8th International Conference on the Properties and Applications of Dielectric Materials (ICPADM 2006), Bali, Indonesia, June 2006

Effect of Non-soluble Contaminants on The Flashover Voltages of

Hydrophobic Polymer Insulators

Basanta Kumar Gautam, Takeshi Yamazaki, Kazuma Yamada, Shuhei Matsushita,

Kenji Sakanishi and Ryosuke Matsuoka
Chubu University, 1200 Matsumoto-cho, Kasugai, Aichi, 487-8501 Japan
* E-mail : y03009@isc.chubu.ac.jp

ABSTRACT: Different non-soluble materials are used
as non-soluble contaminants in the contamination
flashover voltage tests of insulators. Flashover voltage
of a porcelain insulator is influenced by the kind of
non-soluble contaminants. Kind of non-soluble
contaminants may have a significant effect on the
flashover voltage performance of polymer insulators. In
this paper we present comparative investigation results
of the flashover voltage and hydrophobicity recovery
characteristics among three most common non soluble
contaminants like Tonoko, Kieselguhr and Kaolin.
Index terms: Po lymer insulator, non-soluble
materials, hydrophobicity, contamination flashover
voltage
INTRODUCTION
Hydrophobic polymer insulators are increasingly
adopted in many countries and regions for many
advantages they offer over conventional porcelain and
glass insulators. So evaluation of contamination
flashover voltages of these insulators is very important
for the rationalized insulation design of power
transmission systems.
Solid layer method or clean fog method is one of the
popular insulator contamination test methods. In this
method, specimen insulators are contaminated by using
a solution of a constant amount of non-soluble
contaminant and different amount of salt. After drying,
insulators are subjected to a test voltage and wetted by
artificial fog. Different kinds of non-soluble materials
are used in different laboratories. Most commonly used
among them are Tonoko, Kieselguhr and different
types of Kaolin.
Insulator contamination flashover voltages are
influenced not only by the kind of soluble contaminants
mostly represented by salt, but also by the kind of
non-soluble contaminants. DC withstand voltage of a
porcelain insulator contaminated with Tonoko as
non-soluble contaminant has been found to be 20-30%
higher than that with Roger’s Kaolin [1]. Similarly in
another investigation, DC 50% flashover voltage of
porcelain insulators with Tonoko as non-soluble
contaminant has been found to be higher than that with
five different types of Kaolin [2]. Smaller but
discernible differences were confirmed also under AC
voltage [3].
In the case of polymer insulators having the housing
material of silicone rubber, hydrophobicity can be
observed even on the contaminated surface due to
migration of low molecular weight silicone from the
bulk. The rate of hydrophobicity recovery may be
different among different non-soluble contaminants.
We have been investigating the contamination
flashover voltage performance and hydrophobicity
recovery condition of hydrophobic polymer insulators
contaminated with Tonoko, Kieselguhr, Mexican
Kaolin and Roger’s Kaolin as non-soluble
contamination materials. We present some initial
results of our investigation in this paper.
SPECIMENS
A transmission line polymer insulator with hydrophobic
housing was chosen to investigate the effect of
non-soluble contaminants on the contamination
flashover voltage. The specimen dimensional details
are shown in Table 1. For the investigation of
hydrophobicity recovery on the surface of silicone
rubber housing materials contaminated with different
non-soluble contaminants, rectangular silicone rubber
sheets with the dimensions of 16 cm ×4.5 cm×5 mm
thick were also prepared.
TEST METHOD AND TEST ARRANGEMENT
A contamination solution consisting of 10 g/l of the
non-soluble material and 60 g/ l of salt was prepared
for individual non-soluble materials for the evaluation
of contamination flashover voltage of a transmission
line polymer insulator. Specimen insulators were
contaminated after preconditioning according to the
method developed by NGK for obtaining a uniform
contamination layer [4]. Non-soluble contaminant
deposit density (NSDD) and salt deposit density (SDD)
were measured on the sheds at different positions along
8th International Conference on the Properties and Applications of Dielectric Materials (ICPADM 2006), Bali, Indonesia, June 2006

the contaminated insulator. These values were within

Table 1 Dimensional Particulars of the
Specimen Insulator

Contamination and Drying

Specimen Installation in
50 60 min.
Test
Shed Shape, mm

Preconditioning
50 Voltage
Withstand

Fog Chamber

Start of Fog
Injection
Effective Height , mm 660
F.O
Leakage Distance, mm 1530
Core Diameter, mm 26

±15% of the target values of NSDD and SDD which
were 0.1mg/cm2 and 0.12 mg/cm2, respectively.
Contaminated and dried specimens were then subjected
to an AC test voltage under heavy fog conditions in the
fog chamber and 50% flashover voltages were
determined for individual non-soluble contaminants.
The test arrangement and test procedure for the
evaluation of the contamination flashover voltage of the
polymer insulator are shown in Figure 1 and Figure 2
respectively.
Rubber sheets were contaminated by the same method
as insulator specimens. But in this case, three types of
contamination solutions were prepared. Amount of
non-soluble contaminant was the same for all the three
solutions but amount of salt was 0 g/l, 15 g/l and 60 g/l
respectively. Receding contact angles of sessile water
drops were measured on the rubber sheet contaminated
with individual non-soluble contaminants.
5.7 m
Nozzles for Fog
Bushing
6.3 m
Specimen
Plan
Time
Figure 2 Test Procedure
TEST RESULTS AND DISCUSSIONS
I. SURFACE CONDITIONS
Enlarged photographs of contaminated surface on the
rubber sheet with different non-soluble contaminants
are shown in Figure 3. A more uniform surface
contamination is obtained with Kaolins compared with
Tonoko or Kieselguhr.
II. FLASHOVER VOLTAGES
Comparative 50% flashover voltages of the polymer
insulator contaminated with the same contamination
degree but with different non-soluble contaminants are
shown in Figure 4. As all the tests were made under
hydrophilic conditions, any difference in flashover
voltage may be attributable to the difference in the
wetting characteristics among the non-soluble
contaminants. The wetting characteristics depend on
the particle size, uniformity of contaminant layer on the
insulator surface and the kinds of main minerals
included in the non-soluble contaminants. 50%
flashover voltage with Tonoko is slightly higher than
the other three contaminants investigated. This trend is
similar to that reported for porcelain insulators [3].
III. HYDROPHOBICITY RECOVERY

The hydrophobicity or the ability to repel water is one

of the most important properties of polymer insulators.
6.1 m A high degree of hydrophobicity means low wettability
on the insulator surface. Measurement of contact
Specimen angles is one of the widely approved methods for
Bushing
Nozzles for Fog evaluating the surface hydrophobicity. In contaminated
polymer insulators, low molecular weight silicone
migrates from the bulk to the surface and covers over
Elevation the contamination layer, making the contaminated

Figure 1 Test Arrangement

8th International Conference on the Properties and Applications of Dielectric Materials (ICPADM 2006), Bali, Indonesia, June 2006

surface hydrophobic again. The rate of this recovery of
hydrophobicity may be different among different
(i) Tonoko
Its receding contact angle is even larger than the value
for the new specimen. The reason for such a rapid
recovery may be explained by the fact that the
contamination layer with Kieselguhr is microscopically
less uniform as shown in Figure 3. But more
investigation is necessary for finding out the cause of
hydrophobicity recovery to a level more than that for
the new specimen. Recovery for Tonoko and Mexican
Kaolin are comparable. No significant recovery in
hydrophobicity was obtained with Roger’s Kaolin even
(ii) Kieselguhr after 24 hours. This may be attributable to the more
uniformity of the contamination layer in the case of
Roger’s Kaolin compared with the other contaminants.

Specimens were then contaminated with a

contamination solution containing both non-soluble
contaminant and salt. Measurements of receding
contact angles were taken at different time periods until
one week in this case. As shown in Figure 6, recovery
(iii) Mexican Kaolin (iv) Roger’s Kaolin is better with Kieselguhr and not significant with
Figure 3 Enlarged Contaminated Rubber Surface (×50) Roger’s Kaolin.
(Non-soluble material: 10 g/l, Salt: 60 g/l)
From Figures 5 and 6, we can also see that the recovery
120 in hydrophobicity is very rapid during the first 24
Comparative Flashover Voltages

hours after the contamination. The recovery is more

10
gradual after 24 hours. Similarly, we can see that the
(100% for Tonoko)

8 recovery in hydrophobicity is less when contaminated

60 with the solutions containing salt compared with the
case without salt. Hydrophobicity recovery conditions
4 on the rubber surface at 3 hrs and 24 hrs after
2 contamination and drying are compared under all the
0
three contamination conditions as shown in Figure 7. It
Tonoko Kieselguhr Mexican Roger’ can be seen from this figure that hydrophobicity
Kaolin s recovery is larger on the contaminant surface without
Type of Non-soluble Contaminants salt for all the non-soluble contaminants. The
contamination layer on the silicone rubber surface may
Figure 4 Effect of Type of Non-soluble be more uniform if salt exists in the contamination
Contaminants layer resulting in the slower hydrophobicity recovery
on 50% Flashover Voltage of Polymer rate. On the other hand, no significant difference can
Insulators be seen between the two cases where the amounts of
(NSDD:0.1 mg/cm2, SDD:0.12 mg/cm2) salt are 15 g/l and 60 g/l in the contamination solution,
non-soluble contaminants used for the evaluation of which correspond to SDD values of 0.03 and 0.12
contamination performance of polymer insulators. mg/cm2, respectively.

In our investigations, rectangular silicone rubber sheets 120

Receding Contact Angle, °

were contaminated with different non-soluble Kieselguhr

contaminants and contamination degrees. Time 100
variation of the hydrophobicity on the contaminated 80 New and not contaminated
surface after contamination was checked by measuring
60 Tonoko
receding contact angles of sessile water drops.
Measurements were made by using a contact angle 40 Mexican Kaolin
meter. 20 Roger’s Kaolin
At first, time variation of hydrophobicity was measured 0
on the specimens contaminated with non-soluble 0 5 10 15 20 25 30
Time, Hrs
contaminants only. The results are shown in Figure 5.
Hydrophobicity recovery was very rapid in the case of Figure 5 Hydrophobicity Recovery on Contaminated
Kieselguhr compared with the other three contaminants. Silicone Rubber Surface with Different
Non-soluble Contaminants (Without Salt in
the Contamination Solution)
8th International Conference on the Properties and Applications of Dielectric Materials (ICPADM 2006), Bali, Indonesia, June 2006

3) Hydrophobicity recovery is larger on the

contamination layer without salt.
100
Receding Contact Angle, °

New Sample 10

Receding Contact Angle, °.

80 Surface Kieselguhr
80
Mexican Kaolin Kieselguhr
60
60
Tonoko
40
40 Tonoko
Mexican
20 Roger’s Kaolin Kaolin
20
Roger’s Kaolin

0 0
0 50 10 15 20 0 0.05 0. 0.1
Time, Hrs SDD, mg/cm2
(i) SDD: 0.03 (Amount of Salt=15 g/l) (i) 3 hrs. after Drying
12
100

Receding Contact Angle, °.

Receding Contact Angle, °

New Sample Surface 100

Kieselguhr
80 Kieselguhr
80

60 Mexican Kaolin 60 Tonoko

Tonoko
40 Mexican
40 Kaolin
Roger’s Kaolin
20 Roger’s
Kaolin
20 0
0 0.05 0. 0.15
0 ESDD, mg/cm2
0 50 10 15 20
(ii) 24 hrs. After Drying
Time, Hrs
(ii) SDD: 0.12 (Amount of Salt=60 g/l) Figure 7 Hydrophocity Recovery among Different
Non-soluble Contaminants
Fi gur e 6 Time Variatio n in Hydrophobicity
of Silicone Rubber Surface Contaminated
with Different Non-soluble Contaminants
and Salt

REFERENCES

CONCLUSIONS [1]Y. Hasegawa, K. Naito, K. Arakawa, H.M.

Schneider and L. E. Zaffanella, “A Comparative
Effects of kind of non-soluble contaminants on the Program on HVDC Contamination Test”, IEEE
flashover voltage and hydrophobicity recovery of Trans. on Power Delivery, Vol. 3, No.4, pp.
hydrophobic polymer insulators were investigated. The 1986-1993, 1988.
main conclusions are as follows; [2] R. Matsuoka, O. Kaminaga, K. Kondo, K. Naito, Y.
1) Differences in the contamination flashover Mizuno and H. Kusada, “Influence of Kinds of
voltage of hydrophobic polymer insulators can be Insoluble Contaminants on Flashover Voltages of
seen among the most popular non-soluble Artificially Contaminated Insulators”, 9th
contaminants used all over the world. It is International Symposium on High Voltage
modestly higher in the case with Tonoko as the Engineering”, August 28 – September 1, 1995,
non-soluble contaminant. Austria.
2) There are very significant differences in the [3] R. Matsuoka, K. Kondo, K. Naito and M. Ishii,
hydrophobicity recovery among these non-soluble “Influence of Non-soluble Contaminants on the
contaminants. Owing to the difference in the Flashover Voltage of Artificially Contaminated
uniformity in the contaminated surface, Insulators”, IEEE PES Paper No. WM
hydrophobicity recovery is largest in the case of 249-3-PWRD, 1995.
Kieselguhr and smallest in the case of Roger’s [4] K. Naito, K. Izumi, K. Takasu and R. Matsuoka,
Kaolin. “Performance of Composite Insulators under
8th International Conference on the Properties and Applications of Dielectric Materials (ICPADM 2006), Bali, Indonesia, June 2006

Polluted Conditions”, CIGRE Session Paper No.

33-301, 1996.