40 BETTER

TESTING METHODS

Three Tips
for Reducing
Wasted Time
When Testing
Transformers Brandon Dupuis received a B.S. Electrical
Engineering from the University of Maine.
He joined OMICRON electronics Corp. in 2013,
by Brandon Dupuis where he presently holds the position of
Regional Application Specialist for trans-
former testing. Brandon’s focus is currently
on standard and advanced electrical diag-
nostics for power transformers and circuit
breakers. Presently, Brandon is a well-known
OMICRON instructor teaching electrical trans-
former diagnostic testing theory, applica-
tion, and test result analysis, which includes
both presentations and hands-on training.
Brandon is an active member of the IEEE
PES Transformers Committee.

The transformer test that

customers most struggle
with is the transformer
Power Factor test.
 Transformer Technology September 2019 41

Issue 3

Review the “Power Factor 3HUIRUPLQJ RͧLQH HOHFWULFDO WHVWV RQ WUDQVIRUPHUV FDQ EH WLPHFRQVXPLQJ
Checklist” Prior to Testing especially when the field measurements are not captured correctly the first
WLPH7KLVDUWLFOHLGHQWLͤHVWKHHOHFWULFDOWUDQVIRUPHUWHVWVWKDWWHVWHTXLSPHQW
Without a doubt, the transformer XVHUV RIWHQ VWUXJJOH WR SHUIRUP HͦFLHQWO\ DQG FRUUHFWO\ 7KUHH ͤHOGWHVWLQJ
test that customers most struggle tips will be provided, to address the transformer tests that are often “time-
with is the transformer Power Factor suckers”. By building awareness, test equipment users can better avoid
test (which includes the Overall WKH VLWXDWLRQV ZKHUH D VLJQLͤFDQW DPRXQW RI WLPH LV ORVW GXH WR UHWHVWLQJ
Test, the Bushing C1 Test, and the troubleshooting, and collaborating with test equipment manufacturers.
Bushing C2 Test). Unfortunately, the
Power Factor test is highly sensitive
to the test environment, to the test
lead connections, and to the test
̽ Are the surfaces of the bushings
dry (and reasonably clean)?
in resistor”, which has been
known to influence the Power
specimen’s earth-ground connection, Moisture on the surfaces of Factor measurements, when the
among other things. Since the VJGDWUJKPIUECPUKIPKƒECPVN[ transformer is tested in the neutral
Power Factor measurement is highly influence a Power Factor VCRRQUKVKQP5RGEKƒECNN[(GFGTCN
sensitive, obtaining the correct (i.e. the measurement. In most cases, 2KQPGGT'NGEVTKECPF(GFGTCN2CEKƒE
valid) Power Factor measurements in using a clean, dry rag to dry Electric transformers with LTCs
VJGƒGNFKUEJCNNGPIKPI the surfaces of the bushings have been known to exhibit this
KUUWHƒEKGPV+PECUGUYJGTG behavior.
The high sensitivity of the Power Factor excessive surface contamination
test is a double-edged sword: On one
hand, the high sensitivity makes the
on the bushing surfaces is
present, using Windex or Collinite
̽ Ensure that the exterior surface of
the test equipment’s high-voltage
Power Factor test a powerful tool for may be used to clean the surfaces cable is not touching any surface
identifying compromised insulation of the bushings, and thus, of the transformer, at the “far end”
(e.g. moisture ingress, contaminated may improve the Power Factor where the test terminal is being
oil, a “bad bushing”, etc.). On the other measurements. energized.
hand, the high sensitivity makes A conservative approach is to
the Power Factor test prone to
measurement error.
̽ Are the groups of bushing
terminals short-circuited together?
ensure that the last two feet of
the far end of the test equipment’s
All primary side (H) bushing high-voltage cable is not touching
To save wasted time and prevent terminals must be shorted the transformer tank, the surfaces
from retesting, troubleshooting, and together, and all secondary side of the bushings, etc.
collaborating with test equipment (X) bushing terminals must be
manufacturers, the following Power
Factor Checklist should always be
shorted together. Always use
NON-insulated conductor(s) to
̽ Be aware that the test
HQYLURQPHQWFDQVLJQLͤFDQWO\
reviewed, prior to executing any short-circuit the bushing terminals LQͥXHQFHD3RZHU)DFWRU
Power Factor measurement: together when performing a Power measurement.
Factor measurement – Do NOT - Do not Power Factor test in the
̽ Are the transformer tank and the
test-equipment solidly grounded
use insulated shorting leads. If
insulated conductor(s) are used,
rain.
- Avoid testing in high-humidity
to earth-ground potential? then the conductor’s insulation can situations (where excessive
Not connecting the test specimen easily become part of the Power moisture is present).
and the test equipment to a solid Factor insulation measurement. - Avoid Power Factor testing
earth-ground reference is the most Also, connect the shorting when the temperature of the
common mistake test equipment jumpers as tightly as possible transformer oil is below 0°C.
WUGTUOCMGKPVJGƒGNF from bushing-terminal to bushing- - Power Factor test after lunch, if
terminal (in other words, do not let possible (which is typically when
̽ Are the bushing terminals of
the transformer completely
the shorting jumpers sag and/or
touch any surface other than the
the least amount of moisture/
humidity is present).
disconnected and isolated from all terminal being energized).
cable, bus-bar, support insulators, The Power Factor Checklist is
surge arrestors, etc.?
When applying a test-voltage of
̽ Remove all in-service grounds
from any neutral bushing
intended to help test equipment users
ūIGVVJGOGCUWTGOGPVTKIJVVJGƒTUV
10 kV, a minimum clearance of 3 terminals. time”, which is the most practical
in. should be established (between For example, remove the in-service strategy to saving wasted time when
the terminal(s) that is energized ITQWPFEQPPGEVKQPHTQOVJG: testing transformers. In conjunction
and all other surfaces). Avoid bushing terminal, if applicable. with the Power Factor Checklist, the
using a rubber blanket, insulated Variable Frequency Power Factor
gloves, etc. to isolate the bushing ̽ Place the LTC in any off-neutral test can be utilized by test equipment
Photo: Omicron

terminals from external surfaces – tap-position. users, to quickly and easily FRQͧUP
the best insulator for Power Factor Some transformer Load-Tap- that the Power Factor measurements
testing is air! Changers (LTCs) utilize a “tie- obtained are indeed correct.
42 BETTER
TESTING METHODS

Identify Incorrect Power

Factor Measurements
Before Leaving the Job Site

One of the most common scenarios Table 1.

we encounter as a test equipment
manufacturer is when a customer Case Study - Lapp POC 138 kV Bushings (1998)
contacts us to review a set of power
transformer test results. Often, the Measured Power Factor Nameplate Power Factor
data set is provided after the test (10 kV - 60 Hz) (10 kV - 60 Hz)
equipment user has left the job site
H1 0.36% 0.29%
CPFTGVWTPGFVQVJGQHƒEG7RQP
reviewing the data, we (OMICRON) H2 0.24% 0.23%
frequently identify incorrect
Power Factor measurements, and H3 0.35% 0.23%
recommend that the customer retest,
to obtain the correct results.
Consider the bushing C1 Power Factor
Regardless of whether the measurements shown in Table 1,
transformer is still offline or has which were obtained from testing
been re-energized, once this scenario three Lapp POC 138 kV bushings. By
QEEWTUCUKIPKƒECPVCOQWPVQHC only analyzing the 10 kV-60 Hz Power
company’s time and resources have Factor values, it is not obvious that
been wasted. If the crew returns to the the Power Factor measurements are
ƒGNFVQTGVGUVVJGPVKOGCPFTGUQWTEGU incorrect; however, notice that the
are wasted testing a second time. Variable Frequency Power Factor
If the transformer has already been traces for the three bushings in Figure 1
re-energized, then the company are erratic and jagged. In general,
invested time and resources to obtain jagged frequency-sweep tra-
invalid test-results that cannot be ces are indicative of inva- “Getting the measurement
used to assess the condition of the lid Power Factor measure-
transformer. ments. In this case, the cus- ULJKWWKH̨UVWWLPH˽LVWKH
To quickly identify and correct “bad”
tomer determined that they
did not short-circuit the pri- most practical strategy to
Power Factor measurements before
they leave the job site with the incorrect
mary side (H) bushing ter-
minals of the transformer,
saving wasted time when
test results, test equipment users can when the C1 Power Factor
measurements were perfor-
testing transformers.
utilize the Variable Frequency Power
Factor measurement. With a Power med on the bushings.
Factor measurement at one test-
voltage and at one test-frequency
(in other words, with one Power Factor Figure 1. Variable Frequency Power
RGTEGPVCIGXCNWG KVKUFKHƒEWNVHQTVJG Factor results for three Lapp POC
test equipment user to verify that the 138 kV bushings
measurement is even valid; however,
invalid measurements often become 0.60%
obvious when the Variable Frequency H1 H2 H3
Power Factor measurement is 0.55%
performed and analyzed.
0.50%
Power Factor (%)

The Variable Frequency Power Factor 0.45%

test involves performing Power Factor
measurements at a series of different 0.40%
test frequencies (e.g. 15 Hz, 30 Hz,
0.35%
45 Hz, 60 Hz, 150 Hz, 200 Hz, 300 Hz,
and 400 Hz). The general guidelines 0.30%
used to assess the Variable Frequency
Power Factor test (along with several 0.25%
case studies) are provided in the paper
0.20%
Photo: Omicron

titled “The Value of Performing Power

0 100 200 300 400
Factor Sweep Measurements on
Bushings” [1]. Frequency (Hz)
Transformer Technology September 2019 43

Issue 3

To quickly identify and

FRUUHFW˼EDG˽3RZHU
Factor measurements
before they leave the job
Consider the Overall Power Factor Table 2.
site with the incorrect test
measurements shown in Table 2 and
Figure 2, which were performed on a
results, test equipment
Overall Power Factor
-WJNOCPM8QKNƒNNGFVTCPUHQTOGT Measurements users can utilize the
The 10 kV-60 Hz Power Factor
measurements look “normal” for an Kuhlman 118 kV transformer
Variable Frequency Power
QKNƒNNGFVTCPUHQTOGTDWVWRQP
reviewing the Variable Frequency
Measured Power Factor
(10 kV - 60 Hz)
Factor measurement.
Power Factor traces, the customer
CH 0.16%
noticed that the CH frequency-sweep
trace looked abnormal. Notice that as CHL 0.22%
the test frequency increases, the CH
trace approaches 0% Power Factor,
and even becomes negative. Negative
Power Factor measurements are Figure 2. Variable Frequency Power Factor
typically a tell-tale sign of an invalid results for a Kuhlman 118 kV transformer
measurement.
0.30%
If the test equipment user only
had the 10 kV-60 Hz Power Factor
measurements on-hand, then it is 0.20%
conceivable that they would have CH CHL
Power Factor (%)

overlooked the incorrect measurement.

In this case, the customer determined 0.10%
that the transformer tank and the test
equipment were not connected to a
solid earth-ground reference potential. 0.00%
0 100 200 300 400
With the Variable Frequency Power
Factor traces on-hand, test equipment -0.10%
users can quickly identify invalid
measurements, and can retest
before they leave the job site with the -0.20%
incorrect test results. Frequency (Hz)
44 BETTER
TESTING METHODS

Select the Appropriate

Test Current – DC Winding
Resistance

6JGƒGNFVTCPUHQTOGTVGUVVJCVVGUV
equipment users most struggle with
(other than the Power Factor Test) is
the DC Winding Resistance Test. The
DC Winding Resistance Test is the
offline measurement for identifying
bad connections and discontinuities
along the current carrying path of a
VTCPUHQTOGTYKPFKPI=?5RGEKƒECNN[
the DC Winding Resistance Test is
an invaluable tool for identifying bad
connections associated with tap-
changers, which includes both De-
Energized Tap-Changers (DETCs) and
Load-Tap-Changers (LTCs).

In theory, the DC Winding Resistance

Test is a simple concept that relies
on the fundamental application of
Ohm’s Law (V = I x R) [2]; however,
in practice, obtaining the correct
transformer resistance measurements
is challenging, because to obtain
the correct measurements, the
transformer’s core must be saturated.
The most common mistake that
test equipment users make when
performing the DC Winding Resistance
Test is not waiting long enough for
the core to fully saturate (in other
words, the resistance measurements
are captured “too soon”). Typically,
when the resistance measurements
are captured “too soon”, the measured
resistance values are higher than
expected, which makes it appear that
there is a bad connection, even when
no fault exists.

The best strategy for saturating the test current in the range of 5-10 A secondary side, a test equipment
transformer core quickly, and in turn,  KUUWHƒEKGPVŤOQUVTGUKUVCPEG user often has anywhere from
obtaining the correct resistance measurements performed on 57-99 resistance measurements
measurements quickly, is to inject as the primary side (H) winding of a to obtain (that are often well
high of a DC current as possible into transformer have resistance values DGNQYOµKPOCIPKVWFG +P
the winding under test. The higher the ITGCVGTVJCPOµ these cases, it is important that a
injected test current, the faster the
transformer core saturates, the faster
̽ Typically, when testing winding
TGUKUVCPEGUNGUUVJCPOµC
sizeable test-current (e.g. 20-30 A)
is available to the test equipment
the test is performed. The following test current in the range of user, so that they may perform the
guidelines are intended to assist test 20-30 A is ideal – most resistance lengthy test-sequence as quickly,
equipment users with selecting the measurements performed on the and accurately, as possible.
appropriate test current for a given
winding resistance measurement:
UGEQPFCT[UKFG : YKPFKPIQHC
transformer have resistance values
̽ The test-current should not exceed
10% of the rating of the winding
NGUUVJCPOµ under test.
̽ The lower the resistance of the
winding under test, the higher the
̽ Load-Tap-Changers (LTCs) in
North America are most often
̽ The test-current multiplied by
the resistance (of the winding
Photo: Omicron

test-current should be. CRRNKGFVQVJGUGEQPFCT[UKFG :  under test) should not exceed the
̽ Typically, when testing winding
TGUKUVCPEGUITGCVGTVJCPOµa
winding of a transformer. In cases
where the LTC is located on the
maximum compliance voltage
rating of the test instrument’s DC
Transformer Technology September 2019 45

Issue 3

The best strategy

for saturating the
transformer core
quickly, and in
turn, obtaining the
correct resistance
measurements
quickly, is to inject as
high of a DC current
as possible into the
winding under test.
current source. In general, the more for the primary side test, and one for References
power (VA) the test instrument’s the secondary side test) should be [1] B. Dupuis, “The Value of Performing
DC current source is rated for, the applied when performing both tests. Power Factor Sweep Measurements
higher the test current that can be The most common mistake made is on Bushings,” Transformer
injected into a given winding, the NOT injecting enough current when Technology, Issue 1, pp. 56-64, June
faster the DC Winding Resistance performing the DC Winding Resistance 2019.
Test can be performed. :VGUVYJKEJQHVGPNGCFUVQCNGPIVJKGT [2] C. L. Sweetser, “Obstacles
core-saturation time, and inaccurate Associated with Winding
To summarize, test equipment users measurements (especially when an Resistance Measurements of Power
often attempt to apply the same LTC is involved). Transformers,” Omicron.
magnitude of test current for both
the DC Winding Resistance H and DC Again, the best way to reduce
9KPFKPI4GUKUVCPEG:VGUVUJQYGXGT wasted time when testing
since the magnitude of winding transformers is to get the
TGUKUVCPEGKUQHVGPUKIPKƒECPVN[ PHDVXUHPHQWULJKWWKHͤUVWWLPH
different when comparing the DC
Winding Resistance H and DC Winding
4GUKUVCPEG:VGUVUKVKUEQPEGKXCDNG
that two different test currents (one