Full resolution (JPEG)
- On this page / på denna sida
- 1959, H. 1
- Accuracy in Impulse Voltage Tests, by Niels Knudsen
<< 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 been proofread at least once.
(diff)
(history)
Denna sida har korrekturlästs minst en gång.
(skillnad)
(historik)
Accuracy in Impulse Voltage Tests
Professor Niels Knudsen,
Inst. för elektrisk anläggningsteknik, CTH, Göteborg
621.317 : 621.316.933.1
Den noggrannhet, med vilken man genom prov kan
fastställa överslagsspänningen för en isolator i luft,
beror inte i första hand på mätutrustningen.
Mätresultatet påverkas av olika variabla faktorer såsom
omgivningen, luftens sammansättning och tillstånd
samt provkretsens data och utformning. Även om
man i viss utsträckning kan taga hänsyn härtill
genom provningsbestämmelser och korrektioner, är
osäkerhetsmomentet fortfarande ganska stort. Detta
framgår av en serie jämförande mätningar på
kulgap och stånggap, som utfördes vid 14 laboratorier i
Europa under åren 1954—1955. Under det att de
egentliga mätfelen härrörande från mätutrustningen
var av storleksordning 1,5 % var avvikelserna ifråga
om överslagsspänning i stånggap av olika typer av
storlek ±6 ±8, ±10 resp. ±12 %.
Kulgapets roll som mätinstrument i
högspännings-provrum och -laboratorier har ofta varit föremål för
diskussion. Bl.a. if rågasättes huruvida mätkulgap bör
belysas med kvartslampa eller ej. Vissa
experimentella undersökningar utförda vid Chalmers tekniska
högskola tyder på att bestrålning med kvartslampa
inte har någon inverkan på överslagsspänningen vid
negativ stöt, obetydlig inverkan på
överslagsspänningen vid positiv stöt och ingen inverkan på
spridningen. Detta gäller dock endast stora kulgap och
höga spänningar.
According to a fundamental rule, a given quantity
cannot be measured with a higher degree of accuracy
than that with which the said quantity is defined.
It follows from this that in order to determine a
certain property in a test object by measurement,
two forms of accuracy must be taken into account.
On the one hand, there is the actual accuracy of
measurement which is determined by the measuring
equipment, and on the other, an accuracy which is
associated with the nature of the property and that
of the test object.
Analogously, a quantity should not be indicated in
figures with a higher degree of accuracy than that
by which it is defined. For the same reasons that
the width of a river is not stated in cm, the impulse
strength of an insulator should not be given to four
places of decimals. An overestimation of the
accuracy with which impulse strength can be determined
may be misleading, inasmuch as it may cause
reliance to be placed on a margin of safety which is
too low with respect to the insulation level and
over-voltage protection.
In carrying out a test to determine the flashover
voltage with an impulse of 1/50 for an insulator in
air for example, numerous sources of error of
different origins must be taken into account.
1. Haphazard divergence which is due to the fact
that different flashovers follow different paths
although they occur under conditions which are
apparently identical.
2. Influence of variable external factors, primarily,
the quality of the ambient air, density, humidity and
degree of ionisation.
3. Influence of the testing circuit’s data and
arrangement, thus, not the shape of the voltage curve
alone. Influence of the surroundings, on the
electrostatic field around the insulator.
4. Actual measuring errors due to the measuring
equipment.
These sources of error can be more or less
controlled. Corrections can be made for deviations in the
density of the air, and the same applies to some
extent to the humidity. By applying a sufficiently
large number of impulses, the 50 % flashover
value and the standard deviation can be fixed fairly
satisfactorily; the extent to which this fixed value
can be regarded as actually correct will then depend
upon the other factors.
For the purpose of determining the degree of
accuracy that may be expected from measurements
undertaken with great care, a series of
measurements were carried out during the years 1954—55
in 14 European high tension laboratories, of the
flashover voltage on three simple and well-defined
test objects, a sphere-gap with 25 cm balls, a
rod-rod gap and a rod-plate gap with 10 mm circular
bars. The results of the various laboratories’
measurements, which were published in Cigré Report
No. 326/1956 edited by K. Berger, have aroused
considerable interest owing to the fact that the
divergence was greater with the rod gap than had been
anticipated in many quarters.
The two representative cases, namely, a flashover
in a sphere-gap on the one hand, and in a rod gap
on the other, are described separately below, both
with reference to the abovementioned report and
also in the light of two investigations that were
carried out in the high tension laboratory at
Chalmers University of Technology, (CTH).
Sphere-Gap
The significance of the sphere-gap as a measuring
instrument in high voltage tests has often been the
subject of criticism. In the course of the
above-mentioned comparative measurements undertaken in
different laboratories, however, no facts were disclosed
ELTEKNIK 1959 1 1
<< prev. page << föreg. sida << >> nästa sida >> next page >>
Project Runeberg, Fri Oct 18 23:40:44 2024
(aronsson)
(diff)
(history)
(download)
<< Previous
Next >>
https://runeberg.org/elteknik/1959/0005.html
