previously fault of
power transformer is essential to maintain optimum
performance and reliability. For example, further analysis can
be conducted toward routine assessment and maintenance data
such as dissolved gas analysis (DGA), dielectric test, tan delta
of winding and bushing and also sweep frequency response
analysis (SFRA) [1][2]. Nowadays, SFRA has been widely used
to evaluate the internal condition of the transformer (core,
winding, termination, etc.) whether there’s any deformation
after transformer transporting or after severe fault [3-7]. In late
2017, a 150/66 kV 100 MVA power transformer experienced
sudden failure during heavy rain. Damaged 66 kV arrester
nearby the transformer gave a possibility of lightning Fig. 1. Transformer condition after failure

978-1-5386-4739-4/18/$31.00 ©2018 IEEE 196

 2018 10th International Conference on Information Technology and Electrical Engineering (ICITEE)

The specification of the damaged transformer as explained result. Both insulation resistance test and tan delta test were
on the nameplate was shown in Table I. The neutral side of performed in 2015 and 2017, while SFRA test was performed
secondary winding of the transformer was connected to 29  in 2015.
Neutral Grounding Resistor (NGR).
The result of insulation resistance test was shown in Table
TABLE I SPECIFICATION OF THE TRANSFORMER III. The polarization index for all test mode still above fair
condition range 1.11-1.3 [1], hence the insulation of the
transformer before failure was in good condition. Note that the
insulation between primary and earth and also between primary
and tertiary had the lowest PI value among other result (1.44).

TABLE III. INSULATION RESISTANCE TEST RESULT

The result of tan delta test of the winding was shown in

Table IV. According to the reference, the maximum value of
power dissipation is 0.5% [1]. Noted that almost all the tan delta
value of transformer below 0.5%, hence the transformer
Visual inspection of the 66 kV arrester of the transformer
winding insulation before failure was in good condition.
showed ruptured sheath, melted top cover and melted counter
(Fig. 2). The specification of the arrester was shown in Table II. TABLE IV. TRANSFORMER TAN DELTA TEST RESULT

The result of tan delta test of the bushing was shown in Table
V. According to the reference, the maximum value of power
dissipation is 0.5% [1]. Noted that almost all the tan delta value
of transformer below 0.5%, hence the transformer bushing
insulation before failure was in good condition.

TABLE V. BUSHING TAN DELTA TEST RESULT

Fig. 2. Arresters on secondary side of the transformer

TABLE II. SPECIFICATION OF THE ARRESTER

SFRA test result which was taken on 2015 test showed

severe deformation of primary and secondary winding as shown
B. Historical Assessment and Maintenance Data Evaluation in Fig. 3 and Fig. 4. According to the reference, the frequency
of The Transformer region for interaction between winding ranges between 2 kHz
and 20 kHz while the frequency region for winding structure
Historical assessment and maintenance data of the used in ranges between 20 kHz and 1 MHz [2]. For both region, the
the evaluation were dielectric test result (insulation resistance SFRA result for each phase was different one to another on
test), tan delta test result (winding and bushing) and SFRA test primary and secondary winding. Severe deformation indicated

197
 2018 10th International Conference on Information Technology and Electrical Engineering (ICITEE)

I1 I2 I3
any clearance change or shift between one or more winding and I/ A

also between each winding toward transformer tank which 5000

might affected the insulation strength resulting in possible

insulation breakdown. 0
-0.025 0.000 0.025 0.050

-5000

-10000

Fig. 6. Current waveform record during fault

(Phase R=red, phase S=green and phase T=blue)

Noted that after the fault, voltage at phase R was decline and
all current had same phase angle as an indication of phase to
ground fault.
IV. MODELING AND SIMULATION
Fig. 3. SFRA result for primary winding of the transformer Two simulation scenarios was performed using Transient
(Phase R=grey, phase S=yellow and phase T=brown) Analysis Program: lightning strike simulation on 66 kV
transmission line which is connected to the transformer and
internal short circuit simulation between winding and ground.
According to local operator information, transformer failure
occurred during heavy rain. The visible damage on 66 kV
arrester might become indication of lightning strike possibility
on 66 kV side. Simulation model for lightning strike on 66 kV
was generated using Electromagnetic Transient Program as
shown in Fig. 7. Note that two 150 kV transmission lines model
was used to represent the actual configuration of the substation
and the transformer was adjusted according to nameplate data.
The lightning characteristic used in the simulation is 20 kA
1.2/50 s and strike directly to 66 kV transmission line which
connected to the transformer (the 66 kV arrester was assumed
Fig. 4. SFRA result for secondary winding of the transformer
(Phase r=green, phase s=blue and phase t=red) damaged and unable to chop the overvoltage).

C. Historical Fault Data of The Transformer

Based on the fault data in 6 transmission line connected to
the substation and the fault data of the transformer itself from
2016-2017, there were 11 faults which mostly consist of
transmission line fault with distance relay attenuation. Previous
fault on the transformer and the transmission line connected to
the transformer could generate fault current which gave
mechanical stress toward transformer winding [7-9].
Mechanical stress on transformer winding gave huge
contribution toward insulation degradation.
D. Digital Fault Recorder record Fig. 7. Simulation model for lightning strike on 66 kV line
The voltage and current waveform on the primary side of the
transformer during the failure were shown in Fig. 8 and Fig. 9. The simulation result current and voltage waveform on
primary side of the transformer was used as comparison toward
U/ V

100000
available DFR data. The voltage waveform at 150 kV side when
lightning strike 66 kV side of transformer was shown in Fig. 8,
Fig. 9 and Fig. 10, while the overvoltage on the 66 kV side was
-0.025 0.000 0.025 0.050 shown in Fig. 11, Fig 12 and Fig. 13. The maximum magnitude
0
of lightning overvoltage on 66 kV side and 150 kV side reached
8.47 p.u, and 1.69 p.u.
-100000

Fig. 5. Voltage waveform record during fault

(Phase R=red, phase S=green and phase T=blue)

198
 2018 10th International Conference on Information Technology and Electrical Engineering (ICITEE)

300 400

[kV] [kV]

200 300

100 200

0
100

-100
0

-200
0.17 0.18 0.19 0.20 0.21 0.22 0.23 [s] 0.24 -100
(f ile suny aragi.pl4; x-v ar t) v :X0009A v :X0009B v :X0009C 0.188 0.192 0.196 0.200 0.204 0.208 [s] 0.212
(f ile suny aragi.pl4; x-v ar t) v :X0011A v :X0011B v :X0011C

Fig. 8. Voltage waveform of 150 kV side of transformer for lightning strike

Fig. 12. Voltage waveform of 66 kV side of transformer for lightning strike
on 66 kV side phase R (Phase R=red, phase S=green and phase T=blue)
on 66 kV side phase S (Phase R=red, phase S=green and phase T=blue)
300
500
[kV]
[kV]
200 400

100 300

200
0

100
-100

0
-200
0.17 0.18 0.19 0.20 0.21 0.22 0.23 [s] 0.24
(f ile suny aragi.pl4; x-v ar t) v :X0009A v :X0009B v :X0009C
-100
0.1860 0.1912 0.1964 0.2016 0.2068 [s] 0.2120
(f ile suny aragi.pl4; x-v ar t) v :X0011A v :X0011B v :X0011C
Fig. 9. Voltage waveform of 150 kV side of transformer for lightning strike
on 66 kV side phase S (Phase R=red, phase S=green and phase T=blue)
Fig. 13. Voltage waveform of 66 kV side of transformer for lightning strike
on 66 kV side phase T (Phase R=red, phase S=green and phase T=blue)
300

[kV]
Recorded lightning data from lightning detection system
200
(LDS) was analyzed into 3 group: lightning data around the
transformer, lightning data along 2 150 kV transmission lines
100
and lightning data along 4 66 kV transmission lines. For
lightning data around the transformer, LDS detected 9 lightning
0
activity ranging from -40 kA up to 46 kA with closest event
occurred 3.6 km away from the transformer. However, none of
-100
those lightning occurred during time of failure 19.01 (the
closest event occurred at 21.51). For lightning data along the
-200
0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 [s] 0.28
150 kV transmission lines, LDS detected 32 lightning activity
(f ile suny aragi.pl4; x-v ar t) v :X0008A v :X0008B v :X0008C
ranging from -40 kA up to 46 kA with closest event occurred
Fig. 10. Voltage waveform of 150 kV side of transformer for lightning strike 1.72 km away from one of the 150 kV transmission line.
on 66 kV side phase T (Phase R=red, phase S=green and phase T=blue) However, none of those lightning occurred during time of
500
failure 19.01 (the closest event occurred at 20.45). For lightning
[kV]
data along the 66 kV transmission lines, LDS detected 51
400 lightning activity ranging from -40 kA up to 46 kA with closest
event occurred 0.32 km away from one of the 66 kV
300
transmission line. However, none of those lightning occurred
200
during time of failure 19.01 (the closest event occurred at
21.34).
100 According to SFRA test result, the deformation on both
primary and secondary windings were detected. The
0
deformation could indicate winding position shift which might
lead to possible short circuit from transformer winding to
-100
0.1900 0.1944
(f ile suny aragi.pl4; x-v ar t) v :X0011A
0.1988
v :X0011B v :X0011C
0.2032 0.2076 [s] 0.2120 ground due to higher insulation stress. The possibility of phase
to ground fault was supported with the DFR waveform pattern
Fig. 11. Voltage waveform of 66 kV side of transformer for lightning strike where the voltage of one phase was declining (phase R) and all
on 66 kV side phase R (Phase R=red, phase S=green and phase T=blue) the current waveform had the same phase angle. Deformed tank

199
 2018 10th International Conference on Information Technology and Electrical Engineering (ICITEE)

also served as an evidence of sudden pressure increase inside 200

[kV]
the transformer tank which might be caused by internal arcing. 150

If the SFRA result was combined with insulation resistance test, 100

the primary winding had the highest possibility to experience 50

insulation breakdown compared to secondary winding. 0

Simulation model for internal phase to ground fault on primary
-50
winding was generated using Electromagnetic Transient
-100
Program as shown in Fig. 14.
-150

-200
0.00 0.05 0.10 0.15 0.20 [s] 0.25
(f ile suny aragi.pl4; x-v ar t) v :X0009A v :X0009B v :X0009C

Fig. 17. Voltage waveform for phase S to ground fault on primary winding
(Phase R=red, phase S=green and phase T=blue)

8000
[A]
6000

4000

Fig. 14. Simulation model for phase to ground fault on transformer winding 2000

0
The voltage waveform for phase to ground fault simulation
-2000
on primary winding was shown in Fig. 15, Fig. 17 and Fig. 19,
while the current waveform for phase to ground fault simulation -4000

on primary winding Fig. 16, Fig. 18 and Fig. 20. -6000

150
-8000
[kV] 0.00 0.04 0.08 0.12 0.16 [s] 0.20
(f ile suny aragi.pl4; x-v ar t) c:X0008A-X0001A c:X0008B-X0001B c:X0008C-X0001C
100

50 Fig. 18. Current waveform for phase S to ground fault on primary winding
(Phase R=red, phase S=green and phase T=blue)
0

200
-50 [kV]
150

-100
100

-150
0.00 0.04 0.08 0.12 0.16 [s] 0.20 50
(f ile 1502060MVA.pl4; x-v ar t) v :X0008A v :X0008B v :X0008C

0
Fig. 15. Voltage waveform for phase R to ground fault on primary winding
(Phase R=red, phase S=green and phase T=blue)
-50

8000
-100
[A]
6000
-150
4000 0.00 0.05 0.10 0.15 0.20 [s] 0.25
(f ile suny aragi.pl4; x-v ar t) v :X0009A v :X0009B v :X0009C

2000
Fig. 19. Voltage waveform for phase T to ground fault on primary winding
0
(Phase R=red, phase S=green and phase T=blue)
-2000
8000
-4000 [A]
6000
-6000

4000
-8000
0.04 0.06 0.08 0.10 0.12 0.14 0.16 [s] 0.18
(f ile 1502060MVA.pl4; x-v ar t) c:X0008A-X0001A c:X0008B-X0001B c:X0008C-X0001C
2000

Fig. 16. Current waveform for phase R to ground fault on primary winding 0
(Phase R=red, phase S=green and phase T=blue)
-2000

-4000

-6000

-8000
0.00 0.04 0.08 0.12 0.16 [s] 0.20
(f ile suny aragi.pl4; x-v ar t) c:X0008A-X0001A c:X0008B-X0001B c:X0008C-X0001C

Fig. 20. Current waveform for phase T to ground fault on primary winding
(Phase R=red, phase S=green and phase T=blue)

200
 2018 10th International Conference on Information Technology and Electrical Engineering (ICITEE)

V. DISCUSSION REFERENCES
Although the result of simulation for lightning strike on 66 [1] IEEE Guide for Diagnostic Field Testing of Electric Power Apparatus -
kV side of the transformer could give overvoltage on both Part 1: Oil Filled Power Transformers, Regulators, and Reactors,"
in IEEE Std 62-1995 , vol., no., pp.1-68, Dec. 1 1995.
primary and secondary winding of transformer as shown in Fig.
[2] IEC 60076-18 Ed. 1: Power Transformers – Part 18: Measurement of
8 to Fig. 13 (maximum overvoltage reached 8.47 p.u. for 66 kV frequency response, 2011.
side and 1.69 p.u. for 150 kV side), the recorded lightning data [3] L. Yong, C. Yanjie, Q. Rui, J. Shengchang, L. Xianchen, et. Al.,
from lightning detection system showed no lightning activity “Simulation study of influence of transformer bushing fault on frequency
during transformer failure time (19.01). From all 92 recorded response curves,” High Voltage Engineering, vol. 40, pp. 2406-2415,
lightning activity, the closest occurrence time was 20.45 or 1 August 2014.
hour and 44 minutes after the failure time of the transformer. On [4] A. Abu-Siada, N. Hashemnia, S. Islam and M. A. S. Masoum,
the other hand, lightning overvoltage would initially damage the "Understanding power transformer frequency response analysis
signatures," in IEEE Electrical Insulation Magazine, vol. 29, no. 3, pp.
secondary bushing rather than directly deform the transformer 48-56, May-June 2013.
tank because of the lightning overvoltage surpassed the basic
[5] J. A. S. B. Jayasinghe, Z. d. Wang, P. N. Jarman and A. W. Darwin,
insulation level (BIL) of secondary and primary busing (325 kV "Winding movement in power transformers: a comparison of FRA
or 4.48 p.u. for secondary bushing and 650 kV or 4.64 p.u. for measurement connection methods," in IEEE Transactions on Dielectrics
primary bushing) [10]. With shattered bushing happened first, and Electrical Insulation, vol. 13, no. 6, pp. 1342-1349, December 2006.
the possibility of transformer tank deformation afterward was [6] IEEE Guide for the Application and Interpretation of Frequency
small. Therefore, the possibility of lightning strike on 66 kV Response Analysis for Oil-Immersed Transformers," in IEEE Std
transmission line didn’t happen. C57.149-2012 , vol., no., pp.1-72, March 8 2013.
[7] S. M. Islam, "Detection of shorted turns and winding movements in large
Current with high magnitude which was produced form power transformers using frequency response analysis," 2000 IEEE
previously recorded 11 transmission line fault might potentially Power Engineering Society Winter Meeting. Conference Proceedings
(Cat. No.00CH37077), 2000, pp. 2233-2238 vol.3.
give mechanical stress to the winding of the transformer [7-9].
Mechanical stress could shift internal component of the [8] Zhenyuan Wang, “Artificial Intellegience Application in the Diagnosis of
Power Transformers Incipient Faults”, Dissertation for PhD Degree in
transformer such as the winding, hence the internal insulation Virginia Polytechnic Institute and State University, 2000.
strength would be compromised and decreased. The winding [9] P. Henault, "Detection of internal arcing faults in distribution
structure shift due to previous fault on the transformer was transformers," 2011 IEEE PES 12th International Conference on
proven in SFRA test result as shown in Fig. 6 and Fig. 7. The Transmission and Distribution Construction, Operation and Live-Line
compromised insulation inside the transformer might increase Maintenance (ESMO), Providence, RI, 2011, pp. 1-7.
the possibility of short circuit between each winding or even [10] IEC 60071-1 Ed. 1: Insulation co-ordination – Part 1: Definition, priciples
between winding and ground. Since the voltage and current and rules, 2006
waveform from the result of short circuit simulation between [11] "IEEE Guide for Tank Rupture Mitigation of Liquid-Immersed Power
Transformers and Reactors," in IEEE Std C57.156-2016 , vol., no., pp.1-
primary winding phase R and ground (Fig. 15 and Fig. 16) was 26, Dec. 21 2016
identical to the DFR record, it’s predicted that the possible cause
[12] N. Abi-Samra et al., "Power Transformer Tank Rupture and Mitigation—
of transformer failure was internal arcing on phase R of primary A Summary of Current State of Practice and Knowledge by the Task
winding (150 kV) to ground. Force of IEEE Power Transformer Subcommittee," in IEEE Transactions
on Power Delivery, vol. 24, no. 4, pp. 1959-1967, Oct. 2009.
The presence of electrical arc inside transformer will
[13] J. B. Dastous, J. Lanteigne and M. Foata, "Numerical Method for the
increase the oil pressure beyond the structural strength of the Investigation of Fault Containment and Tank Rupture of Power
tank. According to the reference [11-15], internal arcing during Transformers," in IEEE Transactions on Power Delivery, vol. 25, no. 3,
transformer fault could deform or even rupture the transformer pp. 1657-1665, July 2010.
tank which was similar to what happened in Fig. 1. Rapid [14] G. Perigaud, S. Muller, G. de Bressy, R. Brady and P. Magnier,
increase of oil pressure might rupture the tank and shatter the "Contribution to the study of transformer tank rupture due to internal
lead bushing, which explained the missing bushing in Fig. 1. arcing development of a computer simulation tool," 2008 IEEE Power
and Energy Society General Meeting - Conversion and Delivery of
With the bushing shattered, there would be short circuit between Electrical Energy in the 21st Century, Pittsburgh, PA, 2008, pp. 1-8.
transformer lead and tank which caught transformer on fire. The [15] C. Yan et al., "Simulation and analysis of power transformer internal
excessive heat from the burning transformer was radiated to the arcing faults overpressure characteristics," 2014 International
surrounding area and damaged the high voltage apparatus Conference on Power System Technology, Chengdu, 2014, pp. 685-690.
nearby including 66 kV arrester as shown in Fig. 2.
VI. CONCLUSION
According to data and simulation result analysis, lightning
activity has no contribution toward the failure of the 150/66 kV
100 MVA transformer. The possible cause of transformer
breakdown was internal failure in form of arcing between
primary winding phase R to ground.

201