Full resolution (JPEG) - On this page / på denna sida - Review of Published Data on the Humidity Dependence of the Flashover Voltage of Air Insulation, by Jarle Sletbak
<< prev. page << föreg. sida << >> nästa sida >> next page >>
Below is the raw OCR text
from the above scanned image.
Do you see an error? Proofread the page now!
Här nedan syns maskintolkade texten från faksimilbilden ovan.
Ser du något fel? Korrekturläs sidan nu!
This page has never been proofread. / Denna sida har aldrig korrekturlästs.
At high humidity a continuous water film is formed
on the surface, and the flashover voltage
characteristics shift to section C.
The distance between the maximum and minimum
lines depends upon to what degree the flashover
path follows the surface. With suspension
insulators it is significant whether the cap or the pin is
earthed.
The Japanese tests were made with voltages below
200 kV, and with tests on strings of more than two
units, model insulators in scale 1: 2 were used.
Gerber8 has made some measurements to check the
Japanese results, and he found curves of similar
shape, but the number of observation points is
rather low.
Generally the observed points are within ± 5 %
of the average curve, but ± 10 % deviation has also
been observed on certain types of insulators. In
addition to this, the results obtained in different
laboratories deviate somewhat. The total spread
band found when results from different laboratories
are plotted together is expanded to about ± 8 %\
Below are given some characteristic features of
different types of insulators.
When the number of suspension insulator units is
* increased, the curvature of the flashover voltage as
a function of humidity is more pronounced, and
starts at a lower humidity.
The humidity correction of a single unit seems
to vary from type to type, but on the other hand
the EEI-NEMA specifications1 give no particular
correction for a single unit.
With pin type insulators the flashover voltage
seems to be more erratic than with most other
insulators9, and the humidity correction varies
considerably from type to type10. The number of
observations available are inadequate to draw any
definitive conclusions on the reason for this. A few
measurements4,10 indicate that the correction is
reduced somewhat when the flashover voltage exceeds
100 kV, but this reduction may also be due to minor
variations in construction.
With bushings the humidity effect depends closely
upon the construction so that a correction factor of
general validity can hardly be introduced. The
EEI-NEMA specifications give only tentative data for
bushings.
Impulse flashover
Fielder511,12 found that the humidity correction
decreases at short time delays. The amount of
reduction depends on the test object, and Fielder
gives no average value for this, while according to
the EEI-NEMA specifications the percentage
humidity correction has to be reduced linearly with the
time delay from 10 to 0 [.is.
In view of this, it is obvious that the humidity
correction is smaller with impulses of short
duration, such as 1/5-impulses, than for standard
1,5/40 or 1/50-impulses, except for short time delays.
Fielder also found that the correction for relative
air density (r.a.d.) usually employed, is not applic
able for impulses when the flashover delay time
is less than about 10 lis.
As a reasonable approximation, the percentage
r.a.d. correction may be considered constant for
Fig. 2. Humidity correction factors for rod gaps with
positive impulses according to Striegel (curve A)
and Fielder (curve B).
time delays greater than 7 us and reduced
proportionally for time delays below this value. None of the
published observation data contradicts Fielders
results, but the EEI-NEMA specifications do not
include such a reduction.
Most of the published tests are made with positive
impulses, these being the most important ones
because the flashover voltage is usually lower in this
case. The correction is smaller with positive
impulses than a.c. voltages.
It is usually found that the positive impulse
flash-over voltage increases gradually with the absolute
humidity. However, Strigel13 has made tests on a
rod gap and found a correction curve as shown in
fig. 2., curve A. For comparison with Fielder’s
corrections, see curve B, fig. 2.
Belatively few tests are made with negative
impulses. Most of them have given a flashover curve
similar to that with positive impulses, except that the
correction is somewhat smaller.
However, Ishiguro14 found that with negative
impulses, the flashover voltages of rod-plane gaps and
pin type insulators would fall with increasing
humidity. For rod-rod gaps and small suspension
insulators the curves are more complicated, but the
general trend is that the flashover voltage is
independent of humidity. Tests with d.c. on a rod-plane
gap indicate the same general trend as for impulses.
Besides, Ishiguro found a similar dependence of
temperature with positive impulses on insulators as
Nishi and Nakajima found for a.c. voltages, fig. 1.
For a 20 kV transformer bushing gapped and
un-gapped, Ishiguro15 found increasing flashover
voltage with increasing humidity for both positive and
negative impulses.
The correction factor for condenser bushings
depends closely upon construction. The AIEE
standard No 21 (1942) gives the values from the
EEI-NEMA specifications as tentative data, stating that
they are not generally valid.
Discussion of published data and conclusion
As may be seen from the bibliography a great deal
of the publications refer to tests made in
laboratories in U.S.A., which have also participated in the
test program of the EEI-NEMA Subcommitte on In-
ELTEKNIK 1959 1 155
<< prev. page << föreg. sida << >> nästa sida >> next page >>
