21, rue d’Artois, F-75008 PARIS D1-110 CIGRE 2016

http : //www.cigre.org

Long-term performance of composite station insulators with larger diameters:

laboratory tracking and erosion test vs. service experience

I. GUTMAN J.-F. GOFFINET J. SEIFERT F. SCHMUCK

M. GULLO
M. GÅRDESTEDT
S. BUCAN
STRI Elia LAPP Insulators Pfisterer Sefag
Sweden Belgium Germany Switzerland

SUMMARY

At present more than 1.5 million station post and apparatus composite insulators are in service in AC
and DC substations. They have good service records from both a pollution point of view and ageing
(i.e. long-term performance). At the design stage all composite insulators, including station post and
hollow core insulators with large diameter, should be verified by tracking and erosion tests. The main
choice for such tests is between 1000 hours salt fog test (IEC 62217) and 5000 hours test at multiple
stresses (IEC/TR 62730). There are a number of differences in these two main types of tracking and
erosion tests. The pure salt fog test is at present well defined and takes into account differences in
insulator diameter and creepage factor to prevent flashovers during the test (i.e. intending to perform
more “ageing” than “pollution” test). These specific improvements of the salt fog phase of test were
not introduced in the procedure for 5000 hours multiple stresses test as per IEC/TR 62730.
Furthermore, in 5000 hours test the salt fog must be directly sprayed on the test specimens. Experience
with many 5000 hours tracking and erosion tests of insulators with different diameters is compiled and
analyzed in this paper. The results indicate that due to uneven salt fog distribution around and along
the insulators some test objects are severely damaged or the test is multiply interrupted by flashovers,
even causing damages. These observations are in contradiction to service experience. A specially
developed test with two sets of insulators with principally different diameters (the trunk diameter 22
mm and 136 mm) confirmed the generally observed trend, i.e. the insulators with larger diameter
performed much worse due to uneven salt fog distribution. All these data provide an input for the
proposal to re-consider the application of the method with direct spraying especially for the insulators
with larger diameters. Another issue discussed in the paper is the tracking and erosion test for DC,
which is not standardized at present.

KEYWORDS

Composite insulator, Tracking and erosion test, Service experience, Test method

[email protected]
BACKGROUND AND GOAL
At present more than 1.5 million station post and apparatus composite insulators with larger diameters
than for line composite insulators are in service in AC and DC substations. This estimation is based on
extrapolation of the data presented in [1]. Their application seems to be increased exponentially due to
the main driving force, i.e. explosion-free performance in comparison to the conventional porcelain
insulators. Composite insulators also have good service records from both a pollution point of view
and ageing (i.e. long-term performance) [2].

At the design stage all composite insulators, including station post and hollow core insulators with
large diameter, should be verified by tracking and erosion tests. At present, the main choice is between
1000 hours salt fog test (IEC 62217 [3]) or 5000 hours test at multiple stresses (IEC/TR 62730 [4]),
the latter is called multiple stresses test throughout the paper. There are a number of differences
between these two main types of tracking and erosion tests. The pure salt fog 1000 hours test is at
present well defined and takes into account differences in insulator diameter and creepage factor to
prevent flashovers during the test (i.e. intending to create more “ageing” than “pollution” test). These
specific improvements made for the salt fog tests were not introduced in the salt fog phase for 5000
hour multiple stress tests as per IEC/TR 62730. Furthermore, in the multiple stresses tests salt fog
must be directly sprayed on the test specimens, this is again different to the pure salt fog test requiring
indirect spray. While testing composite insulators with larger diameters a number of companies
experienced an indication that using the present test method according to the IEC/TR 62730, some
deterioration and damage as well as flashovers observed in the test would not appear in service. Thus,
the goals of this paper were as follows:
 Compile latest service experience of insulators with larger diameters
 Make critical analysis of the available experience with 5000 hour multiple stresses test as per
IEC/TR 62730 in different laboratories
 Perform standard and non-standard 5000 hour multiple stresses tests using a specially
designed insulator matrix with two principally different diameters of the insulators
 Make recommendations on possible DC tracking and erosion test
SERVICE EXPERIENCE OF INSULATORS WITH LARGER DIAMETERS
The most comprehensive summary of service experience of insulators with larger diameters (in this
case hollow core insulators for apparatus insulators) was provided in CIGRE Technical Brochure (TB)
455 in 2011 [2]. It was stated in this TB that “The experience from service and test stations is
generally positive for all silicone rubber apparatus and only minor degradation such as loss of
hydrophobicity or biological growth has been reported in some cases”. One of the explanations for
this positive behaviour was published as far back as 2004 and was based on long-term field
observations in high polluted areas in the UK and South Africa [3]. The main explanation for less
ageing for apparatus insulators was that for such designs the maximum E-field in the vicinity of HV
flanges is lower and it can be differently located in comparison with line insulators due to the field
grading by the internal parts. Also the leakage current density will be lower due to a large diameter.
Recently two detailed surveys on service experience of the insulators with larger diameters were
published; unfortunately both of them included data only for one manufacturer. The first one
concentrated on specific service/field experience in harsh industrial, coastal and desert environments
with the conclusion that the service performance was good [6]. The second one was based on pure
service inspections performed only hours after the voltage at specific substation were switched off.
This covered 58 insulators in total installed at 11 polluted substations in Argentina, Australia, China,
Oman, etc [7]. The insulators were in good condition and no deterioration was observed. Only a few
minor surface marks were found, e.g. handling-induced deterioration or some biological growth, such
surface marks will not influence the safe operation [7].

Thus, even taking a closer look as in [6], [7], the service experience of insulators with larger diameters
is generally positive world-wide. Screening tracking and erosion tests are an important tool to be used
to verify different designs for possible design/manufacturing weaknesses. In case of apparatus
insulators investigated in [3], most of the prototypes were tested full-scale in 1000 hours salt fog tests
2
using indirect salt spray and initial salinities depending on insulator diameter and profile. This gave
more relevant stress on the insulators both in E-field and tracking and erosion, avoiding high currents,
typical for long and unstable arcs which do not verify the design and material weakness [8], [9]. These
results (initial salinities depending on insulator diameter and profile) were used then as a basis and
later were accepted for the IEC standard 62217 (“Common clauses for composite insulators”).
However, the 5000 hours multiple stresses test (IEC/TR 62730) was not properly updated and at
present uses direct spray of salt fog and only one relatively high salinity level. This issue is needed to
be investigated further.
HISTORY OF DEVELOPMENT OF TRACKING AND EROSION TEST METHODS
A list of historically known tracking and erosion tests (at that time known also as ageing tests) can be
found in CIGRE technical brochure 142 [10] and includes information about the test procedures
developed in Canada, France, Italy, USA and Germany. The main conclusion was that further research
is needed to define the test that can be adapted to the majority of service environments [10].

The first IEC standard which included the tracking and erosion tests for composite insulators was
IEC 1109 issued in 1992. It included three possible tests: 1000 hours salt fog test, 5000 hours salt fog
test and 5000 hours “ageing test under operating voltage simulating weather conditions”. The level of
specification of these tests was low in general, e.g. there was no procedure for the calibration and no
indication on the type of nozzles for any tests. The type of spraying (indirect) and the test creepage
distance was specified only for the salt fog test. The majority of these issues applied for 5000 hours
test were discussed more in detail in a collaborative study [11]. It is important to note that both
1000/5000 hours tests were developed using line composite insulators, thus when applied to the
insulators with larger diameters further issues were expected.

Taking into account above, a new 2nd edition of IEC standard 62217 was issued in 2012 and was
intending to standardize further and harmonize tracking and erosion tests. Three tracking and erosion
tests were defined in the first edition of this standard, i.e. 1000 hours salt fog, 5000 hours multiple
stresses and tracking wheel tests (at present the two last tests are moved to the IEC/TR 62730 and only
the salt fog test remained in 62217). For all these three tests in IEC 62217 and IEC/TR 62730, the test
specific creepage distance was harmonized to 20 mm/kV and common acceptance criteria were
decided. However, there are still a number of differences between the more advanced level of
standardization for the 1000 salt fog test and the 5000 hours test; these are summarized in Figure 1.
These differences might be negligible when testing line composite insulators, but can influence the
results for the insulators with larger diameter.

Figure 1 Difference in test parameters for 1000 hours and 5000 hours tracking and erosion tests.

SUMMARY OF 5000 HOURS MULTI-STRESS TESTS

The authors of this paper has been involved in many 5000 hours tests using both the old version of
IEC (at that time IEC 1109) and the recent version of IEC/TR 62730. The main test cycle is shown in
Figure 2 (left). The recommended set-up according to IEC/TR 62730 is presented in Figure 2 (right).
The intention of this summary was to investigate if there are any trends between the results in the
5000 h test primarily based on the diameter of the test objects and the way to apply the salt fog

3
(indirect or direct spraying). The compilation of all 5000 hours tracking and erosion tests performed
for the last 10 years (about 50 tests in total) resulted in the summary presented in Figure 3 and Figure
4. To visualize this comparison some home-made classification of deterioration/damages was applied.
The test objects were graded on a scale from one to ten, where one indicates no deterioration and ten
indicates severe damage. The results are presented over the average diameter of the test object and
over the Unified Specific Creepage Distance (USCD) according to the IEC 60815. If the USCD was
unknown from the test reports, it was denoted as zero. A few observations from Figure 3 and Figure 4
are as follows:
 There are more deterioration/damages when using direct spraying instead of indirect spraying
even for the insulators with not so large average diameter (50-100 mm).
 There are more deterioration/damages when using direct spraying instead of indirect spraying
even for the insulators tested as required by standard USCD 34 mm/kV (for phase-to-ground
voltage), this corresponds to 20 mm/kV for phase-to-phase voltage. It seems that this
observation is valid even for less stress (52 mm/kV).
 These observations are not in line with positive service experience and can be explained by
uneven distribution of fog around the test objects due to direct spraying. More “scientifically-
based” test to compare direct and indirect spraying on the same insulation matrix is
recommended.
 In some cases severe damages on insulators were observed at the less stressed grounded
bottom part of the insulators indicating uneven distribution of fog around the test objects due
to direct spraying, see Figure 5.

Figure 2 Left: standard test cycle for 24 hours; Right: standard set-up (adopted from IEC/TR
62730).

Figure 3 Comparison of the deterioration after 5000 hours test performed with direct and indirect
spraying over average diameter of test object. Zero indicates no damage, ten indicates
severe damage.

4
Figure 4 Comparison of the deterioration after 5000 hours test with direct and indirect spraying
over Unified Specific Creepage Distance (USCD). If the USCD was unknown from test
reports, it was denoted as zero. Zero indicates no damage, ten indicates severe damage.

Figure 5 Examples of severe damages at the bottom (grounded) part of the tested insulators
indicating different axial and peripheral pollution stress by uneven distribution of fog due
to direct spraying. The condition of these insulators was graded as ten, according to the
home-made classification of deterioration/damages.

COMPARISON OF DIRECT SPRAYING TEST WITH INDIRECT SPRAYING TEST

Test objects and set-up

The test objects specially chosen for this test are presented in Figure 6. These were two pairs of
insulators manufactured from the same material and using the same technology, but with two
principally different trunk diameters, i.e. 136 mm and 22 mm. The creepage distance of the two
insulators was about 860 mm (the maximum difference was 5 mm). The general set-up for the test is
shown in Figure 6, while in Figure 7 it is shown in more detail to illustrate the different positions of
the test nozzles. For the direct spraying the nozzles are close to the insulators and each insulator is
sprayed by the individual nozzle. For the indirect spraying the nozzles are close to the floor level
providing an even flow of the salt fog, which is spread up to the insulators creating a uniform cloud of
wetting around them. The test voltage was 24,9 kV, thus insulators were stressed at 20 mm/kV
calculated for phase-to-phase voltage as prescribed by IEC/TR 62730.

5
Figure 6 Test objects used for the “academic” test with direct and indirect types of spraying and
general set-up (in the middle).

Figure 7 Left: set-up for standard test with direct spraying (salt-fog nozzles are in the middle
oriented towards test objects); Right: set-up for non-standard test with indirect spraying
(salt fog nozzles are at the bottom of chamber).

Test results for direct spraying

The test results were as follows:
 No flashovers during the test.
 Hydrophobicity at the end of the tests was at intermediate level HC 2-3-4 on all four test
objects.
 Deterioration on insulators with larger diameters was more severe than on insulators with
small diameters. On both insulators with larger diameters multiple erosion of depth about
1 mm was observed, while slightly less erosion was observed only in one place on one
insulator with smaller diameter.

6
  The deterioration on insulators with larger diameters started after 1000 hours of test.
 The maximum leakage current was up to 450 mA on the insulators with large diameters and
up to 250 mA on the insulators of smaller diameters (Figure 9). During the test the currents on
insulators in the salt fog phase were often higher than 100 mA independently on diameter,
which makes this test more “pollution” character than “ageing” [8].

Figure 8 Examples of erosion on two different insulators with larger diameter 136 mm,
photographs taken after 5000 hours multiple stresses test with direct spraying.

Figure 9 Development of leakage currents for the insulators with different diameters. Top: larger
diameter 136 mm; Bottom: smaller diameter 22 mm.

Test results for indirect spraying

Unfortunately, the identical test as described above with the same set of test objects, but using the
indirect spraying (see Figure 7, right) was recently started and was performed during only 1500 hours.
However, the indication is already clear. The test results were as follows:
 No flashovers.
 Hydrophobicity is still high (hydrophobic at HC 2, Figure 10).
 No deterioration on any of insulators (see Figure 10).
 The maximum leakage current was lower than for direct spraying (120 mA vs. 250 mA after
1500 hours), see Figure 11. However, leakage current was twice higher on the insulators with

7
 larger diameters than on the insulators with smaller diameters (120 mA vs. 60 mA). Thus, it
might be necessary to define salinity level based on insulator diameter as for 1000 hours salt
fog test.

Figure 10 Excellent condition of insulators after 1000 hours of test with indirect spraying.

Figure 11 Development of leakage currents for the insulators with different diameters after
1500 hours of the test with indirect spraying. Top: larger diameter; Bottom: smaller
diameter.

SUMMARY OF RESULTS FROM OTHER TESTS

Similar results for the standard test according to IEC/TR 62730 with direct spraying, as described
above, were obtained in tests performed in other test. The insulators tested had trunk diameters of
136 mm and 180 mm. Some severe damages were observed on the insulators after the test, see Figure

8
12. The leakage current during the test reached a few hundreds of mA after 1000 hours of test and 600
mA as the maximum, see figure 13, which makes this test more “pollution” character than “ageing”.
This behaviour is confirmed by sporadic flashovers, 7 flashovers were registered in total on two
insulators during the test.

Figure 12 Examples of severe damages on the tested insulators indicating uneven distribution of
salt fog due to direct spraying.

Figure 13 Development of leakage currents during the IEC/TR 62730 5000 hours multiple stresses
test for the insulator with trunk diameter 136 mm.

Similar results to those described above for the standard test according to IEC/TR 62730 with direct
spraying are also reported in [13]. The insulators tested had trunk diameters of 96 mm and 145 mm,
creepage distance was identical. Within the first 1000 hours of the test frequent flashovers were
observed and finally the test was stopped due to the uncertain duration of interruptions (they usually
took place during the night). This makes this test more “pollution” character than “ageing”. Some
erosion was also observed just in the area where the direct spraying hit the insulator surface, main
damages were caused by the flashovers (between the end fitting and housing), see Figure 134.

Figure 14 Examples of severe damage on the tested insulators indicating uneven distribution of salt
fog due to direct spraying in multiple stresses test.

9
COMMENTS ON POSSIBLE DC TRACKING AND EROSION TEST
At present there is no standardized tracking and erosion test for DC voltage. For many test parameters
it can be defined similar as to the standard tests available for AC, i.e. IEC 62217 and IEC/TR 62730.
The most important question, however, is the applied voltage. There is numerous data for tracking and
erosion tests showing that when a DC voltage stress corresponds to the r.m.s. value of AC voltage
stress the level of deterioration will be higher at DC than at AC for the same test insulators [14].
However, one should take into account that DC insulators are always dimensioned based on pollution,
not ageing performance. Further, DC flashover voltage is always lower than AC r.m.s. voltage at the
same level of pollution severity, especially when the hydrophobic effect is temporarily removed. This
means that in general the actual stress in service for DC insulators is estimated to be about 30% lower
than r.m.s. AC. The only reference supporting this issue is an old CIGRE report describing
comparative AC/DC testing of composite insulators [15]. This was performed at 100 kV AC r.m.s and
at 70 kV DC. The results for DC showed that “most of the insulators showed damages equal to or
smaller than those found for AC aged insulators, for the same duration of ageing” [15]. Thus, the
proposal would be to test DC insulators at service voltage stress. Another important point is that for
the testing of DC insulators it is important to implement the recovery phase (typical for multi-stress
test) in the test procedure [16], [17]. This was also decided for the Chinese tracking and erosion test,
where a modified multiple stresses test procedure (with high non-soluble contamination typical for
China and additional long recovery of hydrophobicity during the solar phase) was proposed [18]. The
application of direct spraying is again very questionable and especially for DC, however more
research is needed.
SUMMARY
Comparison of positive service experience of composite insulators with larger diameters (hollow core
and apparatus) with limited laboratory experience of multiple stresses tracking of erosion tests with
direct spraying (IEC/TR 62730) indicated that in some cases the severe deterioration in laboratory
does not correlate with positive service experience.

The first results of the “scientifically-based” comparative test of the present procedure based on
IEC/TR 62730 (direct spraying on the insulators) and the test method based on indirect spraying
revealed the following. Heavy erosion not observed in service can be observed in the test especially on
the insulators with larger diameters and can be explained by uneven distribution of fog due to direct
spraying. High level currents in this test are also not typical for the ageing type of test, converting it in
more pollution test mode. These results were confirmed by observations in other similar tests. It is
proposed to re-consider the application of the method with direct spaying especially for the insulators
with larger diameters in the corresponding Maintenance Team of IEC/TR 62730.

In order to develop a DC tracking and erosion test the following items should be considered:
1000 hours salt fog or 5000 hours multiple stresses tests with common features as indirect spraying in
salt fog phase and test voltage about 70% of respective r.m.s. AC voltage (or service level of test
voltage).
ACKNOWLEGEMENT
The authors acknowledge the key role of “BEST PATHS” (BEyond State-of-the-art Technologies for
rePowering Ac corridors and multi-Terminal HVDC Systems’’) in supporting the funding of part of
this research. It is co-funded by the European Commission under the Seventh Framework Programme
for Research, Technological Development and Demonstration under the grant agreement no. 612748.

BIBLIOGRAPHY

[1] A. Pigini: “HV Insulators Today: Market Size & Reliability”, INMR Magazine, Q.
1, N. 1, 2011, p. 26
[2] CIGRE Working Group A3.21: “Aspects for the Application of Composite
Insulators to High Voltage (≥72kV) Apparatus”, CIGRE TB 455, April 2011

10
[3] IEC Standard: “Polymeric HV insulators for indoor and outdoor use – General
definitions, test methods and acceptance criteria”, IEC 62217, Edition 2.0, 2012-09
[4] IEC Technical Report: “HV polymeric insulators for indoor and outdoor use
tracking and erosion testing by wheel test and 5 000h test” IEC/TR 62730, 2012
[5] I. Gutman, L. Stenström, D. Gustavsson, D. Windmar, W.L. Vosloo: “Optimized
use of HV composite apparatus insulators: field experience from coastal and inland
test stations”, CIGRE Session 2004, A3-104
[6] I. Gutman, D. Windmar, A. Holmberg: “Pollution and Ageing Performance of
Composite Apparatus Insulators with Different Profiles: Service/Field Experience
& Laboratory Testing”, World Congress & Exhibition on Insulators, Arresters &
Bushings, Seoul, Korea, 17-20 April 2011
[7] I. Gutman, C. Ahlholm, U. Åkesson, A. Holmberg, D. Wu, L. Jonsson: ”Long-term
service experience and inspection results of HV equipment made of silicone rubber
insulators”, CIGRE Symposium, Auckland, New Zealand, September 14-20, 2013,
A3-412
[8] I. Gutman, R. Hartings, R. Matsuoka, K. Kondo: “Experience with IEC 1109 1000
h Salt Fog Ageing Test for Composite Insulators”, IEEE Electrical Insulation
Magazine, May/June 1997, Vol.13, No. 3, p.p. 36-39
[9] I. Gutman; R. Hartings: “Single Stress and Multistress AC/DC Tracking and
Erosion Tests for Composite Insulators”, Proceedings of the Nordic Insulation
Symposium, Copenhagen, Denmark, 10-12 June, 1999, p.p. 271-278
[10] CIGRE Task Force 33.04.07: “Natural and artificial ageing and pollution testing of
polymeric insulators”, CIGRE TB 142, June 1999
[11] I. Gutman, R. Hartings, G. Riquel: “IEC 5000 Hours Ageing Test: Critical Review
and Further Test Development”, Proceedings of the International Symposium on
Electric Power Engineering, Stockholm Power Tech, 18-22 June, 1995, volume
“High-Voltage Technology”, p.p. 310-315
[12] I. Gutman, D. Windmar, A. Holmberg: “Pollution and Ageing Performance of
Composite Apparatus Insulators with Different Profiles: Service/Field Experience
& Laboratory Testing”, World Congress & Exhibition on Insulators, Arresters &
Bushings, Seoul, Korea, 17-20 April 2011
[13] F. Schmuck: “Alternative Methodologies for Tracking & Erosion Tests: IEC/IS
62217 vs. IEC/TR 62730”, INMR Magazine, Issue 107, Q. 1, p. 24
[14] I. Gutman, K. Halsan, J. Seifert, W. Vosloo: “Pollution and ageing performance of
DC line composite insulators: service experience vs. laboratory experience”, World
Congress & Exhibition on Insulators, Arresters & Bushings, Seoul, Korea, 17-20
April 2011
[15] S. Bossi, A. Pigini, R. Reali, G.P. Fini, A. Perin, Channakeshava, N. Yasudev, M.
Ramamoorty: ”Study of the performance of composite insulators in polluted
conditions”, CIGRE Session 1994, R. 33-104
[16] J.M. Seifert, D. Stefanini, H. Janssen: “HTV silicone composite insulators for
HVDC applications – long-term experiences with material and design for 500 kV
and above”, 16th ISH -2009, Cape Town, South Africa, 24-28 August, 2009, C-12
[17] A. J. Maxwell, I. Gutman, C. S. Engelbrecht, S. M. Berlijn, R. Hartings, W. L.
Vosloo, D. Loudon, R. Lilja, A. Eriksson: “Selection of composite insulators for
AC overhead lines: implications from in-service experience and test-station
results”, CIGRE Session 2002, R. 33-402
[18] Liang Xidong, Zhang Yibo, Li Zhenyu, Yin Yu: “A New 5000 h Multi-stress test
Procedure for Silicone Rubber Insulators Based on Contamination and
Hydrophobicity Change Simulation”, 17th ISH -2011, Hannover, Germany, 22-26
August, 2011, E-092

11