Beruflich Dokumente
Kultur Dokumente
in distribution insulators
G. Montoya, I. Ramirez and J.I. Montoya
IEE Proc.-Gener. Transm. Distrib., Vol. 151, No. 3, May 2004 335
of silicon rubber. The LC field measurements from 13PH 10,000
are lower than the measurements in the 13PC insulator. In 13PC
Veracruz City, there are two circuits with this type of 13RG
insulation, which have not been washed since for years. 13AC
resistance, Ω ( × 10 )
22PC
3
Porcelain insulators with isolated pins (polymeric con- 1,000
33PC
crete pins-PCP) can be considered in the group of hybrid
insulators. Diverse applications of PCP together with pin-
type insulators are used, such as metallic-oxide surge
arrester with polymeric concrete housing, creepage distance 100
increment and puncture distance increment (see Fig. 5). The
insulators with PCP tested in the laboratory had a smaller
LC than the 13PC insulators.
10
oc 8
ju 9
no 8
au 8
ja 8
99
99
98
99
m 9
m 9
se 98
de 8
ap 9
au 9
9
9
t9
l9
r9
9
l9
p
ay
c
n
n
n
g
b
g
ar
ju
ju
ju
fe
Fig. 6 SR minimum level
10,000
resistance, Ω ( × 10 )
3
1,000
13PC
100 13RG
13AC
22PC
33PC
10
oc 8
ju 9
99
ja 8
no 8
ap 9
99
99
98
de 98
se 8
au 8
m 9
m 9
au 9
9
9
9
t9
9
9
l9
r9
l9
p
ay
g
c
ar
n
n
n
v
g
b
ju
ju
ju
fe
336 IEE Proc.-Gener. Transm. Distrib., Vol. 151, No. 3, May 2004
ESDD measuring permits the detection of the seasonal and measures the leakage peak currents in an insulator
variation model existing in the test bank. According to the energised with high voltage.
IEC 60815 standard, ESDD classifies the pollution level in All LC events were associated with the presence of storms
the zone as a ‘very heavy’ contamination. By means of this and high humidity in the environment. Table 1 shows the
method, it is possible to detect the growth of the pollution activity registered during one year.
level (Fig. 9). This is an offline method, which has a The base current is the current that circulates
disadvantage; the measurements made after rain will be continuously over the insulator surface. The base
lower because the rain has washed the insulation. current reported is the average of 100 readings carried
0.50
13 PC
0.45 13 RG
22 PC
0.40 33 PC
13 AC
0.35
0.30
2
mg /cm
0.25
0.20
0.15
0.10
0.05
0
Oct 98 Nov 98 Feb 99 Feb 99 NORTH May 99 Aug 99 Aug 99 Aug 99
2 months 2 months 3 months 6 months feb 99 3 months 3 months 6 months 12 months
4.3 Non-soluble deposit density out during a sample period of two hours. It means
As a complement to ESDD measurements, the quantity of that during almost two hours the current average had a
non-solubles contained in the contaminated samples was similar value.
determinated. This quantity was normalised between the The LC method joins the characteristics of the previous
areas of the insulator (NSDD in mg/cm2). methods, since it takes measurements in real time and also it
is a direct method. In Mexico, the LC method has been
4.4 Leakage current used successfully for insulation diagnosis in transmission
LC measurements are carried out using the leakage current lines [3, 9].
peaks measuring system ‘SIPICO’ [3]. This system detects
Scale:
Note: 22PC insulator was removed and a 13PH insulator was installed in February 1999
IEE Proc.-Gener. Transm. Distrib., Vol. 151, No. 3, May 2004 337
5 Discussion of field results 6.1 Test method
Several problems were found when moistening some
During the time period of insulation exposure (from August insulator profiles, at is the case of 13AC, In the field, the
1998 to August 1999), the months with the higher insulator is humidified by condensation when the wind
contamination were January to April, 1999. During these impacts the rain drops over the insulator. It was decided to
months, all the insulators reached the minimum SR value of start under the premise that the insulator is completely wet
the scale (Fig. 6). The insulator with the lowest SR is the and for that reason, the Japanese method was followed [10].
13RG, while the insulator with highest SR was the 33PC. The purpose of this test is to find the levels of LC; the
The ‘seasonal variation’ accumulation model of the voltage was only increased until it reached the nominal
pollution level predominates over its ‘growth in time’ voltage of the insulator and not until a flashover was
accumulation model. Owing to this, by January the reached as indicated by the method. The activity of LC was
maximum contamination level registered was reached. This extinguished quickly on drying -of the insulator. For this,
result corresponds to a 13RG insulator with a sample clean fog was generated to saturate the chamber before
period of one month (1.008 of ESDD, Fig. 8). introducing the insulator. As a consequence of this
The 13PC, 22PC and 33PC insulators have high LC modification, the LC activity remains almost a minute,
(Table 1). Even when they are energised to 13 kV and are leading in some cases to insulator flashover.
far away from a flashover by contamination, LC could
cause burns on wood, crossover and poles. The 33PC 6.2 Contaminant reproduction and test
insulator flashed-over during the tests carried out in a clean development
fog chamber to nominal voltage (34.5 kV). This means that Since the 13PC insulator registered high LC activity in the
the 33PC insulator does not support higher values of LC. field, this profile was used primarily to determine the
The 13RG insulator also had a high LC level. Likewise, it mixtures of contamination to use. The first tests were
had a higher pollution accumulation reflected by a very high developed with mixtures prepared according to the IEC
level of ESDD. However, due to its easy natural washing, in 60507[11] with high ESDD levels (over 0.6 mg/cm2).
a single year it returned to minimum levels of contamina- However, it was not possible to reproduce the LC levels
tion. In fact, this insulator had the maximum registered level registered in the field. It was decided to analyse the quantity
of ESDD, as well as the minimum level of accumulation of non-solubles in the samples from the field in order to
after the annual rainy season (see Fig. 9). It is recommended determine its relationship NSDD/ESDD.
for zones with periods of high, prolonged rainfall. Based on the relationship NSDD/ESDD, the mixture
The 13AC and 13PH insulators had the best behaviour generated was 600/200. It is composed of 600 grams of
with a low LC. If a PCP is placed to the 13AC insulator its Kaolin and 200 grams of salt dissolved in one litre of water
performance improves. The arrangement of the 13A (a normalised mixture according to IEC 60507 contains
insulator and PCP with shed (13AAF) in the laboratory only 40 grams of Kaolin). This mixture complies with the
had an LC smaller than 13PH and 13AC. relationship of NSDD/ESDD with an ESDD average of
During the year of observation, there were no flashovers 0.70 mg/cm2 and a NSDD of 4.4 mg/cm2. The new tests
on installed insulators. However, they reached very high LC produced expected LC levels; the obtained LC peaks were
and ESDD values. The losses in high LC should be used as higher than 1.5 A (Table 2).
a criteria for the selection of the insulation or for The initial current level of 800 mA was considered as the
maintenance purpose decisions. maximum permitted before generating a contamination
alarm. This value will be adjusted according to field results
6 Tests in clean fog chamber and new laboratory tests.
New tests were developed to obtain two extra points for
The type of pollution and pollution level found in the field the ESDD vs. LC curve. The first mixture (600/200) was
were reproduced in the laboratory in order to verify the diluted 50% each time, but respecting the proportion of
relationship between ESDD and LC. The highest ESDD NSDD/ESDD. New mixtures were made, 300/100 and
level was selected, which was in January, with a sample 150/50. The last mixture generated an LC under 500 mA.
period of one month. All tested insulators, maintained the same relationship
Scale:
338 IEE Proc.-Gener. Transm. Distrib., Vol. 151, No. 3, May 2004
1800 loses its hydrophobicity. Likewise, the ESDD against LC
1600 curve of the 13PH insulator does not have the same
behaviour as the remaining insulators (see Fig. 11). While
1400
the other insulators maintain the same tendency of growth
leakage current, mA
leakage current, mA
6.3 Tests to NSDD constant 1000
Based on previous results, it was decided to set the NSDD
level obtained in the third mixture (150/50) while the ESDD 800
level was also increased. More tests were generated with the 600
mixtures 150/100 and 150/200. With these points a graph of
ESDD vs. LC was obtained (see Fig. 11). A linear 400
1000
13 PH between them.
33 PC
800 This confirms that the leakage current method is the most
appropriate for insulation diagnosis in distribution net-
600 works.
The Japanese method was modified for the evaluation of
400
several insulator types in a fog chamber. The tests were
200
carried out with a higher level of non-soluble material than
those indicated in the IEC 60507 standard, since the
0 contaminant type found in the test bank was reproduced.
0.1 1
2
ESDD, mg/cm
340 IEE Proc.-Gener. Transm. Distrib., Vol. 151, No. 3, May 2004