C57.152 draft Section 6.1.

7
Page 1 of 5



6.1.7 Induced voltage test


6.1.7.1 General

This test shall be applied for power transformers as
installation tests when assembled on-site
on-site test after repair
test for condition assessment
test for failure identification

The test shall be performed as a withstand test with parallel PD measurement (PD
monitored withstand test). Both, a withstand test without PD measurement or only a PD
test with voltages near the rated voltage of the transformer under test, can only be
recommended when the PD monitored withstand is not applicable.

The three and / or single phase induced test voltages should be within a frequency
range of 80 Hz to 240 Hz. This voltage is generated by a suitable generator and adapted
to the test object by a step-up transformer.

NOTE: The generator is often also used for the loss-measurements at rated frequencies. This increases the
required frequency range from 50 Hz to 240 Hz.

The partial discharges shall be measured according to IEEE Std. C57.113 (2010).
Achieving a sufficiently low PD background noise level is difficult on-site. Therefore the
PD measuring circuit shall be as compact as possible, also grounding loops shall be
avoided and all well known measures for PD noise reduction shall be applied.

The mentioned generator for the induced voltage test may be a motor-generator (M/G)
set or nowadays also a static frequency converter (SFC) set.

Transportable M/G sets for three-phase voltage are usually designed for one or two fixed
test frequencies. In case of a single-phase test the test power delivered by the M/G set
is reduced by a factor 1/3 and the wave shape is more distorted (THD increases) due
to a negative sequence component of the test voltage.

Transportable three-phase SFC sets supply test voltages of continuously variable
frequencies between 80 Hz and 240 Hz, with respect to other tests (loss measurements)
even between 40 Hz and 240 Hz. In single-phase operation the test power is also
reduced by the factor 1/3, but the wave shape remains sinusoidal.

Suitable measures for the reduction of PD noise signals shall be applied for both types
of generators. Then in both cases more or less identical PD background noise levels can
be reached on-site.


C57.152 draft Section 6.1.7
Page 2 of 5

6.1.7.2 Preliminary test procedures
Low-voltage tests (insulation resistance, power factor, ratio, oil dielectric, etc.) should be
performed to determine if the insulation of the transformer under test is suitable for
energization. The insulating fluid should be sampled following prescribed procedures
(refer to ASTM D 923-9 1) and its total dissolved gas-in-oil level analyzed to ensure that it
is acceptable. A moisture content test of the oil should be made to ensure that excessive
amounts do not exist. A turns ratio test should be performed to confirm that the
transformer's tap changer for de-energized operation is properly positioned and that
shorted winding turns do not exist. For the interpretation see table 1.
Table 1Recommended diagnostic characteristics
Procedure New transformer
Service-aged
transformer
Power factor < 0.5 % < 2.0 %
Total dissolved gas
a
< 0.5 % < 0.8 %
Moisture content < 10 ppm < 15 ppm
Turns ratio
Within 0.5 % of
nameplate
Within 0.5 % of
nameplate

a) If units are equipped with nitrogen blankets, total dissolved gas should not exceed 1.0 %.
6.1.7.3 Special precautions before test
Corona discharges at the tank of the transformer under test or at nearby grounded or
energized objects may not only increase the PD background noise level, but also
influence the results of withstand tests. Therefore in preparation for the test, all high-
voltage bushings should be fitted with corona rings of sufficient size, so as to eliminate
all possibilities of air corona discharges. To prevent corona on the ground side, all sharp
edges and points on top of and close to the transformer tank should be masked by
covering them with corona rings galvanically connected to the tank. All high-voltage
bushings should be carefully cleaned and dried. Immediately prior to the test, they should
again be wiped dry. No conductive or semiconductive objects should be left ungrounded
on the transformer or close to it, as this would produce discharges from floating objects.
Therefore, they should either be taken away, when this is possible, or carefully grounded.
All current-carrying connections should be very carefully made to ensure good electrical
contact as contact arcing may produce unacceptably high PD background noise levels.
Any transformer-mounted surge arresters should be disconnected before energizing the
transformer in order to avoid arrester damage and limitation of the test voltages due to
arrester operation.
C57.152 draft Section 6.1.7
Page 3 of 5

6.1.7.4 Power factor pre-test for excitation by M/G sets
After connecting the test set to the transformer under test, a pre-test to determine the
power factor at the source should first be performed to ensure that the amount of
inductive compensation is sufficient to guarantee that the load on the generator is not
capacitive as this could lead to dangerous overvoltages due to generator self excitation. To
perform this pre-test, a high-voltage bushing of the test transformer should first be
temporarily fitted with an external spark gap adjusted to operate at about 50% of the
transformer nominal voltage. During this pre-test, the voltage should not be raised above
30% or preferably only at a level high enough to allow fairly accurate power factor
measurement. Note that the generator is capable of driving a slightly capacitive load
provided that the power margin is sufficient; it is when this margin is exceeded that
generator runaway occurs. To be on the safe side, higher than required inductive
compensation should first be used and it should be adjusted to a value that will allow the
test to be done at full test voltage without exceeding the generator limits as this would
cause generator protection tripping. The user should therefore be absolutely sure that
the generator has a sufficient power margin to reach the maximum test level without
danger of tripping. It is also important that the ratio of the step-up transformer used to
match the voltage output of the generator to that of the transformer under test be as close
as possible to the required value (optimum adaption). This will ensure maximum power
transfer from the generator to the transformer under test. After the pre-test has been
performed and the reactive compensation has been adjusted properly, the temporary
spark gap should be removed and the voltage can then be taken to the test level.

6.1.7.5 Pre-tests for excitation by Static Frequency Converters (SFC) sets

As there is no danger of self-excitation when using static frequency converters, a special
pre-test for checking the power factor is not necessary. In case of small test objects, the
test sequence can simply be started after assembly of the temporary test field on-site.
But in case of huge test objects the ratio between the available test power and the
needed test power should be checked in advance. It has to be taken into account that
the transformer under test is a capacitive test object at high frequencies and an inductive
one at lower frequencies. At the self-compensation frequency in between the power
demand is at its minimum. The SFC set shall be adjusted to that frequency.


6.1.7.6 Test procedure

The induced voltage test shall be performed as a HV withstand test in combination with
a PD measurement (according to IEEE Std. C57.113). Such a withstand test shall be
called a PD monitored withstand test. To keep the PD background noise level low
enough, the test should be performed on a clear day. All outside interferences such as
operating cranes or motorized vehicles should be kept clear of the test side and all
precautions before the test (see 6.1.7.3) shall be carefully considered.
C57.152 draft Section 6.1.7
Page 4 of 5

The sequence of this test is characterized by steps of four voltage levels including the
agreed voltage withstand level U
W
and the agreed PD measuring voltage level U
PD
(Fig.
1). The rate of voltage raise or of voltage reduction between the voltage levels should be
about 2% of U
W
per second. The test sequence is as follows:

























Fig.1: Sequence of a PD monitored withstand voltage test (explanations in the text)


1) The voltage is switched on or raised to a voltage U
1
= 0.5 U
0
= 0.5 U
r
/ 3 (with U
r

rated voltage, U
0
- line-to-ground voltage) at which the PD background level is
measured and recorded. The voltage is held there for a time t
1
3 min which
shall allow the determination of a stable PD measuring value.
2) The voltage is raised to U
2
U
0
and held there for long enough to measure a
stable PD level which is recorded (t
2
3 min).
3) The voltage is raised to U
3
= U
PD
and held there for a minimum duration of
t
3
5 min until a stable PD level is reached and recorded.
4) The voltage is raised to U
4
= U
W
and held there for the withstand voltage test of a
duration
t
4
/ s = 120 (rated frequency/test frequency),
but not less than t
4
= 15 s. The PD level is recorded for information.
5) The voltage is reduced to the agreed PD measuring level U
5
= U
PD
= U
3
with an
agreed duration of t
5
= t
PD
. The PD level should be continuously recorded but in
minimum every 2 min.
6) The voltage shall be reduced to U
6
= U
2
U
0
and held there for long enough to
measure and record a stable PD level (t
6
t
2
3 min).
7) At a voltage level U
7
= U
1
0.5 U
0
the PD background level shall be determined
and recorded for a duration t
7
t
1
3 min. After that the voltage is switched off.
C57.152 draft Section 6.1.7
Page 5 of 5


NOTE: It is recommended that the PD level should be continuously observed on at least one
measuring channel during the whole test sequence. Inception and extinction of significant PD
activity should be noted.

If no PD measurement can be performed (e.g. if it is not agreed or if the PD background
level is too high), the induced voltage test should be performed as a withstand voltage
test at the level U
4
= U
W
of a duration
t
4
/s = 120(rated frequency/test frequency),
but not less than t
4
= 15 s. The test remains a step test, and the levels U
1
to U
3
and U
5
to
U
7
shall be considered as conditioning levels of a duration t 1 min each.

6.1.7.7 Test parameters and test acceptance criteria

The withstand test voltage level U
4
= U
W
(and sometimes its duration t
W
) as well as the
PD measuring voltage level U
5
= U
PD
and its duration t
5
= t
PD
are usually subject to
negotiation. The level may vary from initial factory test voltage level for new transformers
to lower levels depending on the age and history of the transformer
(U
W
1.2 U
0
cannot be recommended.). If no PD measurement is performed, a test
duration t
W
t
4
may be considered. The PD measuring voltage level should be agreed
between 80% and 90% of the withstand voltage level (U
PD
= 0.8.0.9 U
W
) and its
duration t
5
= t
PD
= 20.60 min. The PD limit value shall also bee agreed, a value of the
apparent charge between 500 and 1000 pC may be appropriate.

The transformer has passed the PD monitored withstand test, if all the following criteria
are fulfilled:

1) No breakdown of the test voltage occurs during the whole cycle.
2) None of the PD levels recorded at the test voltage U
PD
exceeds the agreed PD
limit.
3) The PD level measured during the PD test does not exhibit any steady raising
trend or sudden sustained increase by more than 100 pC.
4) The PD levels at identical voltage levels before and after the withstand test (Fig.
1): U2 and U6: U3 and U5) shall not increase by more than 100 pC.

NOTE: If no PD measurement is performed, only the first criterion is applied.