Resumo:
The safe and efficient utilization of any dielectric gas asks for the knowledge of its electric
behaviour and its physics process. Among all the known dielectric gases the most utilized in the
medium and high voltage systems is the air. It is due to the fact of the air is an abundant and
free resource, and it recovers after flashover. However, its electric behaviour changes with
weather conditions, and, in consequence, there is a variation on the values of breakdown
voltage and standard deviation. The analyses of the influence of the weather conditions are a
typical experimental task, where the results are obtained from different environmental
conditions.
The humidity correction factor to lightning impulse tests as stated in national and international
standards is still under consideration. Nowadays, there is an agreement related to the purpose
established by IEC 60. Therefore, considering small spaces that are common in high voltage
equipments applied to medium voltage systems, there is lack of information about the accuracy
presented by information proposed in the standards.
This dissertation aims to investigate the humidity influence in the lightning impulse tests,
observing the critical value and the standard deviation. Taking this in account, the goal is to
obtain tendency curves able to represent the voltage behaviour with the humidity to some basic
electrodes. Then, the first step is to analyse, using a statistic tool, some results obtained in tests
carried out in basic electrodes, i.e., rod-rod and sphere-sphere spacing, and, after this, define a
set of tendency curves to improve the lightning testing results correction.
The tests were carried out in the High Voltage Laboratory Impulse Generator HAEFELY 1,2kJ –
450kV at Federal University of Itajubá. A set of tests was repeatedly performed in order to verify
which of the parameters are necessary to define the features in air insulation, as well as the
environmental influences in the lightning tests results. To the voltage range used in the tests, it
was chosen the 3s-2p connection, which presents the best efficiency in the range of 30 [kV] to
180 [kV]. The test were carried out with the basic electrodes inside the climatic room, where it
can be established specific conditions of humidity, as well as the environmental conditions of
the laboratory can be utilized. The results were statistically analysed by software developed at
LAT-EFEI, where it was not considered any humidity correction factor.
In order to compare the present approach with standard one, it were carried out systematic tests
on the same. The utilized method, Up-and-Down Method validated by NBR6936/ April 1992 –
ABNT (Brazilian Association of Technical Standards) High Voltage Test Techniques.
The analysis of the tests results up to this moment, taking in account the standard deviation of
the corrected breakdown voltage, suggest that a better humidity correction factor can be
obtained to lightning impulse tests.