Too rapid agitation is undesirable, since it introduces an excessive amount of
air into the liquid. Immediately after agitating, the test cup should be fi lled
with oil to a height of not less than 20 mm (0.787 in.) above the top of the
electrodes. To prevent the escape of entrapped air, the container should be
gently rocked a few times and the oil allowed to stand in the cup for 3 min
before voltage is applied.
The temperature of the sample when tested should be the same as that of
the room, but not less than 20°C (68°F). Testing of oil at a lower than the room
temperature is likely to give variable results, which may be misleading.
Voltage should be applied and increased at a uniform rate of 3 kV/s from
zero until breakdown occurs, as indicated by a continuous discharge across
the gap. Occasional momentary discharges that do not result in a permanent
arc may occur; they should be disregarded.
Referee testing: When it is desired to determine the dielectric breakdown
voltage of a new liquid for referee purposes, one breakdown should be made
on each of fi ve successive fi llings of the test cup. If the fi ve values meet the
minimum dielectric values, the average should be reported as the dielectric
breakdown voltage of the sample. If they do not meet the minimum dielectric
values, one breakdown on each of fi ve additional cup fi llings should be
made and the average of the 10 breakdowns reported as the dielectric breakdown
voltage of the sample. No breakdown should be discarded.
Routine testing: When it is desired to determine the dielectric breakdown
voltage of a liquid on a routine basis, one breakdown may be made on each
of two fi llings of the test cup. If no value is below the specifi ed acceptance
value, the oil may be considered satisfactory, and no further tests are required.
If either of the values is less than the specifi ed value, a breakdown should be
made on each of three additional cup fi llings, and the test results analyzed.
Alternative method: When it is desired to determine the dielectric breakdown
voltage of a liquid on a routine basis, fi ve breakdowns may be made on one cup
fi lling with 1 min intervals between breakdowns. The average of the fi ve breakdowns
should be considered the dielectric breakdown voltage of the sample,
provided the breakdown values meet the criterion for statistical consistency.
If the breakdown voltages do not meet this criterion, the contents of the cup
should be discarded, the sample container again gently inverted and swirled,
the cup again fi lled, and fi ve breakdowns made on this second cup fi lling. The
average of the 10 breakdowns should be considered as the dielectric breakdown
voltage of the sample. No breakdown should be discarded.
Criterion for statistical consistency: Compute the range of the fi ve breakdowns
(maximum breakdown voltage minus minimum breakdown voltage), and
multiply this range by three. If the value obtained is greater than the next to
the lowest breakdown voltage, it is probable that the standard deviation of
the fi ve breakdowns is excessive, and therefore the probable error of their
average is also excessive.