Paris 2020

D1-211

Analysis of 400 kV Failed Silicone Rubber Insulators: Role of Micro-Cracks in Glass Fiber
Rod and Electric Field Distribution in Failure Mechanism – Case Study

N.R. SHINGNE*, S.J. BAROT, U.N. PUNTAMBEKAR, S.H. CHETWANI

Electrical Research and Development Association (ERDA), Vadodara 390010, India

J. M. THAKKAR, S. SHAH, A. SADARIA

Adani Transmission Ltd., Ahmedabad 382421, India

SUMMARY

In this paper, root cause analysis of failed composite insulators is presented. The failed insulators
were installed on a 400 kV AC transmission line located in western part of India. The transmission
line having 282 km length was facing frequent failures of composite insulators having service life
of 8 years. All the failed insulators were fractured at the line end. At the fracture point the glass
fiber rod had become black in colour and showed clear sign of degradation. The silicone rubber
sheath from all the failed insulators were punctured near line end of insulators. The failure
investigation was carried out using various material characterization and testing methods. Stereo
microscopic analysis showed that the failed insulators were having micro-cracks in the glass fiber
composite (GFC) rod making them susceptible to partial discharge (PD). The cracks in the
composite rods were more prominent on the line end than the tower end. Our analysis of
insulators indicate combination of increased electrical stress on the insulators and micro-cracks in
GFC rod which lead to partial discharge (PD) in composite insulator. Localised heating due to PD
lead to punctures in the silicone sheath, degradation of GFC rod and eventual failure of insulator.

KEYWORDS
Composite Insulators, Silicone Rubber Insulator, Glass Fiber Composite, Microstructure, Failure of
Composite Insulator

nitin.shingne@erda.org

1
1. INTRODUCTION
Silicone rubber insulators are rapidly replacing ceramic insulators due to their light weight,
toughness, and excellent resistance to pollution. All these advantages notwithstanding,
composite insulators can fail during their service life, such failures are costly and reduces the
reliability of transmission/distribution network. Failures in composite insulators can be caused
by tracking & erosion, corona discharges, mechanical failure, flash under and brittle fracture
of glass fiber composite rod. Several such cases of insulator failures are reported elsewhere
[1]–[3]. However, our investigation of failed composite insulator shows that a new mechanism
accountable for failure of composite insulators. The failed insulators were installed on a 400
kV transmission line located in Western part of India in the state Gujarat. The insulators were
installed in tension as well as suspension configuration. The transmission line where failures
were noticed is 282 km long and it does not pass through a polluted or coastal area which may
exclude environmental conditions for the failures. The failure was detected by visual
inspection of transmission line after tripping.
Several composite insulators failed on the transmission line with operating life of the
insulators of approximately 8 years. The entire transmission line is having composite
insulators from two different manufacturers, differentiated as Make-1 and Make-2 insulators.
In the 282 km transmission line Make-1 insulators were installed on 40 km of line and the
remaining 242 km transmission line is having Make-2 insulators. Interestingly all the failures
were observed only in Make-1 insulators. The failed insulators showed punctured silicone
rubber sheath and broken glass fiber rod near the line end of insulator. Tower end did not
show any punctures in silicone sheath. As a part of preventive action, thermography was
carried out using infra-red (IR) camera to measure hot spot temperature on insulator surface.
The hot spot measurements showed that the temperature at the line end of insulator is
higher by 0.6°C - 18°C compared to the temperature at tower end of insulator. The hot spot
measurements indicate localised heating at the line end of composite insulator.
Recent studies of composite insulator failures of punctures in silicone sheath has indicated a
new mechanism of failure resulting from weak interface between silicone sheath and glass
fiber composite rod [4], [5]. However, in our study of failed insulators revealed that defects in
glass fiber rod played a major role in the failure. Problems related to interface between
silicone rubber sheath and composite rods were not evident.
After the insulators started failing, the working insulator from transmission line was subjected
to routine type and design tests carried out as per IEC-62217 and IEC-61109, the insulators
passed all the tests. Therefore, to understand the root cause of failure, we carried out
comprehensive testing of silicone rubber as well as glass fiber composite (GFC) rod from
failed insulators as well as unused insulators from same batch. FEM simulations were carried
out to evaluate electrical stress on the insulators to understand any relationship between
electrical stress and failures.
2. FAILURE DETAILS AND VISUAL OBSERVATIONS
For easier understanding of failure details and analysis, Figure 1 shows a sketch of composite
insulator indicating different components. At tower end, the insulator is connected to the
transmission tower and the line end is connected with the live conductor. To reduce the
electrical field on the insulator, corona rings are usually attached at both ends for composites
insulators designed for transmission line of more than 220 kV.
For root cause investigation of failure several samples of composite insulators were collected
from transmission line and brought to research laboratory. The investigation results of few
selected samples are reported in this paper. All the failed insulators were broken within 1
meter from the tri-junction at line end of insulators.

2
 Figure 1. Representative sketch of composite insulator

At the breakage point no indication of brittle fracture was observed, the glass fiber strands
were separated from each other and fiber colour looked brown and black, as shown in Figure
2a. Closer inspection of the insulator showed punctures in the silicone sheath, Figure 2b and
2c. The punctures in the silicone sheath were located within 1 m distance from tri-junction at
line end. The tower end of insulator showed no indication of punctures. Removal of silicone
sheath, from line end of insulator, revealed degradation of GFC. However, at the tower end of
same insulator, the GFC showed no indication of degradation as shown in Figure 3. Visual
analysis observations showed that insulator degradation only takes place at the line end and
the tower end remains unaffected.

Figure 2. Silicone insulator showing a) broken end of insulator; b) silicone sheath with
punctures and c) enlarged view of puncture on silicone sheath (yellow circles shows
punctures)

Figure 3. Glass fiber composite (GFC) rod from failed insulator after removal of silicone sheath
a) GFC at line end of insulator showing degradation and b) GFC at tower end of insulator
shows no indication of degradation

3. ANALYSIS OF MATERIALS FROM COMPOSITE INSULATOR

As mentioned in introduction, recurring failures were observed only in one particular Make-1
(manufacturer-1) of insulators. Other insulator of Make-2 (manufacturer-2) did not fail since
installation on the same transmission line. Therefore, the failure site also provided a reference
insulator for study having a good service history without failure. Comparative analysis of

3
failed insulator and unused insulator was carried out for understanding any ageing effects on
the materials. The samples used for investigation are shown in Table 1. Material
characterization carried out separately for near the punctured area and away from punctured
area in order to understand actual effect of defect versus material ageing playing a role in
failure.
Table 1. Composite insulator analysed for failure investigation
Insulator Insulator Detail Insulator
Abbreviation Make
Insulator-A Failed insulator Make-1
Insulator-B Working insulator removed from transmission line after Make-1
detection of hot spot temperature of more than 10°C
Insulator-C Unused insulator from same batch and age Make-1
Insulator-D Unused spare insulator from manufacturer-2 with similar age Make-2

3.1. Analysis of Silicone Rubber

Silicone rubber sample was collected from shade of Insulator-A. Figure 4 shows the ATR-
FTIR spectra of silicone rubber, the peaks near 3435 cm-1 showing hydroxyl peaks
associated with ATH fillers in silicone rubber. The Si-CH3 stretching peak is located at 1259
cm-1, the peaks at 1081 – 1007 cm-1 indicate Si-O-Si stretching vibrations. Overall, the FTIR
spectra did not reveal any abnormality or degradation related to corona activity on the
silicone rubber shades.
Thermogravimetry analysis (TGA) was used for analysing silicone and filler content of
silicone rubber from the insulator shades. The TGA results showed that the silicone
rubber used in insulator shades have silicone content of 37%, filler content approximately
48%, and remaining 15% is the low molecular weight plasticizers. TGA results did not reveal
any abnormality in the silicone compound. Arc resistance of silicone rubber was
measured as per ASTM D495:14. This test measures the ability of insulating material to
resist the arc. Test results from all the insulator samples showed arc resistance more
than 250 seconds compared to the recommended 200 seconds. Basic tests did not reveal
any ageing related problems in the silicone rubber compound used in the weathering
shade of composite insulator.

Figure 4. ATR-FTIR spectra of silicone rubber from insulator – A (failed insulator)

FTIR and EDS analysis did not show any degradation of silicone rubber shades from
the failed insulator. Most of the tests on silicone rubber collected away from the

4
 punctured location did not reveal any deterioration of properties. These tests ruled
out role of silicone rubber in the failure.

3.1.1. Analysis of Silicone Rubber Near Puncture Location

Figure 5 shows the punctured silicone rubber sheath removed from failed
insulators. Under the punctured surface the silicone sheath shows degradation,
Figure 5b. Punctures in the silicone rubber can accelerate the degradation process
due to direct water ingress to the GFC rod.

Figure 5. Punctured silicone sheath removed from failed insulator, a) outer view of
puncture and b) inner puncture surface
Silicone rubber located near the puncture location was analysed using FTIR and
TGA. Figure 6 shows ATR-FTIR spectra of silicone rubber near the puncture
location. Apart from the characteristic FTIR peaks of silicone rubber, the spectra
shows additional peak at 1348 cm-1 indicating presence of nitric acid group on the
sample surface [6]. Also, broad peak between 3800 – 3000 cm-1 indicate presence
of –OH groups or absorbed water. It is possible that near the puncture location
the silicone rubber is partially degraded which absorbs increased content of water.
TGA analysis of inner silicone rubber near the puncture revealed high filler content
of 55% compared to 48% filler content for silicone rubber collected from shades.
The high filler content indicated partial degradation of silicone rubber.

Figure 6. ATR-FTIR spectra of silicone rubber from insulator – A (failed insulator)

Degradation by-product on GFC rod, near the puncture location, was analysed
using Energy Dispersive Spectroscopy (EDS). Punctured Silicone sheath was
removed from GFC rod and the underlying rod had degraded mass (Figure 7a)
which was collected and analysed using SEM-EDS. The EDS spectra showed large
percentage of nitrous compound in the degraded sample, again indicating
presence of nitric acid. Production of nitric acid due electrical corona discharges in
presence of water and oxygen is known phenomenon. Therefore, it appears that
after the silicone sheath is punctured water entered near to the silicone – GFC

5
 interface and in presence of electric filed lead to accelerated degradation of GFC
rod leading to failure.

Figure 7. EDS analysis: a) degraded by-product found on glass fiber rod below a
punctured location, b) SEM image of degraded dust also showing broken glass fibres.
EDS spectrum (below image-a) and elemental analysis table (below image-b)

3.2. Analysis of Glass Fiber Composite Rod

To evaluate quality of GFC rods of composite insulators, water diffusion test and dye
penetration tests were carried out as per IEC 62217.
3.2.1. Water Diffusion Test
In water diffusion test, the GFC rod samples of 3 cm height are boiled in 0.1 wt% NaCl
solution for 100 hours. After boiling water treatment, sample is placed between two
electrodes and leakage current is measured while voltage is gradually increased to 12
kV. Leakage current should not rise above 1 mA during stepwise rise of voltage.
Leakage current above 1 mA indicates that the sample is failed in the test. GFC samples
were collected from line end and tower end of failed insulator-A. Rest of the samples
were collected from line end of insulators. Figure 8 shows the result of water diffusion
test, all the samples collected from the line end of Insulator-A failed in the test. All the
other samples showed very low leakage current including sample from tower end of
failed insulator-A.

Figure 8. Water diffusion test results of GFC rod with leakage current values

6
3.2.2. Dye Penetration Test
Water diffusion test was carried out as per IEC 62217. The samples were collected from
all the insulators described in Table 1. Separate samples were collected from line end
and tower end of Insulator-A. All the samples passed the dye penetration test except
samples from line end of Insulator-A, as shown in Figure 9.

Figure 9. Water diffusion test results from sample of GFC rod from a) line end of
Insulator-A, b) tower end of Insulator-A and c) line end of Insulator-C
Both type the tests, water diffusion and dye penetration, performed on GFC rod of
insulators did not reveal any problem in working insulators. Samples in both the tests
failed only for samples collected from line end of failed insulators. All the GFC rod
samples from working insulators conform to the standard water diffusion and dye
penetration tests.
3.2.3. Optical Microscopy Observations
Since standard test methods did not reveal any problems either in the silicone rubber
or the GFC rod, we investigated the morphology of GFC rod using optical microscopy.
For optical microscopy observations, GFC rod were cut in 10 mm thick sections
without removing the silicone rubber sheath. The GFC sections were then observed
under stereo microscope. Figure 10 shows area near the interface between GFC rod
and silicone rubber, it does not reveal any gap or delamination of the interface for the
unused Insulator-C. Delamination of interface was not visible in the failed insulator as
well, Figure 11, for the area investigated away from the punctured location. However,
number of micro-cracks are visible in the GFC rod. Several such micro-cracks were
visible in all the sections cut from Insulator-A, as shown in Figure 12. The micro-cracks
were even observed in the unused Insulator-C as shown in Figure 13. This indicated
that the micro-cracks in GFC rod formed possibly during manufacturing process. In
insulator-B, which was removed from service due to hot spots, large number of micro-
cracks were observed having thickness of 50 to 100 µm, shown in Figure 14a. GFC rod
samples collected from Insulator-D, Make-2 having good service history, does not
show any indication of micro-crack in Figure 14b. The microscopic observations also
showed that the micro-cracks in failed insulator-A (Figure 12) are thicker compared
micro-cracks from insulator-c (Figure 13).

Figure 10. Stereo microscope images of: a) Silicone rubber with GFC core rod from
Insulator-C and b) enlarged view at interface between silicone and GFC rod

7
Figure 11. Stereo microscope images of section of Insulator-A showing region near
silicone – GFC rod interface, which is intact while the GFC rod shows micro-cracks.

Figure 12. Stereo microscope images of Insulator-A: a) Silicone rubber with GFC rod
showing micro-cracks and b) enlarged view showing many cracks in GFC, bigger cracks
are shown with yellow arrows

Figure 13. Stereo microscope images of Insulator-C GFC rod with different
magnifications showing pre-existing micro-cracks in unused insulator
The pre-existing micro-cracks in un-used GFC rod and gradual growth of number of
micro-cracks with increasing thickness in-field ageing conditions indicated different
mechanism that gradually leads to failure of GFC. It is known that the voids in
electrical field can lead to partial discharge (PD) in an insulating medium. PD in
void/cracks takes place due to change in permittivity at the crack gap which causes
ionisation of gaseous medium leading to gradual degradation of electrical insulation
[7]. It is reported that for same volume of cavity, the elongated cavity will have lower

8
 inception voltage and higher magnitude of PD compared to a spherical cavity [8].
Micro-cracks in GFC are naturally elongated rather than spherical, therefore the
inception voltage required for PD is lower and PD magnitude will be high resulting in
accelerated ageing and degradation. Electric field at the outer circumference, near
silicone rubber interface, will be higher compared to centre of GFC rod. Therefore, PD
will prominently get concentrated near the silicone rubber interface. It is observed
that the crack growth is larger near the interface in the failed insulators as shown in
Figure 12.

Figure 14. Stereo microscope images GFC rod from: a) Insulator-B showing large
number of micro-cracks and b) Insulator-D showing no micro-crack

The micro-cracks in the GFC rod indicated towards PD responsible for accelerated
ageing and degradation of insulators. To understand the PD activated degradation
mechanism in the composite insulators, we analysed the electrical field distribution on
the insulator using finite element simulations.

4. ANALYSIS OF ELECTRICAL STRESS ON INSULATOR

In order to minimise discharge activity on the insulators, it is recommended that the electric
field on the weather shed near metallic fitting should be less than 0.45 kV/mm [9]. Therefore,
composite insulators and electrical stress control devices such as corona rings are designed to
minimize the electric stress. However, improper design of composite insulators or misplaced
corona rings can result into enhance electrical field on the insulator surface. FEM simulations
carried out from specifications of failed composite insulator design showed electric field
distribution within the limits as recommended by IEEE task force [9].

Figure 15. a) Image showing broken corona ring on transmission line indicated by arrow and
b) image shows corona rings are reversed and incorrectly placed away from tri-junction.

9
Corona ring positions on composite insulators are much more fluid and an untrained worker may
fix the corona ring at the wrong position. Figure 15a, shows a case where one corona ring is
broken. In Figure 15b the corona ring is fitted away from the tri-junction and is wrongly oriented in
opposite direction to insulator shades. In such a case the electric field may increase on the
insulator. As per the insulator design corona ring is placed near the tri-junction and it should cover
the first shade of insulator, this is a ‘Normal’ case, shown in Figure 16a. In this normal case, FEM
simulation results showed that at the insulator tri-junction electric field strength is 0.47 kV/mm.
On the other hand if the corona ring is misplaced farthest from the tri-junction then electric field
at the tri-junction increases to 0.92 kV/mm, shown in Figure 16b. Furthermore, it may happen that
the orientation of corona ring is reversed (Figure 15b) or position of line end corona ring is
exchanged with tower end corona ring, where corona ring sizes are different. In each of the
cases, the corona ring may be placed near to the tri-junction or farthest from tri-junction. All these
cases of corona ring placement were studied in the FEM simulation as described in Table 2.

Figure 16. FEM simulation for corona ring position: a) ‘Normal’ case, corona ring placed nearest to
tri-junction and b) corona ring misplaced at farthest distance from tri-junction

Table 2: Different corona ring positions considered for FEM analysis

Case Corona Ring CR Distance

Case Details
No. Position from Tri-joint

1 Nearest CR as per design, placed nearest to TJ

Normal
2 Farthest CR placed farthest from TJ
3 Orientation Nearest CR faces are reversed and placed nearest to TJ
4 reversed Farthest CR faces are reversed and placed farthest from TJ
5 Position Nearest LE & TE CR rings are exchanged & placed nearest to TJ
6 exchanged Farthest LE & TE CR rings are exchanged & placed farthest from TJ
7 None No CR No corona rings are placed on the insulator

Abbreviations
CR Corona Ring
Tri-joint: Position on insulator where metal rod, FRP rod
TJ
and silicone insulation are inter-connected
LE Line end of insulator
TE Tower end of insulator

10
FEM simulation results showed that at each position of corona ring (described in Table-2) the
electric field on insulator changes as shown in Figure 17. The electric field at the tri-junction on line
end of insulator increases from 0.47 kV/mm for normal case to 4.99 kV/mm in the worst case
where the corona ring is not attached. Almost 10 time increase in the electric field can cause
severe damage to the insulator. The FEM simulation results show that in some of the
configuration of corona ring position on insulator there is sufficient electric field available for PD
inception in the insulators.

Figure 17. Electric field stress at tri-junction of composite insulator for various position of
corona ring as per Table 2 (F-farthest position of corona ring from tri-junction)

5. CONCLUSION
Investigation of failed composite insulators showed that combination of micro-cracks and
electrical field at line end of composite insulator are responsible for the failures. Analysis of
failed insulator showed that the glass fiber composite (GFC) rod had micro-cracks. These
micro-cracks act as a site for partial discharge (PD) leading to punctures and degradation of
insulators. The micro-cracks are elongated cavities in the solid insulator rod requiring lower
inception voltage and have higher magnitude of PD compared to spherical void. Increased
electrical field on the insulator due to misplaced corona ring can be an added factor to
increase the possibility of PD. The constant discharges near the interface of silicone rubber
and GFC rod causes localised heating and progressive degradation the silicone rubber leading
to punctures in silicone sheath. Once the silicone sheath is punctured, water and other
impurities can enters in GFC rod leading to increase in leakage current and resultant heating
and degradation. This accelerates the degradation of GFC rod leading to final failure.
Based on the failure investigation results it is recommended that along with type test of
composite insulator, microscopic investigation of GFC rod should be included in the test plan
to evaluate quality of composite insulator. FEM simulation results showed the importance of
corona ring placement; misplaced corona ring can increase the electrical stress on insulator
beyond the recommended limits. Installation technicians should be trained to strictly follow
the design parameters for fitting the corona rings.

11
BIBLIOGRAPHY

[1] Y. Gao, X. Liang, W. Bao, S. Li, and C. Wu, “Failure analysis of a field brittle fracture
composite insulator: Characterization by FTIR analysis and fractography,” IEEE Trans.
Dielectr. Electr. Insul., vol. 25, no. 3, pp. 919–927, 2018.
[2] Y. Zhu, K. Haji, M. Otsubo, and C. Honda, “Surface degradation of silicone rubber exposed
to corona discharge,” IEEE Trans. Plasma Sci., vol. 34, no. 4 I, pp. 1094–1098, 2006.
[3] W.-B. 2. Group and CIGRE, “Guide for the Assessment of Composite Insulators in the
Laboratory after their Removal from Service,” 2011, no. December, pp. 1–62.
[4] B. Lutz, L. Cheng, Z. Guan, and L. Wang, “Analysis of a Fractured 500 kV Composite
Insulator – Identification of Aging Mechanisms and their Causes,” IEEE Trans. Dielectr.
Electr. Insul., vol. 19, no. 5, pp. 1723–1731, 2012.
[5] J. Wang, X. Liang, and Y. Gao, “Failure Analysis of Decay-like Fracture of Composite
Insulator,” IEEE Trans. Dielectr. Electr. Insul., vol. 21, no. 6, 2014.
[6] A. R. Chughtai, D. M. Smith, and L. S. Kumosa, “FTIR Analysis of Non-ceramic Composite
Insulators,” vol. 11, no. 4, pp. 585–596, 2004.
[7] L. A. Dissado and J. C. Fothergill, “Overview of Electrical Degradation and Breakdown,” in
Electrical Degradation and Breakdown in Polymers, G. C. Stevens, Ed. The Institution of
Engineering and Technology, London, United Kingdom, 2008, pp. 49–66.
[8] E. A. Cherney, “Partial discharge. V. PD in polymer-type line insulators -,” IEEE Electr. Insul.
Mag., vol. 7, no. 2, pp. 28–32, 1991.
[9] A. Phillips et al., “Electric fields on AC composite transmission line insulators,” IEEE Trans.
Power Deliv., vol. 23, no. 2, pp. 823–830, 2008.

12