Best Practice For Sweep Frequency Response Analysis

You might also like

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 44

Best practice for sweep frequency response

analysis (SFRA) - Part 1


02 August 2018

Robert Foster - Application Engineer, Megger USA


Sanket Bolar - Application Engineer, Megger USA

Introduction

IEEE and IEC standards reference numerous tests that can be performed to evaluate the
condition of a transformer. One of these is Sweep Frequency Response Analysis (SFRA)
where low voltage, multi-frequency sweeps are performed on the transformer. The results
are analyzed to reveal issues with the transformer’s internal components including the core,
windings, tap leads and connections. SFRA tests can detect small physical changes in these
components, but results can be affected by the way the test set is connected, specific
transformer settings and even tests previously performed on the transformer. This two-part
article explains how proper connections and methodical testing will help to make SFRA test
results more accurate and more repeatable.

Basics of SFRA testing

During commissioning of a three-phase two-winding transformer, 15 tests should be


performed:

 six open-circuit tests (one for each winding);

 three short-circuit tests (performed on the three high side windings with X1, X2, X3
shorted on the low side);

 three capacitive inter-winding tests;

 three inductive inter-winding tests.

All tests are conducted by injecting a low voltage, generally 10 Vp-p, on one terminal of a
winding and measuring the response on the other end of the winding or, for inter-winding
tests, on the corresponding terminal of the secondary winding. The terminals that aren’t
being measured are left open, shorted, or grounded depending on the test. The frequency of
the test voltage typically varies from 20 Hz up to 2 MHz and the ratio of the output to the
input, expressed in dB, is plotted against frequency. Phase versus frequency is also plotted.

Although each transformer has a unique fingerprint, a general pattern emerges based on
transformer type and test performed. If the transformer is faulty, analysis of different
frequency ranges indicates where the fault may lie. With further analysis and possibly
further testing it might be possible to identify what type of fault has occurred and its
probable location within the transformer.
There are three frequency ranges of interest for most transformers: the low, mid, and high
frequencies. The actual frequencies corresponding to each range depend on the size, type
and design of the transformer. However, from the typical sweep for an autotransformer,
which is shown in Figure 1 (taken from IEC 60076-18), it can be seen that effects related to
the core of the transformer dominate the low frequency range up to roughly 2 kHz. The
sweep begins with a decreasing magnitude based on the magnetizing inductances of the
core, extending down to a minimum that occurs at a resonance point between the bulk
capacitances of the transformer and the magnetizing inductance of the core. It should be
noted that this is a typical response where the A phase and C phase have two local
minimums and nearly overlay each other whereas the B phase has a single local minimum
resonance point and responds differently throughout the entire range. This is because the B
phase core leg has symmetrical return paths for the flux unlike A and C phases.

Moving into the mid-frequency range (2 kHz to 20 kHz), the response is most influenced
by the coupling between the windings, so the shape of the curve and resonant points will
vary depending on the type of connection and arrangement of the windings. As the sweep
progresses into the high frequency range of 20 kHz to 1 MHz the leakage inductances,
along with the series and ground capacitances of the winding determine the overall shape.
For both mid and high frequency ranges, the curves for all three phases almost exactly
overlay since the response depends on the winding and, in general, all three windings will
be nearly identical.

Once the sweep extends above 1 MHz for transformers greater than 72.5 kV, or above 2
MHz for transformers 72.5 kV and below, the response depends more on the test setup and
connections than on the transformer itself, although the internal tap leads will have some
influence. At this point the response for the phases will start to diverge. These comments
describe typical results; the actual frequency ranges and the effects of the various
components on the sweeps will vary from transformer to transformer.

Figure 1: General relationship between frequency and transformer component (IEC 60076-
18 Figure B.6)

Key
Recommendations for consistent measurements

IEEE C57.149 states “the test configuration can have an impact on the test results. It may
be difficult to determine if these minor variations are due to differences in test
configuration or some other physical change. Therefore, it is important to document the test
configuration and connections for future test repeatability.” It also states “grounding
techniques will have a significant effect on test results. Grounding techniques, including
selection of ground conductors as well as their routings, should therefore be precise,
repeatable, and documented.” CIGRE Brochure 342 and IEC 60076-18 also highlight the
need for consistent connections and transformer settings. The general recommendations of
these three documents can be summarized as follows:

 Transformer shall be completely isolated from high voltage

 Transformer tank shall be grounded „

 Test instrument shall be grounded „

 Transformer should be as close to “in service” condition as possible

o Note any difference such as lack of oil, transportation bushings, etc.

 All external bushing connections should be disconnected

 If applicable, the tap position for both DETC and LTC shall be recorded

 DETC shall be in the “in service” or “as found” condition

 LTC is recommended to be in the extreme raised position. If sweep is measured with the
LTC set to neutral position it should reach this from the raised position and both tap and
previous tap position should be noted.

 „Grounding leads should be as short as possible (without coiling leads) and of flat braid
type

 Solid connections should be made when attaching test leads to the terminals of the
transformer

 SFRA tests should be performed before the winding resistance test. If a winding resistance
is performed, demagnetize afterwards.

By following these rules, the technician will ensure that the measurements made are as
accurate and repeatable as possible. When analyzing the results it will therefore be easier to
determine if a difference in sweep response compared to fingerprint is due to an actual
mechanical change inside the transformer or simply the result of using a different setting or
connection.

Test setup and transformer settings

To highlight some of the effects that various settings on the transformer as well as the test
setup and the connections can have on the results, several measurements were carried out
over a two-day period on the Dyn1 67 kV/12.47 kV 12 MVA 3Φ transformer shown in
Figure 2. During this time, no internal abnormalities or physical changes occurred in the
transformer. Only the test setup, connections, and tap settings were changed. Additionally, a
winding resistance test was performed. The next section discusses the results in detail.

Figure 2: Dyn1 67 kV/12.47 kV 12 MVA 3Φ transformer

Effect of magnetization

The presence of residual magnetism in the core can influence results in the low frequency
range. Residual magnetism may be the result of performing a winding resistance test. For
such a test, direct current is injected into the winding; the core is magnetized and saturated
so that the voltage drop due to the inductance of the winding is excluded from the
measurement. If the core is not demagnetized after this test, significant differences could
show up on open-circuit SFRA measurements in the low frequency range.

To demonstrate the effect of core magnetization, an open-circuit SFRA sweep was carried
out on the phase B of the LV winding of the transformer with the core in the demagnetized
state. Winding resistance tests were then performed at 10 A, 25 A and 50 A. After each of
these tests, an open-circuit SFRA sweep was carried out. The SFRA sweeps done after the
resistance tests produced the same curve, irrespective of the DC current used for the test.
There was, however, a noticeable difference between the pre-magnetization and post-
magnetization curves, as shown in Figure 3 and 4.
Figure 3: Effect of magnetization on magnitude response (orange traces - pre-
magnetization and after demag)

Figure 4: Effect of magnetization on phase response


Figure 5: Difference between results pre and post magnetization of transformer

In the low frequency area, the curve shifts upwards and to the right after magnetization. A
shift in the first resonant frequency is observed at approximately 500Hz. A difference curve
was plotted to highlight the difference between the pre-magnetization curve and post-
magnetization curve.

As can be seen from the difference curve in Figure 5, there is a considerable difference in
the low frequency range of 10 Hz – 3 kHz. Minor differences are also observed at higher
frequencies.

Based on these observations, it is clear that care should be taken to ensure the core is not in
a magnetized state during SFRA testing. Ideally, winding resistance tests should not be
conducted prior to SFRA tests but if this is unavoidable – or if core magnetization is
suspected – the core should be demagnetized before starting the SFRA test.
Effect of removing the core ground

Often, the core-ground insulation resistance measurement is the first electrical test done on
a transformer. If the technician forgets to ground the core terminal after the test and begins
SFRA testing, the ungrounded core terminal may lead to a deviation in the curve. To
simulate this condition, the core was disconnected from the ground, as shown in Figure 6.

Figure 6: Core terminals disconnected from the ground

Figure 7: Magnitude response of transformer with (blue) and without (red) core grounded
Figure 8: Phase response of transformer with (blue) and without (red) core grounded

Figure 9: Difference in magnitude response between grounded and ungrounded core


Figure 10: (left) Braids
grounded at the bushing flange

Figure 11: (right) Braids grounded at the edge of the grounding pad.
The curves in Figures 7 and 8 were obtained with open-circuit sweeps performed on the
HV phase A of the transformer.

The magnitude difference between the curves shown in Figures 7 and 8 was plotted
separately, and is shown as Figure 9.

As can be seen from Figure 9, differences are noticeable up to 9 kHz but then the curves
essentially overlay up to 900 kHz, after which there are further noticeable differences. For
valid fingerprints and comparisons, the core should therefore be kept grounded during the
SFRA test.

Effect of the grounding braid length

Since SFRA measurements are sensitive, grounding plays an important role. The grounding
loop must be as short as possible and the grounding should be connected using braids at
the flange of the bushing where the signal leads are connected. To demonstrate the effect of
the length of the grounding connections, two measurements were made. The shortest
possible length was used for one measurement, as shown in Figure 10. The whole length of
the grounding braid was used for the other measurement, as shown in Figure 11.

The curves shown in Figures 12 and 13 were obtained. The difference curve is shown in
Figure 14.
Figure 12: Magnitude response with short and long ground connections

Figure 13: Phase response with short and long ground connections

Figure 14: Difference plot of responses with long and short ground connections

There are essentially no differences in the two curves until frequencies of 900 kHz and
higher are reached.

In the next set of open-circuit measurements, shown in Figures 15 and 16, the grounding
braids were removed altogether. The difference curve is shown in Figure 17. Once again,
there was no change in the low and mid frequencies but the deviation increased
dramatically in the high frequency range. There was a small difference between the two
curves at 100 kHz, which is lower than the frequency where differences produced by long
and short grounding braids were noted. Above 200 kHz, the differences increased
significantly.
Figure 15: Magnitude response with proper grounding (blue) and with no grounding (red)

Figure 16: Phase response with proper grounding (blue) and with no grounding (red)

Figure 17: Difference between response with proper grounding and with no grounding
The grounding arrange-ments for the instrument were also changed. But no differences
were observed whether the instrument was grounded to the transformer, to the substation
grid, or even not grounded at all. The magnitude and phase responses are shown in Figures
18 and 19. Note that even though the grounding of the instrument does not affect the
measurements, it must nevertheless be grounded at all times to ensure safe operation.

Figure 18: Magnitude response with test equipment grounded and ungrounded

Figure 19: Phase response with test equipment grounded and ungrounded

Based on these observations, the grounding of the instrument may not affect measurements,
but the position and effective length of the grounding braids does have an effect on the high
frequency range. Standard grounding procedures should therefore be followed to allow
accurate comparisons of SFRA results.

The second part of this article, which will appear in a future issue of Electrical Tester, will
look at the effect of tap position, reversing switches, length of shorting leads and the
influence of the test voltage used on SFRA testing. It will also cover delta stabilizing
windings and how to verify the correct functioning of the test instrument before concluding
with a summary of the main recommendations for dependable and repeatable SFRA testing.

Read part 2 here


Best practice for sweep frequency response
analysis (SFRA)– Part 2
20 September 2018

Robert Foster - Application Engineer, Megger USA


Sanket Bolar - Application Engineer, Megger USA

The first part of this two-part article, which appeared in a previous edition of Electrical
Tester and is still available on the Megger website here, provided a brief introduction to
sweep frequency response analysis (SFRA) testing, and looked at some of the factors that
can affect test results adversely if they are not taken into account when setting up and
performing SFRA tests. This second and concluding part of the article looks at more factors
that can influence test results, and ends by summarizing recommendations for best practice
in SFRA testing. Note that, to avoid confusion, the numbering of the figures follows on
from the numbering used in the first part of this article.

Effect of tap position

The tap position affects SFRA results over a wide range of frequencies so it is important to
record the tap position of all windings in the test report. To demonstrate the effect on results
of the tap position on the winding being tested, three measurements were taken at different
DETC tap positions on the HV winding phase A. The results are shown in Figures 20 and
21.

Figure 20: Magnitude response with DETC set to 1, 3, 5


Figure 21: Phase response with DETC set to 1, 3, 5

Differences are observed from the beginning of the sweep at 20 Hz up to 5 kHz where the
curves overlay until differences again appear from 20 kHz up to 400 kHz. Changing the tap
position on the non-tested winding also affects SFRA results, as can be seen in Figures 22
and 23. The LV winding of the transformer has an OLTC with 33 taps. The open-circuit
SFRA measurement was carried out on the HV phase A of the transformer at three different
tap positions of the OLTC on the LV side.

Figure 22: Magnitude responses with OLTC set to 16R, N, 16L


Figure 23: Phase responses with OLTC set to 16R, N, 16L

As can be seen, when the results obtained at the nominal tap position are compared with
those obtained at the extreme tap positions of 16R and 16L, there are differences at both
low and high frequencies. Once again the differences are observed at low frequencies up to
5 kHz and at higher frequencies between 40 kHz and 900 kHz. This illustrates how
different components of the transformer affect the frequency response in different ranges,
but no individual component is associated with a specific frequency range. In general, the
low frequencies are most affected by the core of the transformer, but just adding extra
windings to the circuit also affects the lowest frequency ranges. Differences would be
expected as the sweep moves into the mid-high and high frequency ranges as, when the
extreme tap positions are compared with the normal position, additional windings have
been added to the circuit. When comparing the two extreme tap positions to each other, the
curves follow each other very closely in the lower frequency ranges and then upon reaching
40 kHz, they diverge similarly to the comparison of the nominal to the extreme raised tap
position, though not as significantly.

Figure 24: Difference in magnitude response between 16R and N tap positions

Figure 24 shows the difference between an open-circuit sweep on the phase A high-side
winding with the OLTC set to the 16R tap position and a similar sweep with OLTC set to
the N tap position. There is a large deviation of 23 dB at 526 Hz and several more
deviations at 60 kHz and above. Although no faults had occurred in the transformer, simply
altering the OLTC tap changer position from nominal to fully raised significantly affected
the response at the lowest and highest frequencies.

Figure 25: Difference in magnitude response between 16R and 16L tap positions

When comparing the open circuit sweep on the phase A high-side winding with the OLTC
tap position set to 16R and the sweep with tap position set to 16L, as shown in Figure 25,
there is essentially no deviation until 40 kHz. Above this frequency, the curves diverge in a
similar way to those shown in Figure 24 but the difference is not as large. Note that the y-
axis scale in Figure 24 is different from that in Figure 25. Even though both measurements
have the same amount of windings in the circuit, there is still a difference in the response
due to the reversing switch, albeit a much smaller difference than between all OLTC
winding turns in the circuit and none.

The results show that the tap position of both the tested and the non-tested winding affects
the SFRA measurement results. If this is not taken into account when SFRA results are
being made, the analysis could lead to a core problem or connection issue being suspected
when in reality the differences relate only to a difference in tap changer position. When
comparing one tap position to the next or fully raised to fully lowered, the differences are
likely to be minor and may be overlooked, but if two taps at the extremes are compared, the
difference in response can easily be misinterpreted as a fault. To ensure proper analysis and
avoid misdiagnosis of faults it is therefore essential to note the tap changer position(s)
while performing SFRA tests and to make future measurements using the same tap
position(s) as those used for the fingerprint measurements.

Effect of the reversing switch

When testing a transformer with an OLTC that has a reversing switch, even if two
measurements are carried out at the same nominal tap position, the results may vary slightly
based on how the tap position is approached. This happens on designs where the reversing
switch contacts are positioned differently when the tap changer is lowered to the nominal
position compared with when it is raised to the nominal position.

Two open-circuit measurements were carried out, one approaching the nominal tap position
from R1, the other approaching from L1. The results are shown in Figures 26 and 27. It can
be seen that in this case there is no deviation between the curves, most probably because
the transformer has a reactive type tap changer. Nevertheless, care should still be taken
during SFRA testing always to approach the nominal tap position from the R1 tap, and this
should be noted on the report to ensure consistent testing in the future.
Figure 26: Magnitude plots comparing response when nominal tap position is approached
from R1 and from L1

Figure 27: Phase plots comparing response when nominal tap position is approached from
R1 and from L1

Effect of shorting lead length in short-circuit measurements

The shorting connections used for short-circuit measurements should be point-to-point –


that is, as short as possible. The length of the shorting leads can influence measurements in
the high frequency range as shown in Figures 28 and 29. The difference is plotted in Figure
30 and as can be seen, the curves follow each other closely for most of the sweep, but there
are differences when the frequency exceeds 1 MHz. Figure 31 shows both point-to-point
shorting (left) and shorting with long cables (right). As can be seen, the length of the
shorting leads can influence results in the high frequency area. It is, therefore, good practice
always to use point-to-point shorting.

Figure 28: Magnitude response comparing point-to-point and long shorting methods

Figure 29: Phase response comparing point-to-point and long shorting methods

Figure 30: Difference curve comparing point-to-point and long shorting methods
Figure 31: Point-to-point shorting and shorting with long cables

Effect of test voltage

When SFRA test sets were first introduced, different manufacturers used different test
voltages. Most modern SFRA test equipment is however standardized on a test voltage of
10 Vp-p. This is important because changing the test voltage can affect the sweep in the
low frequency region because of the core influence. At higher frequencies where the
winding dominates, the sweep becomes independent of the applied test voltage. An
example of this can be seen Figure 32, which shows the results of high-side open-circuit
tests carried out at 10 Vp-p, 5 Vp-p and 1 Vp-p. The transformer under test in this instance
was a Dd0 13.8 kV/480 V 1MVA 3Φ unit.

Figure 32: Magnitude response at 10 Vp-p, 5 Vp-p and 1 Vp-p


Figure 33: Difference between magnitude responses with 1 Vp-p and 10 Vp-p test voltages

Using 10 Vp-p as a reference, it can clearly be seen that as the voltage is decreased to 5 Vp-
p the magnitude curve is shifted upwards by 3 dB at low frequencies. This effect steadily
decreases until about 500 Hz where it increases slightly. The 10 Vp-p magnitude then
becomes slightly greater before the curves essentially overlay at 1.1 kHz and above. The
difference plot is shown in Figure 33. When comparing the 1 Vp-p trace to the 10 Vp-p
trace, the difference curve follows the same pattern but the difference starts at 8 dB rather
than 3 dB.

It can be seen that the test voltage affects the results in the lowest frequency ranges up to
about 1.1 kHz, shifting the response upwards on the dB plot as the voltage is decreased.
Once the frequency reaches 1.1 kHz, the response is essentially a winding response, and the
test voltage no longer affects the results. This means that when performing and analyzing
SFRA tests, the technician must know the test voltage that has been used in order to make
accurate comparisons with previous and subsequent tests. This will help prevent data being
misinterpreted as showing there is a problem in the core when the reality is that a
comparison is being made between a recent sweep carried out at 10 Vp-p and an older
sweep carried out at some other voltage.

Consideration of the delta stabilizing winding

Some transformers have a delta stabilizing winding in which two terminals (Y1 and Y2) are
connected together and grounded during service. It is recommended that these windings
remain shorted together during SFRA testing but be disconnected from ground. Figure 34
shows a set of low-voltage winding open-circuit measurements taken on a YNyn0D11 140
kV/11.5 kV 40 MVA 3Φ transformer. The transformer suffered a short circuit between
phases L1 and L2 on the low-voltage side, which developed into a three-phase fault before
it was cleared. To obtain the results shown in Figure 34, the connection between the Y1 and
Y2 delta winding was broken as well as the connection between the delta winding and
ground.
Figure 34: LV Open circuit measurements with tertiary open

The three phases show the expected response up to 10 kHz. At this point, since the three
windings are similarly constructed, the response from each of the phases would be expected
to be very similar. Instead, they start to significantly deviate after the local maximum of 26
kHz and don’t line up again until above 530 kHz, where slight variation would be expected
due variations in the connecting leads and wiring. These results could easily be interpreted
as showing that the windings had been distorted by the fault.

Figure 35: LV Open circuit measurements with tertiary shorted (Y1-Y2)

However the tests were performed again, this time with the Y1 and Y2 terminals shorted
together but still isolated from ground. The results can be seen in Figure 35. The response
in the low frequencies shows a little more difference than might be expected between L1
and L3 but as soon as the sweeps reach 2 kHz, all three phases essentially line up indicating
that the windings are in good condition. If the results of Figure 34 had been analyzed
without realizing the tertiary was open, it is likely that the transformer would have been
condemned as defective when, in fact, none of the windings was significantly harmed by
the fault. This demonstrates once again that tests must be performed in a consistent manner,
following the recommendations of the standards. In this case, the recommendation it is to
leave the Y1 and Y2 terminals shorted together but disconnected from ground during open
circuit tests.

Verifying instrument function

Even if the transformer is in the same configuration as in previous tests and all the
connections are made in exactly the same way, there is still one issue that could arise and is
largely beyond the control of the technician: there may be a problem with the SFRA test
instrument or its leads. There are two recommended methods for verifying that the
equipment is working properly. The preferred method is to connect the instrument to a test
circuit that has a known frequency response and compare the current trace to an earlier
reference trace. If a reference circuit is not available, the alternative method is to connect
the source lead to the measurement lead and see that the expected 0 dB response is
produced, within the accuracy of the instrument. These checks should ideally be performed
before starting tests on a transformer. At the very least, the reference sweep should be
conducted if the latest transformer trace does not line up with the fingerprint or if the test
equipment is suspected to be faulty. An example of a reference trace spanning a frequency
range 20 Hz to 2 MHz varying from -100 dB to -23 dB is shown in Figure 36.

Figure 36: Reference trace for verification of instrument operation

Summary

The examples in this paper clearly show that various normal conditions of the transformer
as well as connections, grounding, types of shorting cables used, and even tests done before
SFRA testing can all affect SFRA test results, sometimes leading to a false indication of a
problem in the transformer. Several conditions related to the reluctance of the core can
affect the low frequency region below 2 kHz. This was evident when the voltage level of
the input source was varied between 1 V and 10 V as well as when the sweeps were
performed before and after magnetizing the core.

Both instances of magnetization and changing the source voltage shifted the curve in
magnitude and in frequency. A similar phenomenon is observed if the core ground is
removed and not reattached or if, for some reason, the core ground becomes detached
internally. It is also possible for the results to be affected in the highest frequency ranges,
mainly above 200 kHz, leaving the response at lower frequencies unchanged. Factors that
can affect the higher frequencies include how connections are made, in particular how they
are grounded, what type of grounding braid is used and how the terminals are shorted in the
short circuit tests. Finally, changing the tap position, both DETC and LTC, can affect both
the high and the low frequency ranges.

To ensure reliable, repeatable measurements, SFRA tests must be performed methodically


and consistently. The test technician should note such factors as the test voltage used and
the tap position of both the DETC and LTC. It is recommended that the test be performed
with DETC in the “as found” or “in service” position and that the LTC should be placed in
the extreme raised position. If however the LTC is placed in the neutral position, it should
be switched to neutral from the raised tap position and the previous tap position should be
noted.

To minimize the influence of grounding, a flat braided ground that is as short as possible
should be used. When shorting leads are used, these should also be kept as short as
possible. Taking pictures or noting connections for future reference will help to minimize
the possibility of inadvertently making different connections in the future. SFRA testing
should be performed before winding resistance tests, but if a winding resistance has already
been performed or magnetization is suspected, the transformer should be demagnetized
before SFRA testing commences. Finally, any non-standard conditions, such as lack of oil
in transformer or transportation test bushings, should be recorded. By performing SFRA
measurements in a consistent manner, the results will be more comparable to the
fingerprint results making analysis easier and more accurate.

Sweep frequency response analysis


 Are other electrical tests necessary if I perform SFRA?

Yes, each test will provide a different piece of data your transformer. A power factor
test looks at the insulation of the transformer. Turns ratio and winding resistance
tests reveal the condition of the windings. SFRA provides information about the
mechanical integrity of the transformer and can help you determine if a transformer
has sustained any mechanical damage.

Each electrical test you perform gives you a bit more insight, and together they form
a more complete picture of your transformer's health. Sometimes a "second
opinion" from two or more tests on the same component can help you confirm a
suspected problem.

 Can I perform SFRA before oil filling a new transformer?


Per IEEE C57.149, testing with oil is the most common and preferred method for
frequency response analysis. Special consideration should be given to safety when
testing a transformer without oil so that excessive voltages are not applied. Presence
of oil changes the frequency response. Results with and without oil will cause
variations in the SFRA traces. Below is an excerpt from the IEEE guidelines:

"For new equipment, this my require the performance of two FRA tests after receipt
of the equipment at the final destination: 1) one test with the transformer in its
shipping configuration, 2) and one thest with the transformer assembled an oil-
filled as required for insulation resistance testing, to be used as baseline data for
future testing. If no shipping damage is suspected, the test in the as shpped
configuration may not be necessary as a receipt test"

Often, the manufacturer fills and drains the transformer before shipment. You
should be aware of the conditions in which an SFRA test was performed before
shipment from the factory. IEEE further states that:

“If the equipment is to arrive drained of oil, the shipping configuration should
specify that it will be tested pre and post movement without oil. If the equipment is
to be shipped after being drained of oil, it should be tested pre-movement without
oil. Testing the unit prior to shipment in this case without oil and prior to a first fill,
may not be adequate and could lead to false failures due to residual oil being held
in the windings, or additional oil draining from the winding during weeks of
shipment. If the equipment is to be shipped with oil, it should be fully filled for both
pre and post movement tests. If the equipment is to be shipped partially filled, it
should be tested with the same level of oil, or preferentially after oil has been
added. Ensuring oil is at the same level before and after transportation for partially
filled transformers can be difficult and sometimes leads to incorrect assessments.”

 Can SFRA test results from different manufacturers' test sets be compared?

Yes. SFRA test results, when performed correctly and under similar conditions
(correct grounding, same tap position and similar connections), are comparable.
Factors that can affect test results include residual magnetism and poor grounding
practices. Megger's FRAX software has the unique capability to import previous
results from any other manufacturer's test set and compare results.

 Do I have to perform electrical tests in a particular order?

For SFRA:

There is no particular order that needs to be followed for the open circuit and short
circuit tests. However, to increase efficiency, you might want to run the tests in an
order that will help you minimize lead changes

For complete electrical tests:

Excitation current and SFRA should be completed first and winding resistance test
last. This is to avoid any residual magnetism from the winding resistance test from
affecting the results of the other tests. However, If your winding resistance test set is
capable of demagnetizing the transformer efficiently after the test, you will not have
to worry about the test sequence.

 Does ambient temperature affect SFRA readings?

IEEE C57.149 states that “Large temperature difference, typically much more than
10 C, between two measurements will slightly influence the response at higher
frequencies.”

For all practical purposes, the effect of temperature on SFRA measurements is very
small and can be ignored unless there is big temperature variation between the two
comparison traces.

 How is a SFRA open circuit and short circuit test different from traditional tests?

Traditional open and short circuit tests are typically performed in factory to
determine the no load and copper losses taking place in the transformer. Rated
values are typically used when performing these tests. By determining the no load
and copper losses, you can determine the different components in an equivalent
circuit of a transformer.

Although they share similar names and connections, SFRA open circuit and short
circuit tests are completely different. The SFRA open circuit test looks at the
response of the core and winding, and the SFRA short circuit test isolates the
winding response of the transformer. These tests help you narrow down the areas
where the problem might be.

 How would I test a four-winding transformer (e.g.18 pulse or 24 pulse)?

You would have to run a total of 30 different tests.

o 12 open circuit tests, one on each winding (4 windings x 3 phases =12 tests)

o 18 short circuit tests:

 9 tests (From high side with three secondaries shorted one at a time)

 6 tests (From X side with other two secondaries shorted one at a


time)

 3 tests (From Y side with the last secondary shorted)

 In a post short circuit fault condition, do I need to perform both a power factor test
and SFRA?

In such cases, IEEE C57.152 recommends performing all electrical tests including
Power Factor and SFRA. A power factor test may reveal a change in capacitance,
and comparing this to a SFRA trace will help the diagnosis of any issues or failures
associated with transformer windings.
 Is there a published standard for SFRA?

The IEEE guide for SFRA is IEEE C57.149 Guide for the Application and
Interpretation of Frequency Response Analysis for Oil-Immersed Transformers.
Other relevant SFRA documents include: IEC60076-18 Ed. 1 – 2012, Std.
DL/T911-2004, and Cigré Technical Brochure No. 342, April 2008.

 Should I incorporate SFRA into my transformer maintenance program?

Yes. IEEE C57.152 Guide for Diagnostic Field Testing of Fluid-Filled Power
Transformers, Regulators, and Reactors recommends SFRA as a diagnostic test.
SFRA can often pick up mechanical issues that other electrical tests might miss.

 What do the different frequency ranges in SFRA mean?

A SFRA open circuit test will show the response of core and windings, while a
SFRA short circuit test only shows the response of windings. Each frequency range
corresponds to dfferent components in the transformer. Some general frequency
ranges are shown below.

o 20 Hz – 2 kHz: Main core deformation, open circuits, shorted turns, residual


magnetism

o 10 kHz – 20 kHz: Bulk winding component, shunt impedance

o 20 kHz – 400 kHz: Deformation within the main windings

o 400 kHz – 1 MHz: Tap winding

However, keep in mind that each transformer will have specific responses and the
frequency range given is for general reference only. For windings rated less than 72
kV, IEC recommends running the test up to 2 MHz.

 What is the difference between SFRA and Dielectric Frequency Response (DFR)?

SFRA and DFR are two completely different tests. SFRA looks at any kind of
mechanical changes inside the transformer whereas DFR is used to determine the
moisture present in cellulose (solid insulation) of oil filled power transformers. The
two tests have very different applications.

 What is the smallest rated transformer that I can use SFRA on?

There is no industry guideline for using SFRA based on a transformer’s VA


ratings. In theory, you can perform SFRA on a transformer of any size, and if
subsequent tests are performed under similar conditions, the results can be
compared and analyzed. Other electrical tests like transformer turns ratio, excitation
current and DC insulation tests will also give valuable information on smaller
transformers.
 Will SFRA work on dry type transformers?

Yes. SFRA looks at the response of the complex RLC network inside a transformer.
You can perform baseline or reference measurements on dry type transformers and
compare results over the years. For dry type transformers, you need to be aware of
the effect that ground capacitances can have on the traces. Additionally, the
response on the low side may have slight deviations because of low signal levels. A
very good ground plane will produce more repeatable measurements.

Q and A: Variable frequency tests


20 September 2018

Power transformers are costly, difficult to replace assets that are often used in critical
applications. It’s unsurprising therefore that many different techniques have been developed
for diagnostic and routine testing of transformers. So many techniques, in fact, that callers
to the Megger helpline are often confused, especially when it comes to variable frequency
tests. Here are some of the questions we’re regularly asked.

Q: Is variable frequency testing of a power transformer the same thing as an SFRA


test?

A: Yes – and no! SFRA (sweep frequency response analysis) testing is the best known of
the variable frequency transformer testing techniques, but it’s definitely not the only one.
There are many other transformer diagnostic techniques based on frequency, and each of
them has unique diagnostic functions and values. Other widely used techniques include
DFR (dielectric frequency response), narrow-band DFR, and FRSL (frequency response of
stray losses).

Q: What is SFRA testing and why is it useful?

A: SFRA testing, which can be conveniently performed with Megger FRAX test sets, is
used to check the mechanical integrity of transformer components such as the core, the
windings and the clamping structures. To carry out the test, a low-voltage signal is injected
into one end of a winding and measured at the other end so that the electrical transfer
function of the transformer can be determined. The test is typically repeated over a range of
frequencies from 20 Hz to 2 MHz. The results are compared with a reference curve that
was produced using the same technique when the transformer was new or known to be
undamaged. This technique reveals many types of fault, including core movements, faulty
core grounds, winding deformations, winding displacements, partial winding collapse, hoop
buckling and shorted turns. It’s important to note that SFRA is essentially a comparative
test. Without a reference curve for the transformer, the information provided by the test is
much less useful.
Q: Is DFR testing similar to SFRA testing?

A: No, it’s quite different. DFR testing, which is sometimes also known as frequency
domain spectroscopy (FDS), is essentially a series of tan delta tests carried out over a range
of frequencies. The frequencies used are much lower than those used for SFRA – typically
1 mHz (millihertz!) to 1 kHz. The results are usually presented in the form of a capacitance
and/or dissipation factor/power factor curve and, when they are used in conjunction with
insulation modelling they provide invaluable information about the condition of the
transformer’s insulation system, particularly the moisture content of cellulose insulation
and the oil conductivity. Megger offers DFR testing with instruments in the IDAX range.

Q: What’s the difference between DFR measurements and narrow-band DFR


measurements?

A: The measuring techniques used are similar but, as the name implies, narrow-band DFR
uses a much more restricted range of frequencies – usually from around 1 Hz to 500 Hz.
Also, the results are analysed directly rather than by using modelling techniques. It takes
much less time to carry out a narrow-band DFR test than a full DFR test – around two
minutes compared with twenty minutes or more – but the narrow-band test doesn’t provide
the estimated moisture content for the cellulose insulation. What it does do is to provide an
earlier indication of problems than traditional power dissipation factor (PF/DF) tests
performed only at power frequencydo. It also confirms that seemingly good PF/DF delta
values really are good and allows the transformer’s individual temperature correction (ITC)
factor to be determined. The Megger DELTA and Megger TRAX test instruments are used
to perform narrow-band (NB) DFR testing in addition to PF/DF tests.

Q: What is FRSL testing and what is it used for?

A: FRSL stands for frequency response of stray losses, and it’s the only test that can detect
strand-to-strand short circuits in a conductor bundle. It is also sensitive to problems that
lead to increased losses in transformer structural components such as the tank, the clamping
structures and the tie plates. Problems of this type may produce overheating gases in the oil.
FRSL measurements are made with the Megger TRAX test instrument. Also, if you’re
using a Megger FRAX test set for SFRA testing, FRSL test results are made available by
the software at the completion of the SFRA tests.

The do’s and don’ts of insulation power


factor testing – Part 3
20 September 2018

Jill Duplessis - Global Technical Marketing Manager and Editor

This is the third in a short series of articles dealing with power factor (PF)/ dissipation
factor (DF) tests, which are widely used to assess insulation condition in transformers and
other electrical assets. The first two articles, which are still available online, looked at the
theoretical background for these tests, safety and general insulation knowledge. This
instalment looks at terminology associated with PF/ DF testing, and at general testing
knowledge as it applies to these test techniques.

NOMENCLATURE

In any technical area, it’s easy to become confused by unfamiliar nomenclature and
terminology. This is certainly the case for power factor/ dissipation factor tests, so we’ll
start by looking at some of the most common sources of confusion using the same Do’s
and Don’ts format we adopted in the previous article.

Nomenclature Don’ts

DON’T get bogged down by DF versus PF

Figure 1:

Dissipation factor
versus power factor – what are the differences and do they matter? In reality, DF and PF are
calculated differently (Figure 1); however, both are calculated from the measurements
made in the same test and both describe the inefficiency – or efficiency – of an insulation
system. And, for most asset insulation systems that engineers will need to test, the
numerical values of DF and PF will be almost the same. That means deciding between DF
and PF is largely a matter of personal preference.

It is relevant to note, however, that while PF and DF insulation test results may be largely
interchangeable, the phrase ‘power factor’ has multiple meanings in our industry. In
generator and load studies, for example, power factor is the ratio of the active power in a
circuit to the total power in that circuit. For this reason, some power plant engineers – and
others – feel that talking about power factor testing in relation to insulation condition is
confusing. Dissipation, on the other hand, usually refers to using something wastefully.
Since PF/ DF testing actually does report the amount of energy that is lost as heat, it can be
argued that dissipation factor is more accurate and reflective of the measurement being
made. In fact, the δ angle used to calculate dissipation factor (Figure 1) is also referred to as
the loss angle. Nevertheless, there is once again room for personal preference.

DON’T get bogged down with dielectric versus insulation

The terms dielectric and insulation are often used interchangeably, which is reasonable
because a dielectric is an electrical insulator. Bear in mind though that there are many other
types of insulation – thermal insulation, acoustic insulation and more. So if you’re talking
about insulation it’s important to make sure that it’s clear from the context exactly what sort
of insulation you mean. Dielectric is a more specific term – its only meaning is a material
that will support an electrostatic field without allowing conduction.

Nomenclature Do’s

DO be aware of slang

A dissipation factor test is often referred to as a “tan delta” test, while a power factor test is
sometimes called a “doble” test. These terms have no special meaning. They’re just
informal ways of referring to those specific tests.

DO keep test frequency in mind

The name ‘power factor test’ implies that the test is carried out at or near line frequency (50
or 60 Hz). PF/DF tests that are carried out over a range of frequencies show how PF/DF
and capacitance change with frequency but such tests fall into the category of dielectric
frequency response (DFR) testing, which is beyond the scope of this article.

GENERAL TESTING KNOWLEDGE

General Testing Knowledge Do’s

DO familiarise yourself with history


Figure 2:
A test’s history,
particularly one as old as that of a PF/DF test, may provide valuable insights into the hows
and whys of a test, including how common misconceptions about a test can be created and
perpetuate. For PF/DF testing, there are publications from the early 1900s that deal with PF
testing of cable insulation, but capacitance and PF testing really took off with the
popularisation of capacitance graded bushings. Manufacturers of these bushings found that
they couldn’t consistently identify localized problems in the bushings’ insulation systems
using DC insulation resistance testing, so they turned to PF/ DF tests as a better alternative.
Because of their construction (Figure 2), capacitance-graded bushings are the perfect
specimens for these tests. As individually insulated conductive layers within the bushing
short circuit (i.e., a localized problem), C1 capacitance increases, providing a warning of a
potentially imminent failure. Meanwhile PF/DF testing is an averaging measurement so
deteriorating insulation of one layer, even if surrounded by layers in excellent health, will
influence the PF/DF test result. The increase of PF/DF will alert of the problem.

DO remember that PF/ DF testing was first popularised for bushings


This is important because as tests started to be used more widely, exaggerated claims for
the benefits of PF/DF testing followed. ‘Earlier detection of problems in bushings’
(meaning earlier than was possible with DC testing) became, for example, ‘early detection
of problems in insulation’ which in many cases is grossly misleading. And ‘very sensitive
to moisture contamination in the tap compartment of the bushing’ became ‘very sensitive to
moisture contamination’ but, in many cases, this is only true if the PF tests are repeated at
many different frequencies or, in other words, if a DFR test is carried out.

DO simplify testing by understanding dielectric representations and test modes

Figure 3:

An understanding of dielectric
representations and test modes will make PF/DF and capacitance testing much easier. It’s
always beneficial to test the smallest possible section of insulation – smaller is always
better, but how can you segment an asset’s insulation system so that you can test one
segment at a time? The dielectric representation (e.g., as given in Figure 3) shows how this
can be done, and the test modes provide a tool for segmentation. Unfortunately, it often
happens that when you’re doing a test, you have to energize more than one segment
(sometimes the whole insulation system) but you’ll still need to measure just one segment
at a time, and test modes make this possible. As an aside, this means that when you’re
considering the capability of an instrument to charge a test specimen, it’s essential to keep
in mind that (b)that you may be charging much more than the segment you’re actually
measuring.

DO pay attention to the dielectric representation

As we’ve already said, smaller insulation segments are better but the test engineer may not
have total control over how an insulation system can be segmented for testing. For
example, with capacitance-graded bushings, it’s impossible to stick a test lead into the
bushing and test every layer individually. The dielectric representation is a diagram that
identifies each component or group of components that will take up a unique voltage when
the asset is energised. It is left to the engineer carrying out the test to determine whether or
not a component is accessible for a test lead to be attached. Note that the number of unique
voltages/component groups present in an asset determines the number of insulation systems
needed.

DO take note of the way insulation between components is depicted

In a dielectric representation, the insulation between components is depicted as a single


capacitor (Figure 3). This is adequate, because a dielectric representation is simply an aid to
testing. In contrast, a dielectric model is used to predict the electrical behaviour of an
insulation system, and a single capacitor representation would not be adequate for this.

When you’re deciding how to test a transformer, a helpful rule of thumb is to count the
number of separable accessible windings, as demonstrated by these two examples. The first
is an autotransformer with an inaccessible tertiary winding. This transformer has three
windings – series, common and tertiary – but how many of these are separable accessible
windings? The tertiary isn’t because its connections are not brought out of the tank. The
series and common windings are accessible, but they’re not separable, so they count as one
winding. There is therefore only a single separable accessible winding and the transformer
must be tested as a single-winding transformer. Now let’s consider a four-winding Δ-Y-Y-Y
transformer that has all winding terminals accessible except for the three secondary neutral
terminals which are connected together and grounded within the transformer. In this case,
the number of separable windings is two: the delta winding and the three wye-connected
secondary windings, which have to be treated as one winding since they can’t be separated.
As for accessibility, while the delta winding is fully accessible, the neutral terminals of the
secondary group (Y-Y-Y) are not, making it impractical to disconnect them from ground.
This means that although this is a four-winding transformer, it will need to be tested using
the two-winding transformer test procedure, bearing in mind that the test results may not be
entirely meaningful since the secondary winding cannot be lifted from ground.

DO make best use of test modes

The feature of the PF/DF test set that enables it to support test modes is the guard circuit.
This is arranged so that all current resulting from the application of the test voltage will
seek to return to the guard point. The current return paths to the instrument are provided by
up to three test leads – the ground lead and two low-voltage leads (red, R, and blue, B) –
that are connected between the instrument and the asset under test. Three test modes are
available: GST-Ground, GST-Guard and UST (Figure 4).

Figure 4:
The GST modes are grounded specimen tests. The “-Ground” or “-Guard” designation
indicates where the test instrument will internally connect the LV leads, the use of which is
optional. The UST mode is an ungrounded specimen test; current flow in the LV leads is
measured. The ground lead is always used and serves to compare the specimen ground to
the voltage supply ground. It is also a current-carrying lead used for measuring or guarding.

DO be prepared

Dealing with external variables is one of the biggest challenges when performing PF/ DF
tests. These variables include temperature, humidity, surface leakage and the quality of test
preparations. All of these will be dealt with in more detail in later sections.

General Testing Knowledge Don’ts

DON’T forget that success depends on testing the smallest possible amount of
insulation

PF/DF testing provides a result that tells you about the average condition of the insulation
system under test. If the system you’re testing is large, a small, localised problem area may
not significantly influence the average and will therefore not be detected. This means you
should always look at the size of the specimen you are testing, as this will help shape your
expectations for the diagnostic capabilities of the test. A real-life example relates to a large
single-phase autotransformer with an inaccessible tertiary winding. This had severe
tracking on the collar around the lower insulator of one of the bushings. Only one PF/DF
test and capacitance measurement could be performed on this transformer. This was done
with all winding terminals connected together and energised for a single winding-to-ground
insulation measurement. Because of the large size of the test specimen and insulation under
test, the localised contamination made no measurable difference to the test results
compared with historical measurements.

DON’T rely too much on scalability

As mentioned at the beginning of this series of articles, we evaluate PF/ DF test results
rather than relying solely on dielectric loss results because PF/ DF results allow
comparisons to be made between insulation systems of different sizes. It is, however,
important not to take this too far. It’s certainly permissible to make comparisons of this
type, but bear in mind that the test on a larger system will not be as searching as the test on
a smaller system. In other words, the results are not completely scalable. Even if both
systems give identical PF/DF results, this doesn’t mean it’s safe to assume that the
insulation in both is in the same condition. As the earlier example showed, a serious
problem can be concealed behind a good PF/DF value if the problem area is small and the
system under test is large.

DON’T have unrealistic expectations for the test


Figure 5:

Increasing levels of
some contaminants, particularly moisture, don’t result in a noticeable change in PF/DF
until the level of contamination reaches a certain threshold. Figure 5 illustrates this.

On the y-axis, dissipation factor at 50 Hz is plotted on a logarithmic scale. On the x-axis,


percentage moisture in cellulose is plotted on a linear scale. The diagram shows DF curves
for a new transformer (blue) and for a service-aged transformer (red) as they become
increasingly contaminated with moisture. At low moisture levels, the slopes of both curves
are relatively flat, which means that, in this region, DF does not increase noticeably as
moisture increases. Someone relying on DF measurements would therefore be unaware that
moisture contamination is present and is increasing. It is not until the moisture content
reaches 2.5% that the slopes increase marginally and not until 3.5% that they increase
noticeably. From this point on, it can be said that DF testing has increased sensitivity to
moisture. Notice, however, that this increase in sensitivity doesn’t occur until at least 3%
moisture content, whereas most operators of power transformers start to become concerned
when water content reaches 1.7% and by 2.5% they would typically be scheduling a visit
by a processing unit to dry out the transformer. It’s also worth noting the divergence
between the curves for the new and service-aged transformers. This shows that the same
dissipation factor may represent a range of moisture contamination conditions, depending
on the age of the insulation system. For example, a dissipation factor of 0.3% corresponds
to a moisture content of 0.5% in the service-aged transformer, but to a moisture content of
2% for the new transformer. This presents a challenge as it’s not easy to quantify ageing in
a transformer – years of service is not a reliable indicator of ageing for an insulation system
as environmental and system variables, maintenance and many other factors play a role in
how insulation ages.
The fourth article in this series, which will appear in the next issue of Electrical Tester, will
deal with test preparation, the choice of test voltage and analysis of the test results.

The do’s and don’ts of insulation Power


Factor testing – Part 4
01 December 2018

Jill Duplessis - Global technical marketing manager and Editor

This is the last in a series of articles about power factor (PF) and dissipation factor
(DF) tests, which are widely used to assess insulation condition in transformers and
other electrical assets. The first three articles, which are still available online, looked
at theory, safety, general insulation knowledge and terminology. This instalment
covers test preparation, choice of test voltage and analysis of results.
TEST PREPARATION

With any kind of electrical testing, careful preparation is the key to safe working and to
obtaining results that are accurate and reliable. This section provides details of the
preparations recommended when carrying out power factor/ dissipation factor tests, using
the same do’s and don’ts format adopted in the previous articles.

Test Preparation Do’s

DO check that the transformer tank is well grounded

A poorly grounded transformer tank will often lead to very strange test results, possibly
including negative values for DF/PF, and a lot of time can be wasted trying to discover
what’s gone wrong.

DO physically and electrically isolate each winding’s terminals/bushings

Adequate physical isolation is essential. In particular, it is not satisfactory to rely on rubber


blankets sandwiched between the bushing terminals and the busbars to provide physical
isolation. Otherwise, the insulation characteristics of the rubber blanket will influence the
measurement.

DO consider the consequences of leaving busbars attached to bushings

If the busbars are left physically connected to the bushings, the measured winding
capacitance – Cw-g – will be greater than expected because the insulation system between
the winding under test and ground can’t be measured independently; the busbar left
attached to the winding introduces additional capacitance into the measurement, so the
result obtained is actually Cw-g + Cbus. Also, the winding-to-ground PF/DF measurement
is less meaningful as it cannot be assumed to be representative of the winding-to-ground
insulation only. This is because the insulators along the attached bus frame will be included
in the measurement and, since PF/DF is an average measurement, the condition of the
insulation between the busbar and ground may influence the winding-to-ground power
factor, making it appear better or worse than it really is. Problems in the winding-to-ground
insulation will be harder to detect, as more components are included the measurement.
Finally, leaving the busbars physically connected increases the possibility of results being
influenced by surface leakage.

DO think about the prevailing humidity

If the relative humidity is greater than 80%, avoid testing. This is particularly important for
bushings, instrument transformers and other assets that are susceptible to the effects of
humidity and condensation. Note that as the surface area of the test specimen becomes
larger in relation to its overall insulation size, surface leakage – and, therefore, humidity –
becomes more likely to affect the measurement.

DO short-circuit each separable winding

Short-circuit each winding by connecting its terminals/bushings together, and remember


to include the neutral when present. If the neutral is grounded, disconnect the ground
connection for the test and reconnect it when the test is complete. Use only bare copper or
aluminium wire for the shorting jumpers and make sure that all of the shorting jumpers
have adequate clearances from grounded surfaces. If this is impossible to achieve, it is best,
albeit undesirable, to forego short-circuiting of the winding all together. Note that because
the windings must be short-circuited for testing, the turn-to-turn insulation is not stressed
during PF/DF testing. For this reason, it is important to turn to other tests (like excitation
current or ratio tests) to check turns insulation. The effect of failing to short-circuit the
windings depends on the transformer but, in general terms, an inductive component is
introduced into the measurement, which may be seen as an increase in the loss angle, with a
corresponding increase in the measured PF/DF.

TEST VOLTAGE

Test Voltage Don’ts


DON’T simply accept the test voltage set by the instrument

Many instruments automatically set an industry-standard test voltage as soon as the


winding ratings and details of the transformer configuration have been provided. While
most of the time this test voltage will be satisfactory, keep in mind that the automatically
populated test voltage is only as appropriate as the accuracy of the nameplate data entry.
Therefore, it is wise to know the desired test voltage in advance and double check this with
what the test software will be directing the instrument to supply before commencing with
the test. Remember that the line-to-ground single-phase test voltage should not exceed the
line-to-line voltage rating of the winding being energised and, for transformers with a
graded insulation system, the test voltage should always be less than the voltage rating of
the neutral terminal.

DON’T test oil-filled transformers in the absence of oil

Even if a very low test voltage is used, it is not a good idea to carry out (line frequency)
PF/DF tests on service-aged oil-filled transformers in the absence of their insulating fluid.
This is sometimes suggested as a way of assessing moisture content and predicting the
success of a dry out but it’s inadvisable as the risk of fire outweighs the value of the
information gained, particularly as PF/DF testing is relatively insensitive to the presence of
low levels of moisture. Dielectric frequency response (DFR) testing is a better alternative.

Test Voltage Do’s


DO remember the value of power factor tip-up tests

A power factor tip-up test can be a very revealing diagnostic. With oil-paper insulation
systems, PF/DF tests performed at all voltages (up to the recommended maximum) are
expected to give the same value of PF/DF. With the tip-up test, an extra PF/DF
measurement is made with a different test voltage, and the results compared. If this reveals
that the measurements are voltage dependent, this suggests that either the insulation is
degraded or that there is a problem in the test circuit. Despite its usefulness, the tip-up test
is often overlooked.

DO take advantage of the smart tools provided by a (Megger PF/DF) test instrument
When tip-up testing isn’t carried out, these smart tools are particularly useful. Typically,
they assess voltage dependence by looking at the harmonics generated in the dielectric
response of an asset, as the level of harmonics is known to correlate with voltage
dependence. If the smart tool determines that the harmonic levels are high, it warns the
instrument user that further investigation, which usually takes the form of a tip-up test, is
needed.

ANALYSIS OF RESULTS

Analysis Do’s
DO take into account the transformer’s history

Historical information about loading, faults, maintenance, repairs and previous test results
can be a very useful aid to shaping expectations and to interpreting the most recent test
results.

DO assess capacitance and total current first

If possible, refer to previous test results (reports, nameplate data, etc.) and compare the
latest results with the historical data. If the winding capacitance has changed by more than
± 1% (or for C1 for bushings by more than 5 to 10%) further investigation is needed before
PF/DF testing is carried out. Changes may indicate that you are not testing what you
thought you were testing, or that the physical attributes of the insulation have changed
notably. Without access to benchmark values, look out for abnormally low
current/capacitance results. For power transformers this means less than around 1 or 2 mA
for CH and CL measurements. In these cases, the culprit may be, poor grounding,
substandard test connections, or an incorrectly selected test mode. Low values are not,
however, necessarily a concern for CHL (and CHT) measurements, as they may simply
indicate the presence of a grounded shield between the high- and low-voltage windings.
Also, in some multi-winding transformers, one of the windings may itself look like a
grounded shield to the test circuit – for example, CHT with an LV winding between CH and
CT.

DO keep in mind general PF/DF behaviour

PF/DF reflects how efficiently the insulation is fulfilling its purpose of maintaining
electrical isolation between points of different potential. With few exceptions, 0% PF/DF
indicates a system with zero losses but most insulation systems in acceptable condition
have inherent loss. Therefore, measured PF/DF can be “too low” – that is, lower than
expected or even negative. Always remember that the key to analysing PF/DF results in
practice is to look for changes from previous measurements.

DO take into account the challenges of interpreting results

Assigning significance to a change in PF/DF test results is always challenging because


these are averaging tests. This means it’s impossible to know, without further investigation,
whether the change is due to general, widespread contamination, ageing or a localised
problem. This is important because each of these conditions needs to be treated differently.
Localised problems significantly increase the risk of dielectric failure and they require
immediate attention. They cannot be identified by PF/DF testing alone as they may produce
only a small change in the PF/DF value because of the averaging influence of the
surrounding healthy insulation. In other words, even a very small increase in PF/DF could
be the manifestation of a very serious problem. In contrast, ageing of assets is to be
expected and any change in PF/DF that results from the ageing process might merely be an
indication that increased vigilance, in the form of more frequent testing, is desirable. PF/DF
changes resulting from widespread contamination of an asset don’t necessarily need to be
addressed immediately – it may be acceptable to schedule “cleaning” during the next
planned maintenance period. If the contaminant is water, however, by the time it affects the
PF/DF results the transformer may already have passed the point where it would have been
prudent to schedule processing/drying out.

DO make sure you’re comparing apples with apples

Any observed change in PF/DF requires attention, but before you jump to hasty
conclusions and spend time carrying out further tests and evaluations, make sure that
you’ve eliminated the influence of “external” test variables, such as temperature, test
preparation, surface leakage on the bushings and the overall condition of the bushings.

DO remember that temperature is a big deal!

All PF/DF results need to be normalized to their equivalent 20ºC values, but there’s no
“universal” temperature correction curve that will suit all cases! Ageing, moisture and
contamination all affect thermal response and this has led some standards bodies to
recommend discontinuing the use of temperature correction curves and tables. The solution
is Individual Temperature Correction , or ITC – this is correction based on the unique
thermal characteristics of the asset under test, and is implemented in selected test
instruments in the Megger range.

DO take into account the influence of bushings

The active insulation of high, low and tertiary bushings is included in CH, CL and CT
measurements, respectively. Therefore, the condition of a winding’s bushings may
influence the PF/DF results for that winding, making them appear better or worse than they
really are. In either case, the condition of the materials inside the transformer tank is
masked. In some cases however, the bushing insulation systems taken together may make
up only a very small part of the insulation included in the CH, CL and CT measurements. In
these cases, changes in the condition of the bushings do not noticeably affect the winding
PF/DF measurements. Nevertheless, C1 tests should be carried out on the bushings
wherever possible. When this is done, a simple calculation (provided by the test
instrument’s software) is all that’s needed to subtract the bushing insulation contribution
from the CH, CL and CT measurements, eliminating all uncertainty.

DO keep in mind that trending is the best way to analyse PF/DF

Trending the results of PF/DF tests for an asset over a period of time makes it much easier
to see changes and to decide whether those changes are significant. And modern recording
and reporting software, such as Megger’s PowerDB, makes trending easy and convenient
(Figure 1 below).
Figure 1

Analysis Don’ts

DON’T rely on general guidelines to analyse results

General guidelines and PF limits shouldn’t be used unless there are no previous test results
available for comparison. If you have no alternative but to use them, remember that they
are exactly what the name suggests – general guidelines only – so be very wary about
relying on them to make important decisions. Nevertheless, a transformer that is
constructed with relatively low loss materials and which has been subjected to careful
handling during manufacture and assembly in the field should meet the limits in Figure 2.
DON’T panic!

When PF/DF tests give unexpected results that may potentially have serious implications,
remember there are many reasons why PF/DF may have changed, and not all of them are
major causes for concern. Factors that commonly produce changes in PF/DF include: HV
bus not disconnected before testing; one or more HV bushings in poor or deteriorating
condition; excessive surface leakage across the HV bushings due to rain, snow, etc; use of
rubber blankets to achieve electrical clearances; ungrounded tank (which may produce a
negative power factor measurement!); inaccurate temperature correction; moisture or other
contaminants; general ageing of the asset; partial discharge; internal tracking; and an
unintentional resistive path to ground.

DON'T struggle alone!

PF/DF tests are an invaluable aid to assessing the condition of insulation in transformers
and other assets. Most of the time, performing these tests is straightforward, especially with
the aid of the guidance provided in this series of articles and when up-to-date instruments,
like those in the Megger range, are used. Even more information is readily available online
in the form of articles in the TLM (transformer life management) series on the Megger
website. Titles currently available include Moisture in Power Transformers, Dynamic
Measurements of On-load Tap Changers, Individual Temperature Correction, Measuring
Transformer Winding Resistance, Transformer Core Demagnetisation, Power
Factor/Dissipation Factor and Capacitance, Electrical Testing Efficiency through Test
Lead Management , Excitation Current and Oil Tan Delta (Sweep Frequency Response
Analysis is available by request.). It is nevertheless still possible that there will be
occasions when difficulties are encountered in executing tests or interpreting the results. In
such cases, there’s no need to struggle alone – the Megger technical support service, which
has unparalleled practical experience of testing in power environments, is ready to help. To
access the service, simply contact your nearest Megger location.

Q and A: Cable fault location


01 December 2018

Q: Can I detect all types of cable fault with a Megger insulation tester?

A: No! Cables can and do fail in many different ways. Some of the faults can be detected
with a straightforward insulation test, but many cannot. An insulation tester will not provide
useful information about short-circuit, high resistance and flashing faults. To see why, think
of a 33 kV cable with a high voltage flashing fault that has a breakdown voltage of 13 kV –
that is, the fault only appears if the cable is subjected to a voltage of 13 kV or more. It’s
clear that an insulation tester, which will typically have an output voltage of 5 kV or 10 kV,
depending on the model, will not be able to detect this fault. In a case like this, a DC hipot
test is needed to determine the breakdown voltage.

Q: I have a full short-circuit fault and I’m trying to use Digiphone to pinpoint it but I
can’t hear the surge in the headset. What should I do next?

A: Not being able to hear the surge definitely makes it more difficult to pinpoint the fault,
but there are several things you can do about this. The first is to use an audio frequency
tracing unit. With this, as you follow the route of the cable and pass the location of the
fault, you should notice a change in the characteristics of the received signal. Typically, the
signal will quickly become weaker or the instrument’s depth display will show a sudden
change.

Another approach is to use a Digiphone but look at the electromagnetic signal it receives
rather than searching for the sound produced by the surge. Even if the sound is non-
existent, the electromagnetic pulse can usually still be detected as the current discharge into
the cable will create an electromagnetic field around it. Once the location of the fault has
been passed, this electromagnetic field will be much weaker, as there will be little or no
current in the cable after the fault. It is easy to see this change as the Digiphone shows the
received electromagnetic field strength directly on its screen.

Finally, in some short-circuit faults, the outer jacket of the cable is damaged. In these cases,
the normal sheath-fault pinpointing methods can be used.

Q: I have many PILC cables in locations where there’s a lot of moisture in the ground.
I find that most of the time the arc reflection method (ARM) of fault location doesn’t
provide satisfactory results with these. What’s the best method to use in such cases?

A: The reason for this problem is that damping – a result of the high conductivity of the
moisture surrounding the cable – attenuates the returning TDR pulse from the fault location
so much that it becomes impossible to detect. Transient methods are more likely to be
successful. An example is the ICE or impulse current method, which works by monitoring
HV reflections rather than the relatively small pulse produced TDR. Alternatively, Megger
cable fault location instruments and Megger test vans support unique techniques such as
double surge and ARM Plus. These techniques are demonstrably superior to the standard
ARM method and are much better able to identify and locate ‘difficult’ faults.

You might also like