ATACHMENT B - PROTOCOL for VISUAL INSPECTION / MECHANICAL VERIFICATION

Electrical Testing of Power

Transformers
Electrical Testing Of
Power Transformers

In collaboration with
Siemens Power Transformer Division

Siemens AG
Transformatorenwerk Dresden
Overbeckstraße 44, 01139 Dresden, Germany
Tel: (03 51) 8 44 41 00

Foreword

The first considerable long distance transmission of

electric energy was performed successfully on the 24th of
August 1891, only possible with the help of a transformer,
invented by Michail Doliwo-Dobrowolski in Berlin.

Today transformers are indispensable in terms of the

usage in high voltage grids and special applications like
HVDC transmission or electric arc furnace plants. The
complexity of a transformer and its components make it
necessary to monitor the manufacturing process and
verify the operation without failure, fulfilling the required
Standards and customer’s expectations.

Therefore this booklet gives an overview about electrical

testing of power transformers, combining the knowledge
and experience of Siemens’ designers, consultants and
test field engineers.
 Table of contents

1 Introduction 7 6 Measuring the no-load loss &

1.1 Power Transformers Dresden 7 no-load current 39
1.2 Transformer components 8 6.1 Purpose and Standards 39
1.3 Standards and Norms 9 6.2 General
1.4 Types of tests 9 Fehler! Textmarke nicht definiert.
1.5 Test sequences 10 6.3 Measuring circuit 42
1.5.1 IEC Standard 10 6.4 Measuring procedure 44
1.5.2 IEEE Standard 11 6.5 Appendix: Measure arrangement of
1.5.3 Siemens recommendation 12 customized vector group 45
1.6 Reasons for Transformer failure 13
7 Separate source AC withstand voltage test
2 Voltage stresses & dielectric integrity 15 (Applied voltage test) 47
2.1 General 15 7.1 Purpose and Standards 47
2.1.1 Lightning over-voltages 16 7.2 General 47
2.1.2 Switching over-voltages 16 7.3 Measuring circuit 48
2.1.3 Temporary over-voltages 17 7.4 Measuring procedure 49
2.2 Transformer’s insulation verification 17 7.5 Appendix: Measure arrangement of
2.3 Test voltages 18 customized vector group 50
2.3.1 Alternating voltages 18
2.3.2 Impulse voltages 18 8 Switching impulse test &
2.4 Electrical test fields 18 Lightning impulse test 51
8.1 Purpose and Standards 51
8.2 General 51
Routine tests 8.3 Measuring circuit 53
8.4 Measuring procedure 54
3 Measurement of winding resistance 21 8.5 Test equipment for high-voltage tests 55
3.1 Purpose and Standards 21 8.6 Appendix: Measure arrangement of
3.2 General 21 customized vector group 57
3.3 Measuring circuit 22 8.7 Transferred overvoltage 58
3.4 Measuring procedure 23
3.5 Appendix: Measure arrangement of 9 Partial Discharge Measurement 61
customized vector group 25 9.1 Purpose and Standards 61
9.2 General 62
4 Voltage ratio & verification of vector group 9.3 Measuring circuit 63
or phase displacement 27 9.4 Measuring procedure 64
4.1 Purpose and Standards 27
4.2 General 27 10 Induced voltage tests 67
4.3 Voltage ratio measurement 29 10.1 Purpose and Standards 67
4.4 Vector group verification 30 10.2 General 67
4.5 Appendix: Measure arrangement of 10.3 Measuring circuit 68
customized vector group 32 10.4 Measuring procedure 69
10.5 Appendix: Measure arrangement of
5 Short-circuit voltage & load loss 33 customized vector group 72
5.1 Purpose and Standards 33
5.2 General 33 11 Insulation resistance measurement 73
5.3 Measuring circuit 36 11.1 Purpose and Standards 73
5.4 Measuring procedure 36 11.2 General 73
5.5 Appendix: Measure arrangement of 11.3 Measuring circuit 74
customized vector group 38 11.4 Measuring procedure 74
11.5 Appendix: Measure arrangement of
 customized vector group 75
12 Test on tap-changers & auxiliary equipment 77 19 Frequency response analysis
12.1 Purpose and Standards 77 measurement (FRA) 105
12.2 Tests on Tap-changers 77 19.1 Purpose and Standards 105
12.3 Tests on AuxiIiary equipment 78 19.2 General 105
19.3 Measuring circuit 106
13 Check of built-in current transformers 79 19.4 Measuring procedure 107
13.1 Purpose and Standards 79
13.2 General 79 20 FDS measurement for moisture estimation 109
13.3 Measuring procedure 80 20.1 Purpose and Standards 109
13.4 Measuring circuit 80 20.2 General 109
20.3 Measuring circuit 110
20.4 Measuring procedure 111
Type tests
21 Measurement of insulation capacitances &
14 Temperature rise measurement 81 loss factor (tan ) 113
14.1 Purpose and Standards 81 21.1 Purpose and Standards 113
14.2 General 81 21.2 General 113
14.3 Measuring procedure 83 21.3 Measuring circuit 114
14.4 Measuring circuit 86 21.4 Measuring procedure 114
14.6 Appendix: Measure arrangement of
customized vector group 87 22 Measurement of excitation current
14.5 Hot-spot measurement 88 with 400V 115
22.1 General 115
15 Measurement of cooling losses 89 22.2 Measuring procedure 115
15.1 General 89
15.2 Measuring circuit and instruments 89 23 Calibration of the Winding Temperature
Indicator (WTI) 117
16 Sound level measurement 91 23.1 General 117
16.1 Purpose and Standards 91 23.2 Calibration 118
16.2 General 91
16.3 Measuring circuit 92
16.4 Measuring procedure 93

Special tests
17 Measurement of zero-sequence impedance 95
17.1 Purpose and Standards 95
17.2 General 95
17.3 Measuring circuit 97
17.4 Measuring procedure 97
17.5 Appendix: Measure arrangement of
customized vector group 99

18 Measurement of the harmonics of the

no-load current 101
18.1 Purpose and Standards 101
18.2 General 101
18.3 Measuring circuit 102
18.4 Measuring procedure 102
18.5 Appendix: Measure arrangement of
customized vector group 103
 Introduction 7

1 Introduction

This booklet is to be understood as a supportive guide, giving a summary about applied electrical
measurements. These electrical tests are essential to estimate and verify the specifications and the
dielectric strength of the transformer. It also provides the sense of manufacturer’s internal quality
protection and aspiration of constructing a valuable product. Although some calculations are necessary
for preparing and analysing the measurements, we are not considering the matter of scientific approach
or any mathematical explanations. The essential content is to inform and give formal explanations by
means of basic formulas.

Followed by chapters with descriptions of each measurement, the initial chapters cover the issue of
relevant Standards and dielectric integrity in general, simplifying the area of testing high voltage devices,
such as power transformers.

Note: All pictured circuits are exemplary and only serve as better understanding. The
corresponding measuring circuits for the transformer´s customized vector group will be
appended in the last subchapter of each section (where it is useful).

1.1 Power Transformers Dresden

In Dresden, the beginning of transformer

manufacturing was in the twenties, when the founder
Franz Koch and Kurt Sterzel extended their
company “Koch & Sterzel AG”, by purchasing former
airport grounds and built a factory in Dresden-Kaditz.
They developed and produced high-voltage- and
instrument transformers, x-ray apparatus and radio
devices. After the Second World War, the remaining
parts of the “Koch & Sterzel AG” were re-established
to the “VEB Transformatoren- und Röntgenwerk
Founder of „Koch & Sterzel AG“:
(TuR)”, which developed to one of the most
Kurt August Sterzel (1876 – 1960), left
important export establishments of the former GDR. Franz Joseph Koch (1872 – 1941), right
During the time from 1949 to 1990 the number of
employees had been grown to 3500 at this location.

After the fall of the Berlin Wall, the company was

privatized and decommissioned, until 1992 Siemens
took over the ground partly. Today Siemens
employs about 280 co-workers and belongs to the
Power Transmission & Distribution Division. As a
sheer manufacturer, we are focusing on design and
construction as well as testing of Medium Power
Transformers (40 to 250 MVA), which are typical
applications for network- and generator step-up
transformers. They are also widely used for
industrial applications (e.g. arc furnace transformer).
Manufacturing ground (1992)
8 Electrical Testing Of Power Transformers

1.2 Transformer components

Core
On the right track right from the start: The iron core and
windings influence the subsequent efficiency of a
transformer. That’s why we design our transformers as
core types in which the wound and non-wound limbs of the
core are located at the same level and connected by
yokes. The choice of sheet metal also greatly effects the
quality of the core. At Siemens, we use only high-quality,
cold-rolled sheet metal. Depending on your requirements,
we may also opt to laser-treat the sheet metal. Using state-
of-the-art numerical controls, we cut the sheet metal
precisely to the millimeter for step-lap layering. This
ensures favorable flux distribution at joints, providing the
basis for low losses and minimized noise during no-load
operation.

Windings
Transformer windings must be able to withstand high
electrical and mechanical loads in daily operation. Our
windings made from copper wire definitely pass the test of
time. Their high mechanical stability ensures safe
operation, no matter what type of winding you need to
meet your power and voltage requirements. For the
winding we use transposed conductors, consisting of many
enamel-insulated, twisted section wires to minimize losses.

Voltage Variation
To adjust the ratio safely and easily to system conditions,
Siemens Power Transformers have tapped windings. In
this way the ratio of Power Transformers can be changed
gradually - either in no-load condition via off-load tap-
changers or under load by means of on-load tap-changers.

Tank
Another main component besides the core and the winding
is the tank of a Power Transformer: It accomodates the
active part and the oil filling. With the final assembly the
active part (the core, windings, pressed parts, tap-changer
and connecting cables) is installed in the tank and filled
with high-grade insulating oil under vacuum. The
transformer is then ready for testing.
 Introduction 9

1.3 Standards and Norms

Every transformer has to comply with national and international standards, which are published by
certain institutions. The primary Standards Organizations are the ANSI and the ISO. They specify the
respective tests that verify compliance with the requirements. It is important to know that in terms of the
electric field, the ANSI has delegated the writing and publication of standards to the IEEE, as similar as
the ISO handed over its authority to the IEC. All tests are referred to the IEC and IEEE standards. The
German Standards, settled in the VDE, are mentioned as well.

International Institute of Electric Association of electrical

Electrotechnical and Electronics engineering &
Institution
Commission Engineers electronics

Area served Worldwide Worldwide Germany

Headquarter Geneva New York Frankfurt on the Main

specific
IEC 60076-xx IEEE C57.12.xx VDE 0532-xx
Standard
Table 1.3.1: Applied Standards

Usually those standards are used directly to develop national standards. Which standards are valid in
which country can be looked up on the respective websites of IEC and IEEE. The Standards and
requirements based on transformers are settled in IEC 60076 and IEEE Std. C57.12 generally.

1.4 Types of tests

It will be distinguished between different types of tests.

Routine tests
Routine tests are tests required for every single transformer.
typical examples: Resistance measurement, Voltage ratio, Loss measurements

Type- or design1 tests

Those tests will be conducted on transformers which are representative of other transformers,
not covered by routine tests, but needed to withstand the same specified requirements.
“Representative” means identical in calculation and construction. Transformers with slight
deviations in their characteristics can be considered to be representative as well.
typical examples: Temperature rise test, Sound level measurement

Special- or other1 tests

Special- or other tests are additional tests different from type- or routine tests. Usually these
kinds of tests will be agreed upon the manufacturer and the customer.
typical examples: Measurement of zero-sequence impedance, Hot-Spot measurement

1
Terms used in the IEEE Standards
10 Electrical Testing Of Power Transformers

Note: Certain dielectric tests, such as lightning impulse tests for example, can be routine-, type-, or
special tests, depending on the respective standard, the insulation system and the maximum
system voltage.

1.5 Test sequences

There are often discussions about how and in which order tests will be conducted. Dielectric tests are
generally fixed in IEC and IEEE Standards. Existing Standard recommendations and regulations
referred to this issue are following below.

1.5.1 IEC Standard

IEC 60076-3, clause 7.3

“The dielectric tests shall, where applicable and not otherwise agreed upon, be performed in the
sequence as given below. This test sequence is in principle obligatory, but allows other
agreements between customer and manufacturer.”

Recommended test sequence (Dielectric tests only)

- Switching impulse test (SI) for the line terminals
- Lightning impulse test (LI) for the line terminals
- Lightning impulse test (LI) for the neutral
- Separate source AC withstand voltage test (Applied voltage test)
- Short duration induced AC withstand test (ACSD); including PD-measurement
- Long duration induced AC voltage test (ACLD); including PD-measurement

IEC 60076-1, clause 10.5

“In deciding the place of the no-load test in the complete test sequence, it should be borne in
mind that no-load measurements performed before impulse tests and/or temperature rise tests
are, in general, representative of the average loss level over long time in service.
Measurements after other tests sometimes show higher values caused by spitting between
laminate edges during impulse test, etc. Such measurements may be less representative of
losses in service”

This test sequence is a recommendation and not obligatory. It serves quality control purposes, verifying
partial discharge-free operation under operating conditions. There are different categories of windings,
which require different test sequences, specified in the referenced IEC Standard (see Table 1.5.1:
Separation according to IEC Standard).
 Introduction 11

Highest voltage
Winding
for equipment Tests
category
[kV]
LI SI ACLD ACSD AC
Single-phase Three-phase
(phase-to-earth) (phase-to-phase)
72,5 T NA NA NA R R
Uniform 72,5 < < 170 R NA S NA R R
insulation 170 < < 300 R R* R NA S* R
300 R R R NA S R
72,5 < < 170 R NA S R R R
Non-uniform
170 < < 300 R R* R S* S* R
insulation
300 R R R S S R
Table 1.5.1: Separation according to IEC Standard

R – routine test * If the ASCD test is specified, the SI test is not required.
S – special test
T – type test
NA – not applicable

Repeated dielectric tests

Transformers which already have been in service and have been refurbished, dielectric tests shall be
repeated at test levels reduced down to 80% of the original test value. Exceptions of this rule are long
duration AC induced tests (ACLD), which shall always be repeated at 100% test level, (according to IEC
Standard)

1.5.2 IEEE Standard

After IEEE Standards, power transformers are divided into two different classes due to system voltage
and transformer type insulation. This separation has an influence on the test sequence and its
classification (Routine-, Type- or Other test). For more details the IEEE Standards have to be consulted.

Class I Class II
Winding
category Non-graded Graded
/
isolation isolation
Highest system voltage
< 115 115 < 345 345
[kV]
Table 1.5.2: Separation according to IEEE Standard

Recommended test sequence (Dielectric tests only)

Switching impulse test
Lightning impulse test on line terminals and on transformer neutral
Applied voltage test
Induced voltage test
PD measurement
2
Insulation power factor test
Insulation resistance test

2
Term used in IEEE Standard for loss factor (tan )
12 Electrical Testing Of Power Transformers

If required, lightning impulse tests shall precede the AC voltage tests. Switching impulse tests, when
required, shall also precede the AC voltage tests. For class II power transformers, the final dielectric
tests shall be the induced voltage test. This test sequence is a recommendation and not obligatory
either.

IEEE Std. C57.12.90, clause 4.3

“To minimize potential damage to the transformer during testing , the resistance, polarity, phase
relation, ratio, no-load loss and excitation current, impedance, and load loss test (and
temperature-rise tests, when applicable) should precede dielectric tests. Using this sequence,
the beginning tests involve voltages and currents, which are usually reduced as compared to
rated values, thus tending to minimize damaging effects to the transformer.”

IEEE Std. C57.12.90, clause 10.1.5.1

“Lightning impulse voltage tests, when required, shall precede the low-frequency tests.
Switching impulse voltage tests, when required, shall also precede the low-frequency tests. For
class II power transformers, the final dielectric test to be performed shall be the induced voltage
test.”

1.5.3 Siemens recommendation

Based on our own experience and taking into consideration all IEC and IEEE regulations and
recommendations, we suggest the following test sequence:

1. Ratio, polarity and phase displacement

2. Resistance measurement
3. Current Transformer tests
4. Dielectric test:
Switching impulse test (if required)
Lightning impulse test
Separate source AC voltage test
Measurement of Insulation capacitances & loss factor (tan )
Insulation resistance measurement
Induced voltage test (including Partial discharge measurement)
5. Sound level measurement
6. No-load test (followed by sound level test, if specified)
7. Load loss and impedance
8. Zero-sequence impedance test (if specified)

Different from the common Standards (IEC and IEEE), we treat the Insulation resistance measurement
as a routine test, which means every single transformer has to comply with that test additionally.
 Introduction 13

1.6 Reasons for Transformer failure

It is expected that a transformer will experience and survive a number of short circuits during its service
life, but sooner or later one such event will cause slight winding movement, and the ability of the
transformer to survive short circuits in future will then be severely reduced. As the transformer ages, its
components deteriorate and the probability of a failure increases. There are several reasons for
deterioration, summarized in the following.

Paper insulation deterioration

The transformer paper insulation has a limited life, the extent of which depends on thermal, oxidation,
and moisture effects. When the paper insulation has reached the end of its expected life, the
mechanical strength of the paper is much reduced whereas its electrical strength is still satisfactory.
However the reliability of such a transformer is reduced and the mechanical forces of a short circuit or
other outside influence may cause a mechanical breakdown of the insulation leading to an electrical
failure of the transformer.

Core and winding movements

The core, windings, and turns can move due to short circuit forces, vibration, transport jolts, and
loosening of clamping pressure that have accrued during the life of the transformer.

Tap changer
Moving parts in the tap-changer and electrical joints in the tap-changer deteriorate over the years.

Auxiliary components
Auxiliary components such as bushings deteriorate, e.g. bushing gaskets leak causing ingress of
moisture and insulation deterioration.

Gaskets
Main gaskets and pipe work gaskets leak allowing ingress of moisture to the transformer and oil leaks
from the transformer

Rust
Rust causes deterioration of ferrous materials, especially in high humidity climates.
 Voltage stresses & dielectric integrity 15

2 Voltage stresses & dielectric integrity

2.1 General

During their operating time, transformers are exposed to several voltage stresses, which can appear
during normal as well as abnormal operation. In general, over-voltages are distinguished into three
categories:

Lightning over-voltages (A)

with a duration in the order of
microseconds

Switching over-voltages (B)

with a duration in the order of
a fraction of a second

Over-voltages (C)
with a duration in the
order of seconds to minutes

Figure 2.1.1: Types of over-voltages

Dielectric tests are intended to verify transformers integrity in case of over-voltages, as described above.
The different groups of over-voltages have also been considered in a test code. The actual test code for
a particular object, which is specified by a Standard, depends primarily on the size and rated voltages of
the object. Test voltages are primarily sinusoidal AC voltages. DC voltages are usually only relevant for
valve transformers, like HVDC transformers. Today´s test programs have their roots in a test code
based on short time AC-tests only, at voltages considerably higher than normal operating voltages. Test
objects either passed the test or broke down electrically. Later on it was found that other voltage
shapes, which includes transient impulse voltages (e.g. switching- or lightning over-voltages), are more
suitable to describe the stresses during abnormal conditions like lightning strikes. Through the evolution
of electronic diagnostic tools, more tests were added, those like the measurement of partial discharges,
which has become indispensable nowadays.

Standard Section/Clause Type of test

60076-3
IEC Power transformers – Part 3: “Insulation levels, dielectric tests and external
clearances in air”
C57.12.90
Standard Test Code for Liquid-Immersed Distribution, Power and Regulating Routine tests
Transformers, Clause 10: “Dielectric tests”
IEEE
C57.12.00
General Requirements for Liquid-Immersed Distribution, Power, and
Regulating Transformers
Table 2.1.1: Associated Standards
16 Electrical Testing Of Power Transformers

2.1.1 Lightning over-voltages

Lightning over-voltages are caused by weather-

related, atmospheric discharges. The power of the
over-voltage in the grid depends on the lightning
current and the impedance at the strike location. The
typical wave form of a lightning over-voltage can be
seen in Figure 2.1.2. This wave is unipolar (only one
polarity) and propagates along the line, starting at
the location of the voltage strike. It increases to a
peak within a few microseconds (wave front,
distinctive: high surge) and decays back to zero
within about a hundred microseconds (wave tail). By
line impedance and corona discharge, the
propagating wave becomes deformed and
dampened. To preserve from extreme surges
entering the object (e.g. a transformer), protection Figure 2.1.2: Lightning over-voltages
equipment such as surge arresters and spark gaps,
either in individually or in combination, are highly
recommended. The insertion of those protective FW = full wave
devices may conversely cause a steep voltage CW = in tail chopped wave
breakdown, which can be seen as a chopped FOW = in front chopped wave
lightning impulse at the transformer terminal.

2.1.2 Switching over-voltages

Switching operations cause transient phenomena,

what can lead to over-voltages. Shape, duration and
amplitude depending on the grid’s configuration and
point of switching operation related to sinus wave.

Figure 2.1.3 shows an example of an over-voltage,

during a switching operation in an overhead line.

a) configuration of network

b) equivalent diagram

c) oscillogram of switching impulse voltage

Figure 2.1.3: Switching over-voltages

 Voltage stresses & dielectric integrity 17

2.1.3 Temporary over-voltages

Temporary operating and non-operating over-voltages are caused by the following:

load rejection: over-voltage of 1,1 to 1,4 pu (several seconds)

single-phase short-circuit: over-voltage of 1,2 to 1,7 pu
(depending on neutral point configuration)
Ferro resonance (saw-tooth oscillations)
Ferranti-effect
other resonance oscillations

2.2 Transformer’s insulation verification

Figure 2.2.1 shows the relationship of the withstand voltage of conductor insulation to earth and the
duration of the over-voltage. Curve I and II represent the general behaviour of the major insulation to
earth in the area of impulse- and temporary overvoltages. Each area is specified in duration and
magnitude.

A: lightning impulse
overvoltage

B: switching impulse
overvoltage

C: temporary
overvoltage

Diagram I: oil insulation

Diagram II: air insulation

Figure 2.2.1: Insulation levels

For each area (A/B/C) tests are specified, verifying the impulse and AC withstand voltage of the
Transformer. The magnitude of the three test voltages depends on the highest voltage for equipment
and is defined in IEC and IEEE. For a comparison diagram I & II show the withstand voltage
characteristics of oil and air insulated transformers. It should be mentioned that a significant decrease in
withstand voltage in the area of switching over-voltages occurs. That is why a switching impulse test is
required here in every case, whereas an additional AC voltage test is not necessary. Even though a
transformer’s durability depends on more factors such as, insulation construction, oil purity, temperature,
partial discharge, etc. If the transformer exceeds or runs close to its major insulation level, the operating
life time may be decreased.
18 Electrical Testing Of Power Transformers

2.3 Test voltages

2.3.1 Alternating voltages

Regarding the standards, an alternating test voltage may either consist of a voltage electrically
energizing a circuit called a separate source, or a voltage across two terminals of a winding, needed to
conduct a test called an Induced voltage test. The duration of the alternating test voltage has been
traditionally 60 seconds, which is called “one-minute test” at low frequency (close to the rated
frequency). For voltages substantially above rated value during an induced voltage test, the core will
saturate unless frequency is increased in proportion.

For large high-voltage transformers, the short-time induced voltage test has often been replaced by a
combination of a long-time induced voltage test with measurements of partial discharges, together with
a switching impulse test. The switching impulse is then considered crucially for insulation integrity, while
the level of partial discharges is a qualitative measure of the insulation.

2.3.2 Impulse voltages

As seen in section: 2.1, there are actually two different types of transient impulse voltages: The
Lightning impulse with a “short” duration and the Switching impulse with a “long” duration. The Lightning
impulse is characterized by a steep voltage rise and a fast decay. On the other hand, the switching
impulse possesses a longer duration for wave front and –tail. The total duration of the switching impulse
is generally ten to twenty times longer than the lightning impulse.

For a lightning impulse, the length of the winding conductor is long compared to the propagation speed
of the impulse along the conductor. The wave characteristics of the winding have to be considered. For
a switching impulse the rate of change in voltage is low enough to permit a model where wave
characteristics can be ignored and transformer behaviour is similar to that under normal AC voltage and
power frequency conditions.

2.4 Electrical test fields

A modern high-voltage test system is mandatory to conduct all common dielectric tests, verifying
transformer’s insulation and integrity. It consists of several parts, which are developed and produced by
only a handful of manufactures. Due to our vicinity to “HighVolt Prüftechnik Dresden“, specialized in
high-voltage test devices, their expertise is very useful in testing of power transformers.

At our grounds we are running two test fields. Because of their spatial dimensions and the installed
equipment, they are not convenient for all high-voltage configurations, which depend on size, rated
power and voltage of the transformer.
 Voltage stresses & dielectric integrity 19

Test field 1 ( 425 kV) Test field 2 ( 145 kV)

Test hall Test hall

1. Control room
1. Control room (1)

2. Control room 1. Control room (2)

Test field 1 is used as the routine test field for Test field 2 is feasible for transformer with
almost every measurement. Because of its less rated voltages. Because of small
larger dimensions (distances to walls etc.), it distances to the walls, lightning impulse
is allowed to conduct measurements with tests are limited to values where no flash
higher voltages than test field 2. overs will occur.
20 Electrical Testing Of Power Transformers

Digital measuring systems

Digital measuring systems are used for measuring and displaying impulse voltage and current waves as
well as AC transients, evaluating impulse parameters for peak voltage, current and time. By means of
an analyzing system, test results are stored to a database for evaluation and comparison from different
tests. Moreover, in combination with HV components, such as voltage dividers, shunts, chopping gaps
or impulse generators, we can control the circuit for easier and faster adjusting for certain test
specifications. The measuring systems are integrated in our control rooms (pictured above).

Power and instrument transformer controls

AC Resonant test system (HighVolt)
Impulse voltage test system “MIAS” (HighVolt)

To cover all electric measurements, further devices like AC/DC Peak voltmeter, instrument transformer
controls etc. are installed. In fact there are also portable measuring devices (e.g. winding resistance
meters, voltage ratio meter, etc.), which are not placed in control rooms.
 Measurement of winding resistance 21

3 Measurement of winding resistance

3.1 Purpose and Standards

To know a transformer’s winding resistance is essential. The value is needed for further calculations
and serves a number of functions like:

Checking the internal winding connections

Load loss calculations (short-circuit measurement)
Indirect method, establishing winding temperature and temperature rise within a
winding (see section: 14 Temperature rise measurement)
Verifying electric continuity within in a winding

Standard Section/Clause Type of test

60076-1
IEC
Clause 10.2: “Measurement of winding resistance”

C57.12.90 Routine test

IEEE
Clause 5: “Resistance Measurements”

VDE 0532-76-1
Table 3.1.1: Associated Standards

3.2 General

Winding resistances are always measured in the unit Ohms [ ] and is defined as a direct current
resistance.

Temperature
It is important to know that the resistivity of the conductor material strongly depends on the temperature.
The general statement is with increasing temperature, material’s resistivity also rises. For temperatures
within the normal operating range the following relation between resistance and temperature is
sufficiently accurate:

= resistance at temperature
+ = resistance at temperature
=
+ , = temperature in °C
C = constant (function of material type)
22 Electrical Testing Of Power Transformers

Material IEC [1] IEEE [51] A value of resistance without the respective temperature is
not useful, otherwise the resistance will be considered at
Copper C = 235 C = 234,5 room temperature. The indirect method of establishing the
winding temperature can be achieved by repeating the
Aluminium C = 225 C = 225
resistance measurement at a further random temperature.
Table 3.2.1: Material constants

Winding characteristics at resistance measurement

According to the law of induction, a winding has not only a resistance, but also a large inductance.
When a voltage is applied across two terminals (e.g. of a winding), the relationship between voltage and
current can be described as:

u =
= · + · i =
L =

Readings first may be taken after the current

has reached a steady value. Establishing a
magnetic field in the winding and switching
phenomena influence the resistance value
readings (see Figure 3.2.1).

Figure 3.2.1: current properties

3.3 Measuring circuit

There are several methods measuring the winding resistance. The basic methods are the “Kelvin-
Bridge” and the “Voltmeter-ammeter Method”. According to IEEE Standard an accuracy of ± 0,5% is
required for resistance measurement and ± 1°C for temperature measurements.

Our measuring devices are based on the “Voltmeter-ammeter Method” (see Figure 3.3.1), which is
carried out using DC current. The principle is grounded on Ohm’s law, calculating the resistance by the
recorded values. The measurement itself is performed using digital meters (see Figure 3.3.2), but can
also be performed with conventional analog-meters, which requires simultaneous readings and is not
common anymore.
 Measurement of winding resistance 23

=
=
=

Figure 3.3.1: Principle of voltmeter-ammeter

measuring method
=

=
=
=

Figure 3.3.2: Measuring circuit (schematic)

3.4 Measuring procedure

If the transformer is equipped with an off-circuit ratio adjuster or on-load tap changer, the preliminary
test of the core-and-coil assembly of this measurement is carried out in all switching and tap changing
positions. Whereas in most cases the final measurement of the winding resistance is carried out in the
principal tapping position (rated position) and in the tap position with the minimum and maximum
number of windings. The winding resistances are always measured between the phases. Depending on
the measuring system used, up to four or three windings of the test object connected in series may be
measured at the same time. The test is made up by the transformer test system of Tettex Instruments
type 2291 or Raytech resistance meter WR50R. Depending on the winding type, phase-to-earth or
phase-to-phase resistances are measured.

As noted, the ohmic winding resistances are highly depending on the temperature. For this reason the
measurement is effected with a maximum of 10% of the rated current, in order to avoid an unnecessary
heat-up of the winding. Additionally the winding temperature is measured, which is implemented by
means of a digital thermometer at the bottom of the tank, half way up of the tank’s height and on the
tank cover. The information obtained are averaged and the results taken to be the mean oil temperature
and to be equivalent to the mean winding temperature. An alternative method of test is based on the
direct withdrawal of an oil sample and the measurement of its temperature by means of a precision
mercury thermometer.
24 Electrical Testing Of Power Transformers

Condition
DC-current value, used for measurement:
Maximum value – 10% of the rated winding current, (IEEE Standard permits 15%)
Minimum value – 1,2 times the magnetizing current crest value

Uncertainty in measurements
Using analog instruments the uncertainty is typically ±0,5% (accuracy class 0,2 for instruments and 0,1
for standard resistors used for current measurement). For digital instruments a typical uncertainty is
±0,15%.

Interpretation of the measured values

As pictured in Table 3.4.1, the resistance might have two different values, depends on the inner circuitry
of the transformer (star- or delta connection). This has to be considered when calculating the phase
resistance , which is the required value.

Star connection Delta connection

Phase
= · = ·
resistance

Line resistance = · = ·

Table 3.4.1: Difference between Star- and Delta - connection

 Measurement of winding resistance 25

3.5 Appendix: Measure arrangement of customized vector group

Test circuit

Measuring instruments
TETTEX 2291 Raytech WR50R

High current micro-ohmmeter Resistance measurement of high inductivities

Measuring Measuring
2m …20k 0,5 …1k 1m …0,1M
range: range:
± (0,1% of rdg. + ± (0,5% of rdg +.
Accuracy: ± (0,05% of rdg. +0,05 of scale) Accuracy:
2 digits) 3 digits)
26 Electrical Testing Of Power Transformers
 Voltage ratio & verification of vector group or phase displacement 27

4 Voltage ratio & verification of vector group or phase

displacement

4.1 Purpose and Standards

The voltage ratio and the phase displacement are principally of interest when it comes up to parallel
operation with two or more transformers.

Standard Section/Clause Type of test

60076-1
Clause 6: “Connection and voltage displacement symbols for three-phase
IEC transformers”
Clause 10.3: “Measurement of voltage ratio and check of phase
displacement”
Routine test
C57.12.90
IEEE Clause 6: “Polarity and phase relation test”
Clause 7: “Ratio tests”

VDE 0532-76-1
Table 4.1.1: Associated Standards

4.2 General

Voltage ratio
The voltage ratio is considered to be the ratio between the no-load voltages of HV- and LV-side. Vector
groups and their characteristics are only defined for poly-phase transformers. They can be connected in
star-, delta- or zigzag-connection, depending on the requirement. The phase displacement between the
windings can only be influenced within 30° - steps from 0° to 330°, which depends on the connection
method. Transformers parallel operation needs a similar no-load ratio and the same vector group.
Otherwise circulating currents would occur between the parallel transformers, what have to be avoided.

The term “voltage ratio” is defined as:

=
=
= =
=
=
28 Electrical Testing Of Power Transformers

Vector groups
The most common vector groups according to IEC 60076-1 (Figure 4.2.1).

Figure 4.2.1: Common connections (IEC 60076-1)

 Voltage ratio & verification of vector group or phase displacement 29

The vector diagram of the high voltage winding

is actually placed on a clock face, so that the tip
of vector I (1U) is at 12 o’clock. When the
vector diagram for the low voltage winding is
placed on top with the same phase orientation,
the direction of vector I (2U) identifies the clock
number of the vector group. Thereby one hour
is equal to 30° phase displacement.
For a zig-zag connection the winding half
closest to the terminals determines the terminal
markings. If the winding half closest to the
terminals is on limb V, the terminal is also
called 2V.
Figure 4.2.2: Illustration of clock-number

For three-phase transformers the phase angle

of the intermediate and low voltage winding is
referenced to the high voltage winding for the
vector group. If there is a neutral, it is defined
by the letter “N” (high voltage winding) or “n”
(low voltage winding); e.g. Dyn11.

IEEE Standard
Vector groups according to IEEE Std.
C57.12.90 are essentially the same as the IEC
vector groups. The designation of the terminals
is H1, H2 and H3 for the high voltage side and
X1, X2 and X3 for the low voltage side.
Figure 4.2.3: Examples of clock numbers

4.3 Voltage ratio measurement

Measuring Circuit
In our test field the digital measuring
instruments are comparing the voltage ratios of
two windings (Voltmeter-method). The principle
measuring circuit is shown in Figure 4.3.1,
which is the same for all transformer vector
groups. If the low voltage side can be measured
without using voltage transformers, it is better to
Figure 4.3.1: Principle Voltmeter-method
feed the high-voltage side, since the voltmeter measuring circuit
is the only load on the transformer to be tested.
30 Electrical Testing Of Power Transformers

Procedure
As a condition, the distribution of magnetic flux in the core has to be the same. This means that only
windings, winding segments and winding combinations which have the same magnetic flux, can be
compared with one another. Normally a reduced voltage is used for the supply. The measurement is
carried out with a minimum of three voltage values (increasing in steps of 10%). The average value
represents the correct measured value and the readings must lie within a range of 1%. Digital
instruments with a sufficiently high resolution are required. They have practically no load to the
transformer because of their high input resistance. A stable voltage source is mandatory to obtain an
accurate measured result.

4.4 Vector group verification

Measuring circuit
Determining the vector group is only valid for three
phase transformers. This can be applied by
connecting a terminal from the low voltage side to a
terminal on the high voltage side (see bold lined
path in Figure 4.4.1). The terminals 1V and 2V are
then galvanically connected, having the same
potential. If a three-phase supply is connected to the
high voltage winding, potential differences appear Figure 4.4.1: Principal measuring circuit for
verification of the vector group
between the open terminals and are used to
determine the vector group.

The measured values of the individual voltages are

entered on a phasor diagram, from which the correct
polarity and vector group can be determined. In this
example terminal 2V has the same voltage as
terminal 1V. The two points are then identical on the
phasor diagram. The points 2U and 2W can be
determined by measuring voltages 1U-2U, 1W-2U,
and 1U-2W, 1W-2W, from which the transformer
vector group can be derived.

Procedure
The measuring instruments, which are the same for
Figure 4.4.2: Graphic display of voltages
voltage ratio, energize the transformer with a
symmetric 3-phase voltage supply and conducts the
procedure (described above) automatically, creating
a chart as a printout, by which the respective vector
group is determined (example in
Figure 4.4.3). Generally both measurement (voltage
ratio & voltage group) can be conducted by the
“Tettex 2793a” or “Raytech TR-Spy Mark III”.
Voltage ratio & verification of vector group or phase displacement 31

Figure 4.4.3: Example for measuring results of

transformer´s vector group (Yy0)
32 Electrical Testing Of Power Transformers

4.5 Appendix: Measure arrangement of customized vector group

Test circuit

Measuring instruments

TETTEX 2793a Raytech TR-Spy Mark III

voltage ratio, vector group (3-phase) voltage ratio, vector group (3-phase)

Measuring range: 0.8:1…13,000:1 Measuring range: 0.8:1…13,000:1

Accuracy: 10V 40V 100V Accuracy: 10V 40V 100V

0.8…100 ±0.05% ±0.05% ±0.05% 0.8:1…2000:1 ±0.15% ±0.08% ±0.08%

101…1000 ±0.1% ±0.05% ±0.05% 2001:1…4001:1 ±0.22% ±0.1% ±0.1%

1001…1500 ±0.12% ±0.05% ±0.05% 4001:1…13000:1 ±0.5% ±0.3% ±0.3%

1501…2000 ±0.14% ±0.08% ±0.08%

 Short-circuit voltage & load loss 33

5 Short-circuit voltage & load loss

5.1 Purpose and Standards

Knowing the exact load loss and short-circuit voltage is interesting, not only for the operating costs It is
important to know in case of large power transformers (non-permissible temperature rise, see section:
14. Temperature rise measurement). Moreover it gives an indication about the eddy-losses caused by
leakage fluxes in mechanical parts like the tank wall, by comparing the calculated and measured values

For transformers with tapped windings the short-circuit voltage has to be measured in the principal tap
position and the two extreme tap position additionally.

Standard Section/Clause Type of test

60076-1
Clause 10.1: “General requirement for routine, type and special tests”
IEC Clause 10.4: “Measurement of short-circuit impedance and load loss”

60076-8
Clause 10: “Guide to the measurement of losses in power transformers” Routine test

C57.12.90
IEEE
Clause 9: “Load losses and impedance voltage”

VDE 0532-76-1
Table 5.1.1: Associated Standards

5.2 General

The generally applicable short-circuit equivalent schematic can be seen in Figure 5.2.1.

The definition of short-circuit voltage and load loss:

An AC voltage is connected to one winding system of a transformer with the opposite winding
system short-circuited. When rated current flows in the short-circuited winding system, the
appearing voltage between the terminals is the short-circuit voltage. The absorbed active power
corresponds to the transformer load loss.
34 Electrical Testing Of Power Transformers

a)
= 1
= 2
= 1
= 2
=
=
=

b)
Because of the short-circuit, the iron
losses and the main reactance don’t
have to be considered.

c)
= +
= +
=
=

Figure 5.2.1: Scheme for short-circuit

A component of the transformer’s no-load losses ( + ) will also be measured. It can be neglected,
since the short-circuit voltage is minimal compared to the rated voltage (exceptions are starting
transformers with an air gap, reactor transformers, etc.). The manufacturer ensures all data about short-
circuit voltage and load loss.

Short-circuit voltage
According to Figure 5.2.1 c) the short-circuit voltage can be determined as follows:

=
( )
=
=
=
=

Figure 5.2.2: Short-circuit

vector diagram
 Short-circuit voltage & load loss 35

= + = · %

describes the relative short-circuit voltage. It is given as a percent of the rated voltage. With this
value it can be determined which current flows in case of a short-circuit during operation.

=
= ·
=

For power transformers a small value of relative short-circuit is requested unlike furnace arc
transformers which have higher relative short-circuit values.

Load loss
The total losses occurring within the transformer when rated current and frequency is applied are
represented in the load loss. It is made up of the ohmic losses of the windings and internal connections,
as well as the stray losses (eddy current losses) caused by leakage fields in the windings and the
mechanical parts. The load loss is referenced to the winding temperature (75°C according to IEC and
85°C according to IEEE).

=
= + =
=
36 Electrical Testing Of Power Transformers

5.3 Measuring circuit

Figure 5.3.2 and Figure 5.3.1 showing the measuring circuit for load loss measurements on single-
phase and three-phase transformers. Usually the short-circuit is applied on the low voltage side of the
transformer, because it is more practicable to adapt the test equipment in the test field.

In case of testing three-phase transformers there are principally two different wattmeter configurations
possible: the two- and the three-wattmeter method. The three-wattmeter method is preferred, because it
more accurate and allows to determine the power per phase (for investigation tests).

Figure 5.3.2: Measuring circuit for single-phase Figure 5.3.1: Three-wattmeter method for
transformers three-phase transformers

5.4 Measuring procedure

Winding resistance and winding temperature measurements must be carried out before the actual load
loss measurement. If there are built-in current transformers, they must be shorted during the test to
avoid saturation of their iron cores and prevent over-voltages at their secondary terminals. The bushing
taps must be earthed. If the transformer is equipped with an on-load tap-changer, the first loss
measurement is carried out at the principal tap and subsequently at the highest and lowest taps.

During the test, the current is adjusted steadily upwards (from zero to full measuring current) in order to
avoid inrush currents. Its DC component can lead to instrument transformer errors, which cannot be
corrected (pre-magnetization of the current transformers).The duration of the test should be as short as
possible avoiding any significant heating in the windings. Therefore the measuring time at rated current
should be about 30 seconds (rule of thumb).

Moreover the measuring current should be as close as possible to the rated current, although IEC
specifies that the current should not be below 50% of rated current. To confirm the measured results a
second measurement with approximately 10% lower current, is recommended. The values should be
agreed by extrapolating the two points.
 Short-circuit voltage & load loss 37

Procedure for three-winding transformers

For three-winding transformers (e.g. additional stabilizing winding) the short-circuit and load loss cannot
be determined by a single short-circuit test. They have to be calculated within three different short-
circuit tests, allocating the results to every single winding. The following short-circuit tests must be
carried out. Losses and impedances have to be adjusted to rated current and reference temperature
(75°C or 85°C).

Measuring arrangements Legend:

Test Winding 1 – High-voltage winding
Winding Winding
between Short- 2 – Intermediate-voltage winding
supplied open
windings circuit
3 – Low-voltage winding
1-2 1 2 3

1-3 1 3 2

2-3 2 3 1

Table 5.4.1

Calculation of the equivalent short-circuit voltage per winding

All three results should be referred to a common apparent power (e.g. rated power of the high-voltage
winding). The short-circuit voltage is recalculated linearly with the reference power.

Checking the results:

+
=
= +
+ = +
=
= +
+
=

Calculation of the load loss per winding

Once more a common reference power must be selected (e.g. rated apparent power of the high-voltage
winding). The load loss per winding can be recalculated to correspond to the operating conditions. The
load losses per winding are not always the same as the actual losses during operation because of the
differing configurations of leakage flux during operating conditions with three windings.

+
=

+
=

+
=
 38 Electrical Testing Of Power Transformers

5.5 Appendix: Measure arrangement of customized vector group

Test circuit

Measuring instruments

Yokogawa WT3000 Rack

Digital Precision Power Analyzer contains Power Analyzer, Votlage-
and Current transducer controls
0,01% of
Accuracy: basic readings
0,02% of
power readings
Bandwidths: DC and 0,1Hz-1MHz
Direct inputs: up to 30 A/1000 V

Tettex TMS-PT-581
Voltage transducer
Output voltage: 100 V
0,5, 1, 2, 5, 10, 20, 50,
Measuring range:
100, 200, 400, 800 kV
0,04 % (5-800 kV)
Overall accuracy: 0,1% (1-2 kV)
0,15% (0,2-0,5 kV)

Tettex TMS-CT-582
Current transducer
Test field 1 Test field 2
Output current: 1A 1A
1, 2, 10, 20, 40, 100, 1, 2, 5, 10, 20, 50, 100,
Measuring range:
500, 1000, 2000, 4000A 500, 1000, 2000 A
0,005% (40-4000 A) 0,005% (20-2000 A)
0,01% (10-20 A) 0,01% (5-10 A)
Overall accuracy:
0,03% (4 A) 0,03% (2 A)
0,05% (2 A) 0,05% (1 A)
 Measuring the no-load loss & no-load current 39

6 Measuring the no-load loss & no-load current

6.1 Purpose and Standards

The no-load loss is a very important value, representing a considerable amount of energy during the
life-time of the transformer. The no-load losses are caused by the no-load current which is necessary to
excite the transformers core. The actual loss figure has to be guaranteed and depends on the sheet
material of the core and its manufacturing process.

Standard Section/Clause Type of test

60076-1
Clause 10.1: “General requirement for routine, type and special tests”
IEC Clause 10.5: “Measurement of no-load loss and current”

60076-8
Clause 10: “Guide to the measurement of losses in power transformers” Routine test

C57.12.90
IEEE
Clause 8: “No-load losses and excitation current”

VDE 0532-76-1
Table 6.1.1: Associated Standards

6.2 General

Unloaded Transformer
An energized but not loaded transformer can be seen as an iron core reactor. The equivalent diagram
for this case is shown in Figure 6.2.1. The secondary winding can be neglected, because there is no
current flow.

=
=
=
=
=
E =
Figure 6.2.1: No-load equivalent diagram
40 Electrical Testing Of Power Transformers

The magnetizing characteristic of the iron core is pictured by the

hysteresis curve (Figure 6.2.2). The area inside this dynamic loop
is a measure of the energy required to change the flux for on cycle
(one period). The smaller the area, the less energy is required.

B =
H =
=
=

Figure 6.2.2: Magnetic loop =

No-load current
The no-load current express the sum of the current
needed for the magnetization of the core and a
capacitive current reflecting the capacitances
between the windings. Because of the non-linear
magnetization characteristic of the iron core, the
corresponding no-load current is naturally distorted
when a sinusoidal voltage is applied (see Figure
6.2.3). The measured no-load current is as RMS
value. It is generally expressed in percentage of
the rated current and is about 1 to 5 % for small
power transformers and 0,1 to 0,3% for large
power transformers. For three-phase transformers
Figure 6.2.3: No-load current
the value is the average of the three windings.

No-load loss
The no-load loss of a transformer consists of several components:

a) Iron losses = +

( = hysteresis losses; = eddy losses)

b) Dielectric losses = · · ( )

c) Winding losses = ·

+
 Measuring the no-load loss & no-load current 41

Usually dielectric and winding losses are negligible, because for power transformer they are several
orders of magnitude smaller. That means that the no-load losses are equal to the iron losses ,
which consist of hysteresis losses (depending on sheet metal of the core) and eddy losses (depending
on sheet thickness). There’re exceptions like starting transformers with an air gap.

Relationship between no-load loss and voltage distortion

The voltage distortion is caused by the non-sinusoidal no-load current of the transformer under test,
which causes a voltage drop in the internal impedance of the supply (generator/matching transformer).
Hysteresis losses depending on the average value of applied voltage. If this voltage is properly set,
is not affected by voltage distortion. The eddy losses on the other hand are a function of the
square of the RMS voltage, the same as the losses in a DC resistance. However, the RMS value is
affected by the voltage distortion, which includes the eddy losses.

Those voltage distortions do not occur during operation because the impedance of the supply system is
much smaller than the transformers main inductance . Comparing the losses of different transformers,
the no-load losses are guaranteed using a sinusoidal supply. Losses measured with distorted voltage
must be recalculated.

Relationship between no-load loss and iron temperature

No-load loss temperature dependence can be seen only at relatively high temperature variations. That
is why transformers should be tested at the ambient temperature in the test field (about 20°C, according
to IEC/IEEE). Corrections of the measured no-load losses are not required, if the following conditions
are hold:

a) The average oil temperature is within +-10°C of the reference temperature

b) The difference between the top and the bottom oil temperatures does not exceed 5°C

If the above conditions cannot be met, the measured no-load loss can be referred to the above
reference temperature of 20°C using the following empirical formula:

= ·( +( )· )

= 20°
=

=
=
42 Electrical Testing Of Power Transformers

6.3 Measuring circuit

Single-phase transformer
The position of the circuit-earthing may create an error on the measured no-load loss figure when
measuring no-load loss on single-phase transformers. The source of the error is a capacitive earth
current caused by parasitic cable capacitances and generator and matching transformer windings.
Normally the three capacitive currents , and are approximately of the same magnitude and
shifted in phase by 120°. In this case the resulting current is zero. If the supply voltages do not have
the same potentials to earth, there will be a residual current , which can affect the no-load current,
depending on the location of the earth termination. This source of error can be eliminated by placing the
earth ahead of the current transformers (seen from the supply side, see Figure 6.3.1).

Figure 6.3.1: Influence of capacitive currents

=
=
=
AT =

Three-phase transformer
When testing three-phase transformers, the voltages induced in the individual windings (phase voltages)
must be measured to determine the form factor of the voltage. The form factor for star-connected
windings is different, because it cannot include harmonics divisible by three. The individual elements of
the measuring circuit should fulfil the following requirements:

the generator must maintain a constant frequency with changing load

the short-circuit impedance of the generator and the matching transformer should be
as small as possible
 Measuring the no-load loss & no-load current 43

Vector groups Remarks Measuring circuit

Voltage transformers and

wattmeters are connected in
Yd star
Voltmeters are connected
YNd between two phases and
measure the line-to-line
Dd voltages, which are also the
phase voltages for a delta-
connected transformer under
test.

Voltage transformers and

wattmeters are connected in
star
Voltmeters measure the line-
to-line voltages (allowed,
Dyn because third order harmonic
currents can flow in the high-
voltage winding due to the
delta connection voltage
distortions on the low-voltage
winding are avoided)

Yyn
Voltmeters must measure
YNyn the phase voltages

Voltage transformers are

Yy delta-connected on the
secondary
YNy
Voltmeters measure the
line-to-line voltages

Table 6.3.1: Measuring circuits

44 Electrical Testing Of Power Transformers

6.4 Measuring procedure

Because there are lower voltages required, the no-load measurement is carried out on the low-voltage
side of the transformer under test. Built-in current transformers must be shorted during the test and
condenser bushing taps must be earthed. Before carrying out the no-load loss test, the voltage ratio
must be checked. For oil transformers the bushings and Buchholz relay must be vented and the oil level
of the transformer must be checked. Before the actual loss measurements take place the transformer
must be excited by 1,1 to 1,15 times rated voltage. The over-excitation reduces the effects of
remanence caused by direct current excitation during resistance measurement or from the Switching
impulse test. Until the measured figures are not steady, the actual loss measurement cannot start.

Typically, measurements are taken starting at 110 and decreasing to 100, 90 and 80% of rated voltage.
When testing large three-phase units the three watt-meters may show differing figures. It is even
possible for one wattmeter reading to be negative. The magnetic asymmetry of the iron core causes
asymmetrical no-load currents. Depending on the flux density in the core, the phase displacement
between current and voltage in one phase is greater than 90°, which will be seen as a negative power
in one wattmeter. After all, the actual input power is the sum of the readings of three watt-meters. The
measuring instrument “Yokogawa WT3000” does this automatically, since it has more than three input
channels.
 Measuring the no-load loss & no-load current 45

6.5 Appendix: Measure arrangement of customized vector group

Test circuit

Measuring instruments

Yokogawa WT3000 Rack

Digital Precision Power Analyzer contains Power Analyzer, Votlage-
and Current transducer controls
0,01% of
Accuracy: basic readings
0,02% of
power readings
Bandwidths: DC and 0,1Hz-1MHz
Direct inputs: up to 30 A/1000 V

Tettex TMS-PT-581
Voltage transducer
Output voltage: 100 V
0,5, 1, 2, 5, 10, 20, 50,
Measuring range:
100, 200, 400, 800 kV
0,04 % (5-800 kV)
Overall accuracy: 0,1% (1-2 kV)
0,15% (0,2-0,5 kV)

Tettex TMS-CT-582
Current transducer
Test field 1 Test field 2
Output current: 2, 4, 8 A 2, 4, 8 A
1, 2, 10, 20, 40, 100, 1, 2, 5, 10, 20, 50, 100,
Measuring range:
500, 1000, 2000, 4000A 500, 1000, 2000 A
0,005% (40-4000 A) 0,005% (20-2000 A)
0,01% (10-20 A) 0,01% (5-10 A)
Overall accuracy:
0,03% (4 A) 0,03% (2 A)
0,05% (2 A) 0,05% (1 A)
46 Electrical Testing Of Power Transformers
 Separate source AC withstand voltage test (Applied voltage test) 47

7 Separate source AC withstand voltage test (Applied

voltage test)

7.1 Purpose and Standards

The test is designed to check the main insulation of the transformer. This main insulation is generally
understood as the insulation system between two windings (major insulation), but also the insulation
between the winding and earth (end insulation) and all connections to earth.

Standard Section/Clause Type of test

60076-3
IEC
Clause 11: “Separate source AC withstand voltage test”

C57.12.90
Clause 10.6: “Applied voltage tests”
IEEE Routine test
C57.12.00
Clause 5.10: “Insulation levels”

VDE 0532-76-3
Tabelle 7.1.1: Associated Standards

7.2 General

In this test a separate AC source is applied to the transformer (“applied-voltage test”). For transformers
with uniform insulation every part of the winding is exposed to the test voltage , which relates to the
insulation level of the transformer between the windings and earth. In case of transformers with graded
insulation between their windings, the test voltage must be adjusted to the lowest insulation
requirements (usually the winding end or neutral).
48 Electrical Testing Of Power Transformers

7.3 Measuring circuit

The principle test circuit is shown in Figure 7.3.1. The equivalent diagram of the transformer under test
is an R-C parallel circuit consisting of the effective capacitance , representing the capacitance
between the winding under test and earth (including the bushing capacitance) or between the other
windings, and the resistor R, representing the insulation (Figure 7.3.2). The used principle measuring
circuit in test laboratory is shown in Figure 7.3.3.

G =
=
=
=
= Figure 7.3.1: Test connection for
a three-phase transformer
=
K =
=
=
=
HM =
VT =
Figure 7.3.2: Capacitances of the transformer
ransformer

CT =

Figure 7.3.3: Measuring circuit

 Separate source AC withstand voltage test (Applied voltage test) 49

7.4 Measuring procedure

Before the test can be carried out, it must be ensured that the whole winding insulation is calculated
properly for the test voltage. Care must be taken for transformers with graded insulation (non-uniform),
where the voltage level must be adjusted to the insulation of the transformer neutral. In this case each
phase will be tested individually, see test connections in Induced voltage tests for non-uniform
transformers

The Buchholz-relay and all bushings must be degassed before the test; any surge arresters and
bushings arcing horns must be removed. Electrodes, such as spheres or similar protection shields
mounted on the outer terminal of the bushing are permitted because they increase the external electric
strength, whereas the applied voltage test verifies the internal electric strength. Generally the test must
be carried out at ambient temperature (at least 10°C). During the test, the voltage is applied directly to
the bushings of the transformer windings by a single-phase source. All other windings and the tank
must be earthed (see Figure 7.3.1).

Voltage, frequency and duration

The test voltage is a RMS value [kV] and
Highest voltage for Rated separate source
represent the withstand voltage of the
equipment AC withstand voltage
insulation (given in the IEC 60076-3, clause 7).
[kV] [kV]
It depends on the respective highest voltage
… …
for equipment . The table below shows an
extract for some values. The applied test 52 95
voltage should be raised rapidly from 25% or 60 115
less to the test voltage and remain 72,5 140
100 185
constant for 60 seconds. At the end of the test
the voltage should be reduced rapidly (in 123 230
about 5 seconds). The test frequency should 145 275
be the rated frequency. The measured values 170 325
may be read directly by means of a capacitive … …
voltage divider and an AC peak voltmeter. Table 7.4.1: Rated withstand voltages for
This method is characterized by a high transformer windings
accuracy, achieving precise measurements.

Acceptance criteria
The test is successful if the test voltage does not collapse or if there is no indication of other faults
(smoke, bubbles, sudden current increase).
50 Electrical Testing Of Power Transformers

7.5 Appendix: Measure arrangement of customized vector group

Test circuit 1 ( 230 kV) Measuring instruments

FPEO 2400/ 600 (HV-Transformer)

MU 14 (AC/DC Peak Voltmeter)

Measuring
; 2; RMS -value
quantities:
Accuracy: max. ±0.5%

Test circuit 2 ( 375 kV) Measuring instruments

Reactor

Rated Power: 1750 kVar

Inductance: 235 H

Resonant AC test system

 Switching impulse test & Lightning impulse test 51

8 Switching impulse test & Lightning impulse test

8.1 Purpose and Standards

The purpose of the test is to verify the integrity of the insulation for transient voltages, which are caused
by atmospheric phenomena (lightning strikes), switching operations or network flaws.

Standard Section/Clause Type of test

60060-1
High-voltage test techniques – Part 1: “General definitions and test
requirements”
60060-2 Routine tests
High-voltage test techniques – Part 2: “Measuring systems” LI for
60060-3 > 72,5 kV
High voltage test techniques – Part 3: “Definitions and requirements for on- SI for
IEC
site tests” > 300 kV
60076-3
Power Transformers – Part 3: “Insulation levels, dielectric tests and external for the rest:
clearances in air” Special tests

60076-4
“Guide to lightning impulse and switching impulse testing of power
transformers and reactors”
Routine tests
C57.12.90 LI for
Clause 10: “Dielectric tests” > 115 kV
SI for
IEEE
C57.98 > 345 kV
“Guide to impulse testing techniques, interpretation of oscillograms and
failure detection criteria” for the rest:
Other tests

0532-76-3
VDE acc. to IEC
0532-76-4
Table 8.1.1: Associated Standards

8.2 General

Because impulse tests always mean high dielectric stresses for the insulation system, it is possible that
hidden damages occur, which will be identified during the voltage insulation tests. That is why, if not
otherwise specified, the impulse test (SI and LI) precede the insulation tests (applied voltage test and
induced voltage test in combination with PD measurements).

Impulse shapes
Switching impulses as well as lightning impulses have standardized wave shapes. The following table
shows the essential differences.
52 Electrical Testing Of Power Transformers

Switching impulse (SI) Lightning impulse (LI)

Front: : 100 s Front: : 1,2 s (± 30%) = 1,67 · T

Tail: : 1000 s Tail: : 50 s (± 20%)
: 200 s1 Polarity: negative2

Polarity: negative2

one characteristic wave form Three different wave shapes

a) full wave (pictured above)
b) tail chopped wave
c) front chopped wave (only required by
IEEE Std.)

chopped waves
peak should be 10% higher full wave
value
chopping after =2–6 s
overshoot not higher than 30% of test
amplitude

Test sequence (acc. IEC) Test sequence (acc. IEC)

impulse at about 60% of the specified one reduced level full impulse (calibration)
insulation level one full level full impulse
three impulses at the specified insulation one reduced level chopped impulses
level (if requested)
two full level chopped impulses
Consider: It may force the core into saturation, (if requested)
because switching impulse represents a time two full level full impulses
integral of the applied voltage to the transformer.
(Chopped wave should have an amplitude of
110% of full wave amplitude)
Table 8.2.1: Characteristics of Switching- and Lightning impulse

1
time above 90% of peak voltage
2
for oil insulated transformers
 Switching impulse test & Lightning impulse test 53

To prevent flashovers on the air-side of the transformer bushings, the polarity of the impulse voltages is
generally negative. Because of their unusual aperiodic waveform transient voltages cannot be
generated perfectly. Therefore the definition of rise time needs additional specification: A straight line
from the bottom to the peak, passing through the voltage values 0,3 and 0,9 times the crest value, gives
the rise time (see figures in Table 8.2.1; SI and LI). The rise time is then 1,67 times the time between
these two voltage values.

8.3 Measuring circuit

The following figure shows the impulse test circuit with all its components. It is illustrated in three ways
to make out the different parts of the circuit:

bold lines: main circuit (discharge circuit)

fine lines: measuring circuit
dashed lines: circuit to apply chopped wave

= =
= =
= =
SFS = , =
TFS = =
= SG =
= CG =
= MK =
= IPV =
= IMS =
54 Electrical Testing Of Power Transformers

The configuration of the test connection

should be selected so that it corresponds
with conditions under operation. So there
are lots of test configurations possible. Only
single-phase impulses are considered
occurring most of all in the grid. Therefore
non-tested winding terminals have to be
earthed. Figure 8.3.1 shows the most
common test connections. The impulse test
sequence is applied to each of the line
terminal. Other line terminals (three-phase
transformers) shall be earthed directly or
through a low impedance.

According to IEC Standard, transformers

with tapped windings are tested with the two
extreme- and the principal tapping position
(one tapping for each of the three individual
phases of a three-phase transformer),
unless nothing else has been agreed. If the
tapping range is 5% or less only the principal
has to be tested.

=
R = Figure 8.3.1: Common test connections
O =

8.4 Measuring procedure

Preparation
Before conducting the impulse tests (Lightning- and Switching impulse), the transformer will be checked
for the following points:

standing time
confirming measurement of voltage ratio, polarity and winding resistance
oil level and quality
vent the Buchholz relay and porcelain bushings
short-circuit and earth any built-in transformers
earth the capacitive bushing taps
check on-load or off-load tap changer positions
 Switching impulse test & Lightning impulse test 55

Steps
The actual impulse tests:

calibrating and verifying the impulse voltage waveform

applying the impulse voltage to the transformer under test (see test sequence)
verifying that the transformer under test had withstood the stresses without damage
(by comparing oscillographic recordings)

Acceptance criteria
The switching and lightning impulse tests are successfully if the recordings of the oscillographics do not
show any changes (close similarity in wave form; between calibrated voltage and applied test voltage).
If the waveforms are deviated from each other, the reasons have to be determined.

8.5 Test equipment for high-voltage tests

In this section the most important components of our high-voltage test field are introduced. They are
mandatory to simulate Switching- and Lighting impulse voltages.

Impulse voltage generator

An Impulse voltage generator is an electrical apparatus which produces very short high-voltage surges.
They are used to test the strength of electric power equipment against lightning and switching surges.
The main components are capacitors, which are separated in stages, connected in series by spark
gaps. By charging the stages with positive or negative DC voltage, the charging voltages totalize to the
required impulse voltage. Additionally the waveform of the impulse (front/tail) can be adjusted by the
contained resistors. The applicable impulse voltages of Impulse Generator in Test field 2 are limited,
because of the close distance to the wall.

Test field 1 Test field 2

IG 300/2000G IG 120/1600G
Number of stages: n = 10 Number of stages: n = 10 (extended)
Capacitance per stage: = 1500 nF Capacitance per stage: = 1500 nF
Charging voltage per stage: = 190 kV Charging voltage per stage: = 190 kV
Total charging voltage: = 1900 kV Total charging voltage: = 1900 kV
Total charging energy: 300 kJ Total cha ging energy: 150 kJ
Lightning impulse voltage: Lightning impulse voltage:
= 1675 kV = 650 kV
(permitted) (permitted)
Switching impulse voltage: Switching impulse voltage:
= 1100 kV = 350 kV
(permitted) (permitted)
56 Electrical Testing Of Power Transformers

Multiple chopping gap

To simulate impulse voltage with chopped waveforms, the applied voltage has to be reduced to zero
within a certain time ( s). Therefore a multiple chopping gap is necessary to force voltage collapses
within the applied impulse voltage. The chopping can be realized at the front or at the tail of the impulse
wave.

AFC 890/1800 AFC 200/1200

Number of stages: n=9 Number of stages: n=6
Gap range/stage 5…120 mm Gap range/stage 5…120mm
Operating range/stage: n · (60…200)kV Operating range/st ge: n · (60…200)kV
Chopping voltage: 540…1800kV Chopping voltage: 360…1200kV
Chopping time: 2…6 s (±50ns) Chopping time: 2…6 s (±50ns)
Capacitance/stage: 9 nF Capacitance/stage: 8 nF

Capacitive voltage divider

Voltage dividers are useful for measuring high voltages, which cannot be handled by the measuring
system. Similar to an potential divider with ohmic resistors, two capacitances are connected in series,
where one serves as measuring capacitance with a considerably larger capacitances referred to .
By means with the common formula for voltage dividers, the voltage can be reduced to a harmless
dimension for the corresponding measuring equipment.

=
+
 Switching impulse test & Lightning impulse test 57

SMCF 1000/1600 SMCF 1333/1200

Rated voltage: 2000kV Rated voltage: 1200 kV
Number of divider Number of divider
4 (400-800-1600-2000 kV) 5 (200-400-600-800-1200 kV)
stages: stages:
Capacitance of Capacitance of
1 nF 1,333 nF
voltage divider: voltage divider:

8.6 Appendix: Measure arrangement of customized vector group

Test circuit
58 Electrical Testing Of Power Transformers

8.7 Transferred overvoltage

Lightning and Switching surges can be transferred from one voltage level to another through
transformer couplings. A distribution system, which may not be directly exposed to the overvoltages of
atmospheric origin, but is connected to a utility system through a transformer of high turns ratio will be
exposed to overvoltages on the secondary side due to overvoltages on the primary windings. The
resulting stresses on the distribution system may exceed the basic insulation levels (BIL). The
inspection of transferred overvoltages is only justified for large generator step-up transformers, which
have a large voltage ratio and high-voltage network transformers with a low-voltage tertiary winding.

Standard Section/Clause Type of test

60076-3 Annex B
IEC “Overvoltage transferred from high-voltage winding to a low
voltage winding” Special test
VDE 0532-76-3
Table 8.7.1: Associated Standards

The issue of transferred overvoltage is also treated from a system viewpoint in IEC 60071-2 Annex A.

Procedure
When the low-voltage winding cannot be subjected to lightning over-voltages from the low-voltage
system, this winding may be impulse tested with surges transferred from the high-voltage winding. This
method is also be preferred when the design is such that an impulse directly applied to the low-voltage
winding could result in unrealistic stressing of higher voltage windings, particularly when there is a large
tapping winding physically neighbouring to the low-voltage winding.

With the transferred surge method, the tests on the low-voltage winding are carried out by applying the
impulses, at lower impulse levels (approximately 25% of original impulse level) to the neighbouring
higher voltage winding. The line terminals of the low-voltage winding are connected to earth through
resistances of such value that the amplitude of transferred impulse voltage between line terminals and
earth, or between different line terminals or across a phase winding, should be as high as possible but
not exceeding the rated impulse withstand voltage. The magnitude of the applied impulses shall not
exceed the impulse level of the winding to which the impulses are applied.

This method is also applicable, as a special

procedure, with low voltages (usually 100V).
Therefore a mini impulse generator with
integrated chopping gap and adjustable
impedances is needed.

Mini impulse generator

 Switching impulse test & Lightning impulse test 59

Test circuit

The impulse voltage will be applied on each High-voltage terminal Low-voltage terminal
high-voltage terminal (except neutral). The (impulse applied) (transferred overvoltage)
transferred overvoltage will be measured on 2U
each low-voltage terminal separately, which 1U 2V
means that nine measurements have to be 2W
taken in total. The acceptance criteria are the 2U
same as for the standard impulse voltages.
1V 2V
2W
2U
1W 2V
2W
 Partial Discharge Measurement 61

9 Partial Discharge Measurement

9.1 Purpose and Standards

A partial discharge measurement is another important way to verify the transformers insulation system,
ensuring that no harmful PD sources exist. The purpose of a PD-measurement is to detect and localize
areas where partial discharges are about to occur, typically cavities or conducting particles inside the
insulating material. Those areas are exposed to higher dielectric stresses, which in the long run can be
very harmful to transformers insulation (e.g. electric breakdown). So usually PD-measurements are
carried out in conjunction with dielectric tests (Induced voltage test).

Standard Section/Clause Type of test

60076-3
Annex A “Application guide for partial discharge measurements during AC Routine test
IEC withstand voltage test on transformers according to 12.2, 12.3 and 12.4” for
60270 > 72,5 kV
“Partial discharge measurements”
Routine test
C57.12.90
IEEE for
Clause 10.8/10.9: “Induced voltage tests”
115 kV

VDE 0532-76-3 Routine test

Table 9.1.1: Associated Standards

Because the partial discharge measurement and the induced voltage test are carried out
simultaneously, the corresponding measuring circuit of the customized vector group is consolidated in
the next section (“Induced voltage test”, see 10.5 Appendix: Measure arrangement of customized vector
group).
62 Electrical Testing Of Power Transformers

9.2 General

A partial discharge is a partial voltage breakdown

within a series of insulating elements between two
electrodes (of different potentials). It can be
interpreted as an electric charge from one position
to another. For very fast changes the individual
insulation links between two line terminals can be
regarded as a number of series connected
capacitors, where is a potential source for
discharges (Figure 9.2.1 and Figure 9.2.3). The
excessive stress in these “weak” spots can result
from design- or insulating material flaws or
deviations in the manufacturing process. A damage
to the insulation caused by previous tests is also
possible.

BU =
HV =
NT =
, , = Figure 9.2.1: Schematic representation of
( ) a part of the transformer
insulation
=
= ( )

Figure 9.2.2 and Figure

9.2.3 are showing the RLC-
network of a transformer as
an example, illustrating the
complexity of possible spots
where partial discharges
may occur.

HV =
LV =
RW =
NT =

Figure 9.2.2: Winding and Figure 9.2.3: Equivalent circuit

insulation system
 Partial Discharge Measurement 63

9.3 Measuring circuit

If the two line terminals are connected together

via an external capacitors (Figure 9.3.1), a
charge within the series connected capacitances
(inside the insulation, Figure 9.2.1) will also be
reflected in the charge of external capacitor .
Under the assumption that << , potential
discharges only occur inside the transformers
insulation system (represented by ). Then the
charge movements can be detected as
circulating current impulses i(t). All PD measuring
methods are based on this principle. Therefore
the basic equivalent circuit for PD measurements
is presented Figure 9.3.1.
The preferred method is to transform the signal to Figure 9.3.1: Equivalent circuit for PD
an apparent electric charge, measured in measurement
picocoulombs (pC) which requires an integration
of the PD current impulse i(t).

G =
Q(t) = Q( ) + ( ) dt Z =
=
=
Q(t) = =
Q( ) = i(t) =
~ ,~ =
Worth mentioning is the different approach of PD q =
measuring by the IEEE, transforming the signal =
to a Radio Interference Voltage (RIV), generally
measured in microvolts ( V). Most PD systems
available on the market perform a “quasi-
integration” of the PD current impulses in the
frequency domain using wide-band filter.

PD measurements on transformers can only be

conducted on their bushings. For power
transformers the measuring impedance is generally
connected between the bushing measuring-tap and
earth (parallel with ).

= (= )
=
=
=
= Figure 9.3.2: Calibration circuit for PD measurement
= (bushings with capacitive tap)
64 Electrical Testing Of Power Transformers

For bushings without a capacitive tap, external

coupling capacitances ([ ] picturedi in 10.5) must
be connected in parallel with the bushing (compare
Figure 9.3.2 and Figure 9.3.3). To avoid external
discharge (corona) in the PD measuring circuit all
transformer bushing tops should be covered with
shielding electrodes (including earthed bushings).

= (= )
=
= Figure 9.3.3: Calibration circuit for
= PD measurement (bushings
without capacitive tap)
=

9.4 Measuring procedure

Although there are some differences between the IEC and IEEE Standard, the PD measurement is
basically integrated in the Induced voltage tests (see section: 10 Induced voltage tests). Considering
the IEC Standard, PD measurement is mandatory for long- and short duration induced voltage tests
(ACLD and ACSD). The time sequences are shown in Figure 10.4.1 and Figure 10.4.2, according to
IEC. PD activity will be checked on all bushings where the system voltage is higher than > 72,5kV.
The measurements (in pC) are carried out at each voltage level, except the enhanced level (part C) and
should be documented.

Acceptance criteria
The PD test is considered successful if no partial discharges activity greater than the specified apparent
charge amplitude in pC is detected at any bushing, and if there is no rising trend in the apparent charge
amplitude during the long duration test. The recommended acceptable values of apparent charge,
according to IEC Standards are:

300 pC at 130%
500 pC at 150%
the level of continuous PD activity does not exceed 100 pC at 1,1

If the PD activity cannot comply with the requirements, the type and the location (external or internal
insulation system) of the PD source has to be detected. In
Table 9.4.1 typical PD sources are shown, which is just for a better visualization. Those PD pattern can
be regarded as a fingerprint of the partial discharge activity of a specific defect in the transformers
insulation system. For getting more details about the subject of PD source investigation, professional
literature has to be consulted.
 Partial Discharge Measurement 65

Schematic drawing of the

PD source Typical PD pattern
PD source

Conducting material
(tip electrode) with
direct contact to
metallic electrode

Conducting material
without any contact to
metallic electrode

Non-conducting
material (cavity) with
direct contact to
metallic electrode

Non-conducting
material (cavity)
without any contact to
metallic electrode

Non-conducting
material (cavity)
without any contact to
metallic electrode with
changing surface due
to the partial
discharge

Table 9.4.1: Typical PD sources in the transformer insulation

66 Electrical Testing Of Power Transformers
 Induced voltage tests 67

10 Induced voltage tests

10.1 Purpose and Standards

The induced voltage test verifies the AC withstand strength of each line terminal and its connected
windings to earth and to other windings, but also between phases and along the winding (turn-to-turn
insulation).

Standard Section/Clause Type of test

60076-3
IEC Clause 7.3: “Dielectric tests”
Clause 12: “Induced AC voltage tests (ACSD, ACLD)”
C57.12.90
Clause 10.7/10.8: “Induced voltage tests” Routine test
IEEE
C57.12.00
Clause 5.10: “Insulation levels”

VDE 0532-76-3
Table 10.1.1: Associated Standards

10.2 General

The testvoltages of this measurement are nearly the same as for the Applied Voltage Test (see
section 7). The difference is the measuring procedure and the possibility to conduct a Partial Discharge
Measurement (during the whole test duration), which can indicate insulation faults before an electric
breakdown occur. There are essential differences between the IEC and IEEE understanding of induced
voltage tests. Generally IEC distinguishes between:

Short- and Long duration AC tests (ACSD/ACLD)

Uniform- and non-uniform (graded) insulation
Highest voltage levels

Note: Partial discharge measurements (PD; see section: 9 Partial Discharge Measurement) has to be
conducted simultaneously, except for transformers with < 72,5kV.

IEEE distinguishes only between class I ( < 115kV) and class II ( 115kV) transformers. For class
II transformers a long duration test in combination with a PD measurement is always required. Table
10.2.1 gives an overview about the separation, according to IEC and IEEE Standards.
68 Electrical Testing Of Power Transformers

Three-phase Single-phase
Highest voltage
ACLD ACSD ACLD ACSD
Category of winding for equipment
[kV] Phase- Phase- Phase-
to-earth to-phase to-earth
test test test
According IEC
< 72,5 X X
72,5 < < 170 PD PD
Uniform
insulation 170 < < 300 PD PD

> 300 PD PD
72,5 < < 170 PD PD PD

Non-uniform insulation 170 < < 300 PD PD

> 300 PD PD
According IEEE
Class I
< 115 X X
Non-graded insulation
Class I X X X
Graded insulation

Class II 115 PD PD
Table 10.2.1: Induced voltage tests for three- and single-phase transformers

X : Routine test without PD measurement

PD : Routine test with PD measurement

10.3 Measuring circuit

The principle of the induced voltage test is

shown in Figure 10.3.1. In reality the test-circuit
is more complicated than shown in this figure,
because the leakage reactance of the
transformer and the generator have been Figure 10.2.1: Simplified test circuit
omitted for better understanding. The load
impedance of the generator is a current
resonance circuit (R-C-L). The generator =
current reaches its minimum value when the = ( )
test frequency f is equal to the resonant TT =
frequency . Because the test voltage is C =
normally higher than twice the rated voltage, R =
the test frequency has to be adjusted (double =
test voltage means double frequency) to avoid =
over-excitations in the iron core (see formula; k
= variable depending on transformer design).
To keep the generator current as low as
= · · =
possible the variable reactor can influence the ·
resonance circuit and its resonant frequency.
 Induced voltage tests 69

10.4 Measuring procedure

Preparations
The actual test should not be carried out until the impregnation of the windings has been completed.
The necessarily “standing-time” depends on the rated voltage of the transformer (e.g. about 3 days for a
220 kV transformer).

The Buchholz-relay and all bushings must be degassed before the test; any surge arresters and
bushings arcing horns must be removed. Unlike in the applied voltage test: Electrodes, such as spheres
or similar protection shields mounted on the outer terminal of the bushing are highly recommended,
especially for high voltage tests. These electrodes are mandatory for partial discharge measurements in
combination with the induced voltage test.

Test voltage
The test voltage depends on the highest voltage for equipment and can be looked up in the
Standards. Generally these are the same values used in the separate source AC withstand voltage test
(except ACLD) and is about twice the rated voltage.

Test duration at full test voltage

The test time should not exceed 60 seconds (at full test voltage ), but not less than 15 seconds
(independent of the test frequency). If the test frequency is adjusted and is higher than twice the rated
frequency, the test time will be decreased after the following formula:

= (C)
= · =
=

Test sequence (according IEC)

Short duration AC withstand

voltage test (ACSD)

= 1,3 · ( )
= 1,3 · ( )
3
= 60076 3, 7
C =

Figure 10.4.1: Voltage level sequence (ACSD)

70 Electrical Testing Of Power Transformers

Long-duration AC withstand
voltage test (ACLD)

= 1,5 · ( )
3

= 1,7 · ( )
3
C =
30 min < 300
=
60 min > 300
Figure 10.4.2: Voltage level sequence (ACLD)

Test sequence (according IEEE)

Long-duration AC withstand
voltage test (ACLD)

ACLD tests always in combination with

PD measurement, Table 10.2.1.

A = time for PD measurement

B = 7200 cycles (ca. 2min for 50Hz)
C = 60 minutes

Figure 10.4.3: Test sequence for class II transformer

The voltage shall first be raised to the 1h level and held for a minimum of 1 min or until a stable partial
discharge level is obtained to verify that there are no indications of partial discharges (voltage level of A
and C are the same). The level of partial discharges shall be recorded just before raising the voltage to
the enhancement level (B), which is held for 7200 cycles. Then the voltage shall be reduced directly to
the 1h level (C). Partial discharge measurements shall be made at 5 min intervals.

Acceptance criteria
The test is successful if the test voltage does not collapse. In combination with PD measurements the
criteria for success of the test are further explained in section: 9 Partial Discharge Measurement.
 Induced voltage tests 71

Transformers with non-uniform insulated HV windings

Induced AC voltage tests for non-uniform insulated HV windings are treated separately, settled in IEC
60076-3, clause 12.3. On single-phase transformers, only a phase-to-earth test is required. For three-
phase transformers, two sets of ACSD tests have to be carried out:

phase-to-earth test at rated withstand voltages + PD measurement

(performed as (three) single-phase tests, see Table 10.4.1)

phase-to-phase test with earthed neutral at rated voltage

(performed as three-phase test)

In case of ACLD tests, single-phase connection (phase-to-phase) are possible as well, but are not
obligatory. Recommended test connections which avoid excessive overvoltage between line terminals
are shown below. Other separate windings shall generally be earthed at the neutral if they are star-
connected and at one of the terminals if they are delta-connected. Depending on the low-voltage
winding, three different generator connections are possible.

I II III This test connection can be used when the

neutral is designed to withstand at least one-
third of the voltage

Test connection I is only possible for

transformers with 5 limb core or shell-type
transformers

Test connections for ACLD test, if performed

phase-to-phase in single-phase connection

only possible when the transformer has a no

wound limb for the magnetic return path for
the flux in the tested limb (five limb core)

Table 10.4.1: Single-phase test connections for Induced voltage AC test

72 Electrical Testing Of Power Transformers

10.5 Appendix: Measure arrangement of customized vector group

Test circuit 1 (with capacitive tap) Test circuit 2 (without capacitive tap)

Measuring instruments

Lemke LDD-5 Omicron CPL 542 & MPP600 WMC 0,8/150

measuring impedance (CPL 542) rated voltage: 150 kV

+ battery pack (MPP600) rated capacitance: 0,8 C

Omicron CAL542 WCF 1,25/200

Partial Discharge Detector Charge calibrator Coupling capacitances

detectable 0,1pC-25 C necessary for bushings without
complies with all parts of IEC 60270
apparent charge: (extendable) capacitive taps
measuring range: 5, 20, 100, 500 pC fed a defined impulse charge into rated voltage: 200 kV
the measuring circuit. The partial
Accuracy: ± 10 % of readings discharge detection system is then rated capacitance: 1,25 C
calibrated to this value

Note: The power analyser “Yokogowa WT3000” is used additionally, observing applied voltage and current
 Insulation resistance measurement 73

11 Insulation resistance measurement

11.1 Purpose and Standards

The insulation resistance test (also commonly known as a “Megger test”) is a spot insulation test which
uses an applied DC voltage and is performed to determine the insulation resistance from a winding to
earth or between individual windings. This value is important to know, when assessing the transformers
insulation condition. The measured insulation resistance is to be understood as a reference value used
for comparison in measurements made at a later stage. In our test field the Insulation resistance
measurement is treated as a routine test, different from the applied Standards.

Standard Section/Clause Type of test

Routine test
> 72,5 kV
60076-1
IEC
Power transformers – Part 1: “General”
Special test
< 72,5 kV

Routine test
for class II transformers
C57.12.90
IEEE
Clause 10.11 “Insulation resistance tests”
Other test
for class I transformers
Table 11.1.1: Associated Standards

11.2 General

Although there are more sophisticated measuring methods, like the FDS measurement (section: 20
FDS measurement for moisture estimation), the advantage of this method is the very easy procedure.
The value the insulation resistance is generally measured in megohms (M ), which depends strongly
on transformers design, temperature, dryness and cleanliness (especially bushings). Considering those
factors can help to explain and find existing insulation uncertainties. It has to be noted that insulation
measurement varies with the applied voltage. To compare measurements with each other, the applied
voltage should always be the
74 Electrical Testing Of Power Transformers

11.3 Measuring circuit

The measuring circuit is quite similar to that of

the applied voltage test. All terminals of one
winding system (e.g. HV side) must be
connected with each other. The megohm -
meter used for this test has an integrated DC
power supply, which generates a voltage of
5kV. In case of arc furnace transformers the
low-voltage side is usually not designed for
5kV. The voltage has to be adjusted to avoid
flashovers in this case.
Figure 11.3.1: Principal measuring circuit
(MO = megohm-meter)

11.4 Measuring procedure

In accordance with the IEC Standard, the duration for the measurement is 1 minute. The readings take
place after 15 seconds ( ) and 60 seconds ( ). The ratio of / is a criteria for the insulation
condition and should be in the range of 1,3 – 3.

The IEEE defines the measurement with a third reading of the resistance after 600 seconds ( ).
With that value the “Polarisation Index (PI)” can be determined, which serves as an evaluation of the
insulation as well. PI values < 1 stand for unsatisfactory condition, whereas PI values > 2 indicate a
really good condition.

Polarisation Index: =

The number of measurements increases with the presents of a tertiary winding, because more
combinations of connections are possible. The temperature of the transformer should be close to the
reference temperature of 20°C.
 Insulation resistance measurement 75

11.5 Appendix: Measure arrangement of customized vector group

Test circuit

Measuring instruments

Gossen-Metrawatt Metriso5000 Megger MIT520/2

Digital High-Voltage Insulation Tester Insulation Resistance Tester

5kV 2,5kV 1kV 500V

Accuracy: ± (10% of readings + 8 digits) Accuracy: ±5% 1T 500 G 200 G 100G

±20% 10T 5T 2T 1T

Measuring range: 0,4 M - 1 T Measuring range: 10 k …15 T

100V, 250V, 0,5kV, 1kV, 1,5kV,

Test voltages: Test voltages: 250V, 500V, 1kV, 2,5kV, 5kV,
2kV, 2,5kV, 5kV
 Test on tap-changers & auxiliary equipment 77

12 Test on tap-changers & auxiliary equipment

12.1 Purpose and Standards

Although all on-load tap-changers are tested individually during the manufacturing process, it is
necessary to check the correct operation in combination with the fully assembled transformer, verifying
a proper performance without failure during operation.

Standard Section/Clause Type of test

60076-1
Clause 10.8: “Test on on-load tap-changers”
IEC Routine tests
60076-3
Clause 10: “Insulation of auxiliary wiring”

Routine tests
for class II
C57.12.00
transformers
Table 19 and clause 8.2.3: “Dielectric test for low voltage control wiring,
IEEE
associated control equipment and current transformer secondary circuits, on
Other tests
Class II power transformers”
for class I
transformers

VDE 0532-76-1 acc. to IEC

Table 12.1.1: Associated Standards

12.2 Tests on Tap-changers

Specified by the IEC Standards, the following tests for tap-changers have to be conducted:

Un-energized transformer Energized transformer

1 complete cycle of operation with 85% of rated 1 complete cycle of the tapping range at no-load at rated
voltage voltage and frequency

10 switching operations across the range of ± 2 stages on

8 complete cycles of operation at rated voltage
either side of the principal tapping. The switching
(one cycle is defined as going from one end to the
operations are carried out, if possible, at rated current in
other and back again of the whole tapping range)
the winding with taps with one winding short-circuited

Table 12.2.1: Test procedure (acc. IEC)

The test circuits with rated voltage is similar to that for measurement of no-load loss (see section: 6
Measuring the no-load loss & no-load current) and the test circuit for the operation test with rated
current is similar to that for measurement of load loss (see section: 5 Short-circuit voltage & load loss).
The tap-changer should be observed and checked for any abnormal activities during the test.
78 Electrical Testing Of Power Transformers

12.3 Tests on AuxiIiary equipment

In addition to the on-load tap-changer test, the auxiliary equipment have to be tested. Therefore all
wirings for auxiliary power and control circuits are subjected to an AC separate source test (2kV RMS to
earth for 1 minute, acc. IEC). The IEEE Standard specifies an AC applied voltage test at 1,5kV RMS for
the auxiliary equipment, excluding current transformer circuits, which are tested at 2,5kV RMS. If some
of the auxiliary devices (e.g. motors and other apparatus) have a lower test voltage than specified for
the wiring alone, they have to be disconnected before the test. This test will be carried out with the high-
voltage testing unit “Schleich GLP1-g”. The test is successful when voltage breakdowns occur.

IEC IEEE

Auxiliary equipment: Current transformers:

2kV RMS to earth
1,5kV RMS to earth 2,5kV RMS to earth
for 1 minute
(AC applied voltage test) for current transformers

Schleich GLP1-g
manual adjustable AC high-voltage source
arc detection

Measuring range: 0…6000V

Accuracy: ± (2% of rdg + 0,1mA)
 Check of built-in current transformers 79

13 Check of built-in current transformers

13.1 Purpose and Standards

Current transformers, together with voltage transformers (VT), are known as instrument transformers.
They are used for measuring alternating currents and voltages, operating on the same principle of
common transformers. Because measuring and recording instruments are really sensitive, they are not
allowed to be exposed to high currents. Therefor the current transformer produces a reduced current
accurately proportional to the current in the circuit, which can be conveniently connected to measuring
system. Additionally current transformers isolate the measuring instruments and protect them from
over-voltage in the monitored circuit.

Standard Section/Clause Type of test

60044-1
IEC Routine test
Instrument transformers - Part 1: Current transformers

C57.13
“…current and inductively coupled voltage transformers used in
IEEE Routine test
measurement of electricity and the control of equipment associated with the
generation, transmission, and distribution of alternating current.”

Table 13.1.1: Associated Standards

13.2 General

Like any other transformer, a current transformer

has a primary winding, a magnetic core, and a
secondary winding. The alternating current flowing
in the primary produces an alternating magnetic
field in the core, which then induces an alternating
current in the secondary winding circuit. An
essential objective of current transformer design is
to ensure that the primary and secondary circuits
are efficiently coupled, so that the secondary
current bears an accurate relationship to the
primary current. The most common design of CT
consists of a length of wire wrapped many times
around a silicon steel ring passed 'around' the
Figure 13.2.1: Principle of a current
circuit being measured. The CT's primary circuit
transformer
therefore consists of a single 'turn' of conductor,
with a secondary of many tens or hundreds of
turns.
80 Electrical Testing Of Power Transformers

13.3 Measuring procedure

The portable Omicron CT-Analyzer performs excitation, ratio and polarity tests on current transformers
using the voltage-comparison method. Current transformers can be tested in their equipment
configuration, such as being mounted in transformers, oil circuit breakers or switchgear. It is necessary
for the equipment to be totally isolated from the electrical system prior to testing.

Polarity Test
Connected to the high voltage winding, the method is used will be conducted by switching on a DC
current source briefly. The polarity can be checked on a polarized voltmeter, connected to secondary
winding (low voltage side). The voltmeter must be calibrated to be sure that the deflection is correct.
Usually a series resistor is applied in the circuit to protect the instrument. If the polarity of the
transformer is correct, the deflection must be in the same direction when the DC current is switched on.

Saturation Test
Measuring the excitation voltage and excitation current that results as the voltage applied to the CT
under test increased. As the CT under test begins to saturate, a large increase in current will be
detected for a small increase in voltage. With this test the magnetization curve and magnetization
characteristics will be confirmed. Saturation test shall be conducted before ratio test and after polarity
test, since residual magnetism left in the core due to DC test (polarity, resistance), which leads
additional error in ratio test.

Ratio Test
The ratio test is performed by comparing a voltage applied to the secondary winding to the resulting
voltage produced on the primary winding. For example, if 1 volt per turn is applied to the secondary
winding, the voltage present on the primary winding would be 1 volt. More specifically, if 120V were
applied to the secondary of a 600/5 current transformer (120:1 ratio), 1V should be present on the
primary winding.

13.4 Measuring circuit

Omicron CT-Analyzer
Primary Secondary Testing of current transformer
(Transformer (CT Conductor)
Conductor)
Automatic evaluation according to IEC & IEEE
Measuring of polarity, ratio, excitation/saturation,
winding resistance, burden impedance
Accuracy: 0,02% of readings
 Temperature rise measurement 81

14 Temperature rise measurement

14.1 Purpose and Standards

The rate of aging or decomposition depends strongly on the temperature the insulation material is
exposed to. At a temperature of 100°C is already a noticeable degree of decomposition. According to
Montsinger´s law, on which different loading guides are based, the aging rate doubles with each
temperature increase of about 6 K in the range from 80°C to 140°C. As a result, the Standards have not
only established the highest permissible temperature values for the windings and oil, but also rules for
permissible overloads, including their influence on life expectancy. Defined by IEC, the term
temperature rise is the difference between the temperature of the average winding temperature and the
temperature of the external cooling medium. It confirms the guaranteed temperature rises of oil and
windings. It is also useful for detecting possible hot-spots in- and outside the winding. Measuring the
average- and top oil temperature rise, as well as knowing the winding-oil gradient, may be really
important in terms of future upgrades (overload considerations).

Standard Section/Clause Type of test

60076-2
Clause 5: “Test of temperature rise”
IEC
60354
“Loading guide for oil immersed power transformers”
Type- or
C57.12.90
Clause 11: “Temperature rise” design test
IEEE
C57.12.00
Clause 5.11: “Temperature rise and loading conditions”
VDE 0532-76-2
Table 14.1.1: Associated Standard

14.2 General

The temperature inside the transformer has to comply with certain temperature limits, specified by IEC
and IEEE Standards. Those limits are necessary to guarantee a long and reliable service of the
transformer. To compare between different transformer concepts and suppliers properly, the
permissible temperature rise above the cooling medium (winding-oil gradient), as well as the ambient
temperature, have to be regarded.

Cooling methods and Identification symbols (according IEC)

Generally the cooling methods are specified in the IEC Standards as a test code (see Table 14.2.1).
The cooling system which is applied, influences the temperature rise but may be also the sound
emission. In the end it depends on the area of application.
82 Electrical Testing Of Power Transformers

First letter Second letter Third letter Fourth Letter

Internal cooling medium Circulation mechanism for External cooling Circulation mechanism
(in contact with windings) int. cooling medium medium for ext. cooling medium

mineral oil or
synthetic insulating natural convection flow through
O N A air N natural convection
liquid with fire point cooling equipment and in windings
300°C

forced circulation through cooling forced convection

insulating liquid with
K fire point 300°C
F equipment and natural convention W water F fans (air)
flow in the windings pumps (water)

forced circulation through cooling

insulating liquid with
equipment, directed from the
L no measurable fire D
cooling equipment into at least the
point
main windings

Table 14.2.1: Identification symbols (acc. IEC)

A common example:

ONAN/ONAF
The transformer has a set of fans which may be put into service if required at high loading. The
insulating liquid circulation is by thermosiphon-effect only, in both cases.

Measuring principle
During the temperature rise test only , , and can be measured directly. The winding
temperature will be determined indirectly by calculation (resistance measurement). Measuring the
winding temperature, using sensors directly attached to the winding are carried out with fiber optics and
useful, detecting possible hot-spots and predict transformer´s operational life span. A simplified
temperature distribution model is shown in Figure 14.2.1.
 Temperature rise measurement 83

Figure 14.2.1: Simplified illustration of temperature distribution

= =
( ) =
= =
= = ( )
( ) C =
= 1, 2 =
( ) g =

14.3 Measuring procedure

Short-circuit method
There are several methods to determine
the temperature rise. Because of practical
reasons the short-circuit method has been
established for the determination of
steady-state temperature rises. As seen in
Figure 14.3.1 the principle of short-circuit
connection is applied. During this test the
transformer is not subjected to rated
voltage and rated current simultaneously,
Figure 14.3.1: Principal test connection (three-phase)
but to the calculated total losses, with matching transformer
previously obtained by two separate
determinations of losses, namely load loss
at reference temperature and no-load loss.
84 Electrical Testing Of Power Transformers

The supply voltage is about the same as the short-circuit voltage, which means there are no losses in
the iron core practically. But to obtain the correct top-oil temperature rise, the total losses are required.
That’s why the no-load losses must be simulated in the windings by injecting a current slightly higher
than rated current.

=
= ·
=
+ =
=
=

The power required for the test is therefore:

=
= · · %
=

Preparations
The transformer has to be placed that the cooling system (inlet and outlet) will not be affected by any
objects in the test field. Besides protective devices like the Buchholz relay must be equipped. Any
indication from these devices during the test shall be noted and investigated. All cold resistance
measurements have to be performed before the test, in exactly the same measuring configuration as
the one used for the warm resistance measurement.

Procedure
The purpose of the short circuit-method is to establish several values:

top oil temperature ( )

in a steady-state of total loss injection
average oil temperature rise ( )

average winding temperature rise ( )

top oil temperature rise ( ) at rated current
hot-spot winding temperature rise ( )

Additionally to the short-circuit method, a winding resistance measurement will be carried out before the
temperature rise test (reference values: and ) and after disconnection of power supply. To get
correct values for the winding resistance after shutdown, the winding temperature variation have to be
extrapolated backwards in time to the instant of shutdown. Therefore the resistance measurement
should be conducted as soon as possible after short-circuit method was applied.
 Temperature rise measurement 85

°C
A temperature rise
Short-circuit Total loss injection B settling time
Connection C 3h test time
Rated current injection D 1h test time
Winding resistance winding resistance
DC current E
measurement measurement

A B C D E

Figure 14.3.2: Temperature rise procedure h

As seen in Figure 14.3.2 the whole measurement is divided in several steps. Unless otherwise specified,
the temperature rise test is conducted with the transformer connected on the maximum current tapping.

a) Total loss injection

test current corresponds to the total losses of the transformer (above rated
current to the extent necessary for producing an additional amount of loss equal
to the no-load loss at rated voltage)
top-oil temperature and cooling medium temperature are monitored
test will be continued until steady-state liquid temperature rise are established,
which is terminated when the rate of change of the top-oil temperature rise has
fallen below 1K per hour
readings will be taken at regular intervals (readings during the last hour is taken
as the result of the test)

b) Rated current injection

test current reduced without a break to rated current, which condition
during testing time (1h), continuous temperature records of top-oil, winding hot-
spot (if measured) and external cooling medium will be taken at least every 5min

c) Winding resistance measurement

cooling devices maintained to the same condition as during the current injection
measurement starts after power supply and short-circuit connections have been
removed; DC current sources will be connected to the windings (short break)
fast decreasing of winding temperature and its resistance after shutdown
measured resistance values have to be corrected backwards
86 Electrical Testing Of Power Transformers

Acceptance criteria
IEC specifies the maximum permitted ambient temperature with 40°C, and the corresponding permitted
temperature rise of the winding with 65 K. The maximum average winding temperature is therefore:
65 + 40 = 105°C.

14.4 Measuring circuit

(1) Measuring circuit for temperature rise

The main measuring circuit and the

circuit for measurement of losses,
voltages and currents is principally the
same for the measurement of load
losses (section: 5 Short-circuit voltage
& load loss).

= ,

=
=
=
= Figure 14.4.1: Principle test circuit for
temperature rise

(2) Measuring circuit for determination of winding resistance after shutdown

The winding temperature must be

determined indirectly, measuring the
winding resistance. It is measured
before the test and once again
immediately after the completion of
the current injection (several values
needed for extrapolation of winding
temperature to the instant of
shutdown). The windings will be
connected to separate DC circuits.
Care has to be taken that the
readings are correct, because direct
current needs to stabilize to a steady
state (inductive voltage drop in the
winding). A total duration of 20 min for
the measurement is appropriate for
the determination of the cooling down
curve.
Figure 14.4.2: Test circuit for winding resistance
(different from figure: windings connected
in series for simultaneous measurement)
 Temperature rise measurement 87

14.5 Appendix: Measure arrangement of customized vector group

Test circuit

Measuring instruments

Yokogawa MV100/MV1000
data acquisition recorder/system
(for thermal sensors)
12 channels (MV100)/
24 channels (MV1000)
data collection over
Ethernet network
external storage media (floppy
disk, zip disk, ATA flash memory)
MV 100 built-in control room

LumaSHIELD (Luma Sense)

Fiber Optic Winding Hot Spot Temperature Monitor/Controller
Accuracy: ± 0,8°C (total accuracy, includes signal conditioner
and sensor errors)
Resolution: 0,1°C

The arrangement is similar to the short-circuit voltage & load loss measurement. The respective measuring
equipment (power analyzer: Yokogawa WT3000, instrument transformers), which is needed for power
observation, has been illustrated before.
88 Electrical Testing Of Power Transformers

14.6 Hot-spot measurement

The hot-spot temperature ( ) is the hottest temperature of winding conductors in contact with solid
insulation or insulating liquid. The hot-spot winding temperature rise ( ) is the difference between
hot-spot winding temperature and the external cooling medium temperature. It can be determined
through calculation based on the result of the temperature rise test (at rated power <20MVA) or by
direct measurement, as a special test and by agreement between manufacturer and purchaser (at rated
power >20MVA). Usually we are only considering the direct measuring method.

Standard Section/Clause Type of test

60076-2
IEC clause 7.10: “Determination of the hot-spot winding
temperature” Special

VDE 0532-76-2
Table 14.6.1: Associated Standards

Direct measurement during the temperature rise test

A number of thermal sensors (optical fibre sensors) will
be mounted inside the windings in position where it is
supposed the hot-spots are located. Usually hot-spots Windings

emerge at the top of the windings. When more than one

sensor is used on the same winding, the maximum
Standard Isolation
reading shall be taken as the hot-spot winding spacer
temperature. The hot-spot winding temperature rise
( ) is then obtained by: Grooved Isolation
spacer

Thermal
= + sensor

=
=
1
=

Figure 14.6.1: Mounting of thermal sensors

Dissolved gas-in-oil-analysis
For large mineral oil-immersed power transformers, in which additional flux effects are potential risk
factors, a chromatographic analysis of dissolved gas-in-oil may allow the detection of possible local
overheating. Therefore the gas-in-oil analysis gives another opportunity of hot-spot indication and is a
routine test for transformers with > 72,5kV (IEC 60072-2, Appendix D), conducted by intern or
extern laboratory.
 Measurement of cooling losses 89

15 Measurement of cooling losses

15.1 General

The value of cooling losses will be determined to get a completion of the overall losses of the
transformer (additionally to winding and core losses). The cooling losses are defined as the power
which is taken by the installed cooling system (fans, liquid pump motor). These losses will be measured
by means of the power analyser “Yokogowa WT3000”.

15.2 Measuring circuit and instruments

Figure 15.2.1 shows a simplified extract of the electric cabinet’ circuit layout (only essential parts).
Generally the electric cabinet is a box for electrical or electronic equipment to mount switches, knobs
and displays, to control and monitor auxiliary equipment (e.g. fans, liquid pumps, sensors) of the
transformer. Moreover it has an important function in preventing electrical shock to equipment users
and protects the contents from the environment. In the figure below only fans are pictured exemplary
and for better comprehension.

Power Measuring Load

t Supply Unit

400V/50H

Motor circuit breakers +

Auxiliary switches

Yokogawa WT3000 Protective ground conductor (PE)

3 of 4 input elements used

Electric cabinet

Fans

Figure 15.2.1: Example of measuring configuration for cooling losses

90 Electrical Testing Of Power Transformers

The power, which is taken by fans and liquid pump motors, is measured with the digital precision power
analyzer “Yokogawa WT3000”. It is switched in between power supply and load and is not part of the
actual circuit layout. Because there is a three-phase power supply different measuring configurations
are feasible. An appropriate setup is the three-wire system, measuring three voltages and three
currents (one for each phase), illustrated below.

U
V
W U
Load
V
W

Yokogawa WT3000
Voltage inputs
Digital Precision Power Analyzer
Current inputs
4 Input terminals available
Accuracy: 0,01% of basic readings
back surface
0,02% of basic power readings
Direct Current Input: up to 30 A
Direct Voltage Input: up to 1000 V
 Sound level measurement 91

16 Sound level measurement

16.1 Purpose and Standards

Today sound emissions, caused by electric installations such as transformers, have to be considered,
especially in populated areas. To protect from these noise inconveniences, electric components are
required to operate within specified noise limits. For this reason knowledge about sound emissions from
the transformer is really important.

Standard Section/Clause Type of test

60076-10
Part 10: “Determination of sound levels”
IEC
Part 10-1: Draft: “Determination of transformer and reactor sound
levels” – User Guide Special- or
C57.12.90 other test
IEEE
Clause 13: “Audible sound emissions”

VDE 0532-76-10
Table 16.1.1: Associated Standard

16.2 General

Operating transformers generate sound, or more correctly noise. Sources of a transformers sound are:

No-load sound (noise from the core)

No-load sound is caused by magnetostriction (elastic length variations of iron core parts) generated
during the magnetizing process. The oscillations are transferred by the oil as mechanical vibrations to
the tank walls and radiated to open air. The amplitude depends on the flux density in the core and the
magnetic properties of the core steel. The frequency spectrum of the audible sound consists mainly of
twice the rated frequency and its even multiples. For example in a 50 Hz system, the audible sound
consists of the harmonics: 100 Hz, 200 Hz, 300 Hz etc.

Load sound (noise from the windings)

Due to its magnetic forces load current generates vibrations in the winding, tank wall and magnetic
shields. The load sound power is strongly dependent on the load current. Depending on the type of
cooling, pump and fan noises are added.
92 Electrical Testing Of Power Transformers

16.3 Measuring circuit

According to IEC, the transformer has to be energized during the test. The condition of “energized” will
be agreed between manufacturer and customer. It means either no-load or load condition in the
following permissible combinations:

(1) Transformer energized; cooling equipment and any pumps out of service
(2) Transformer energized; cooling equipment and any pumps in service
(3) Transformer energized: cooling equipment out of service, pumps in service
(4) Transformer not energized: cooling equipment and any pumps in service

The measuring circuit for load noise measurement is the same for load loss measurement (see section:
5 Short-circuit voltage & load loss), just like the circuit for no-load noise measurement is similar to no-
load loss measurement (section: 6 Measuring the no-load loss & no-load current). To get the total
sound level, load- and no-load noises have to be sum up.

Measuring points (acc. IEC)

The position of the microphones has to be at the prescribed contour (Figure 16.3.1), with equally
distance to each other (not more than 1 m apart). In case of a detached cooling system, which is placed
more than 2 meter away from the transformer, the Sound level of the cooling system has to be
measured separately.

Natural air cooling (e.g. ONAN)

Contour must be spaced 30 cm from the
principal radiating surface

Forced air cooling (e.g. ONAF)

Contour must be spaced 2 m from the principal
radiating surface

Tank height < 2,5 m Tank height 2,5m

Measurements carried Measurements carried
out at half the tank out at 1/3 and 2/3 of the
height tank height

Figure 16.3.1: Typical microphone positions

 Sound level measurement 93

16.4 Measuring procedure

Test conditions
During the entire measurement the excitation voltage should be sinusoidal and keep its rated
magnitude and frequency. It is important to allow enough time for any DC magnetization to decay
before starting the test, because the remaining DC flux causes odd harmonics in the sound spectrum
and increases the total sound level. Moreover the transformer under test should be placed as far as
possible from reflecting walls, especially parallel to walls, or other equipment to minimize reflections.

Calibration
The measuring equipment has to be checked with a calibrator before and after the test. The test
instruments are not allowed to differ more than 0,3dB (acc. IEC) or 1dB (acc. IEEE).

Test sequence
The A-weighted sound pressure level of the background will be measured right away before and after
the actual transformers sound level test (A-weighted means all hearable frequencies for humans; 20Hz
- 20kHz). The positions of the microphones are therefore the same as for the actual sound level
measurement. The background level is the arithmetic average of these measuring points, if the variation
is 5dB or less (acc. IEC) or 3dB or less (acc. IEEE). Otherwise a certain formula will be used to get the
average background sound level. The actual sound measurement will be conducted similar to
background noise level with transformer permissible conditions (see 16.3).

Acceptance criteria
The sound level measurement is valid if the difference between the first and second background
measurement is below 3dB (acc. IEC).

Measuring instruments

Typ B&K 2260 Typ B&K 4189 Typ B&K 4231

Measuring device Microphone Sound level calibrator

linear operating range: wide dynamic ranges: from calibration accuracy ±0.2dB
80dB adjustable to give 14.2dB(A) to 146dB and 20dB(A)
full-scale readings from to 162 dB (3% distortion limit) conforms IEC Standard
80dB to 130dB in 10dB
steps very wide operating temperature
range and low ambient-
max. peak level: 3dB temperature coefficient
above full scale reading

6.3 Hz - 20 kHz frequency

range in real-time 1/3-
octave bands
 Measurement of zero-sequence impedance 95

17 Measurement of zero-sequence impedance

17.1 Purpose and Standards

It is usual in performing system design calculations, particularly those involving unbalanced loadings
and for system earth fault conditions, to use the principle of symmetrical components. This system is
described in positive, negative and zero-sequence impedance values to the components of the
electrical system. For a three-phase transformer, the positive and negative sequence impedance values
are identical to that value described above, but the zero-sequence impedance varies considerably
according to the construction of the transformer.

Standard Section/Clause Type of test

60076-1
IEC Clause 10.7 “Measurement of the zero-sequence impedance(s) on
three-phase transformers”
Special- or
C57.12.90 other test
IEEE
Clause 9.5 “Zero-phase-sequence impedance”

VDE 0532-76-1
Table 17.1.1: Associated Standards

17.2 General

Calculations in a symmetrical three-phase system are only possible as far as the values of the phases
are simply shifted by 120° and have the same magnitude. Otherwise each phase has to be considered
and calculated separately, which is relevant in case of single short-circuits for instance (impedances
differ from those in a symmetrical network). Generally it is necessary to convert the given unbalanced
three-phase system into a balanced three-phase system. It will be distinguished in:

Phase system Rotation order

Positive-sequence U-V-W
Negative-sequence U-W-V
Zero-sequence All phases in same direction

Table 17.2.1: Phase systems

Corresponding with voltage- and current components, the transformers positive- and negative-
sequence impedances are the same, whereas the zero-sequence impedance can differ significantly,
depending on winding connection and design. An example for an unbalanced three-phase system
consisting of positive-, negative- and zero-sequence system is shown in Figure 17.2.1.
96 Electrical Testing Of Power Transformers

Figure 17.2.1: Unbalanced three-phase system

Zero-sequence impedance
After definition the zero-sequence impedance is the impedance measured between phase terminal
(three phases connected) and neutral, which is only possible in star- or zig-zag- connected windings.
The zero-sequence impedance, attributed to each phase, is three times the measured value.

= ·

= ( )
=
Figure 17.2.2: Zero-sequence impedance

Normally it is given as a percentage value of the basic impedance , which can be calculated
as follows:

= · =

= ( )
=
 Measurement of zero-sequence impedance 97

17.3 Measuring circuit

It should be noted that the directly measured impedances are not used. For further calculation
equivalent zero-sequence circuits are more preferred. The equivalent circuits and principal test
connections for different neutral- and network conditions are shown in Figure 17.4.1 respectively.

Figure 17.3.1: Principle measuring circuit

= =
= =

17.4 Measuring procedure

The measurement must be carried out at rated frequency and always with the active part in the tank,
because of its high influence on the zero-sequence impedance. The zero-sequence flux may cause
excessive heating in metallic structural parts such as tank, cover or clamping construction. That’s why
the measuring current must not be higher than 30% of nominal current . Currents up to nominal
current are only permitted for a very short time. The applied voltage must not exceed the phase-to-
neutral voltage, which occurs during normal operation.
98 Electrical Testing Of Power Transformers

Equivalent circuit for the zero

Transformer group Measuring circuit
sequence system

Figure 17.4.1: Examples of test connections for zero sequence impedance measurement
 Measurement of zero-sequence impedance 99

17.5 Appendix: Measure arrangement of customized vector group

Test circuit

Measuring instruments

Yokogawa WT3000 Rack

Digital Precision Power Analyzer contains Power Analyzer, Votlage-
and Current transducer controls
0,01% of
Accuracy: basic readings
0,02% of
power readings
Bandwidths: DC and 0,1Hz-1MHz
Direct inputs: up to 30 A/1000 V

Tettex TMS-PT-581
Voltage transducer
Output voltage: 100 V
0,5, 1, 2, 5, 10, 20, 50,
Measuring range:
100, 200, 400, 800 kV
0,04 % (5-800 kV)
Overall accuracy: 0,1% (1-2 kV)
0,15% (0,2-0,5 kV)

Tettex TMS-CT-582
Current transducer
Test field 1 Test field 2
Output current: 1A 1A
1, 2,10, 20, 40, 100, 1, 2, 5, 10, 20, 50, 100,
Measuring range:
500, 1000, 2000, 4000A 500, 1000, 2000 A
0,005% (40-4000 A) 0,005% (20-2000 A)
0,01% (10-20 A) 0,01% (5-10 A)
Overall accuracy:
0,03% (4 A) 0,03% (2 A)
0,05% (2 A) 0,05% (1 A)
100 Electrical Testing Of Power Transformers
 Measurement of the harmonics of the no-load current 101

18 Measurement of the harmonics of the no-load current

18.1 Purpose and Standards

Generally it is of interest to keep the amount of harmonics as low as possible. For modern power
transformers the harmonic content of the no-load current is actually of little interest, because the no-
load current is only about 0,1 – 0,5 % of the rated current, which does not affect the grid effectively
enough. It could be interesting in case of old transformers. It has to be noted that there are no
applicable Standards according this measurement.

18.2 General

Although a sinusoidal voltage is applied at the transformers terminals, a non-sinusoidal magnetizing

current occurs, caused by the non-linear relationship between magnetizing force H and the flux density
B (see Figure 6.2.2: Magnetic loop). As seen in Figure 18.2.1 the magnetizing current, as well as the
harmonic oscillations (Figure 18.2.2), depends on the applied voltage, which represent the flux density
B. The more voltage is applied, the odd-numbered harmonics of the current increase as well, whereas
the fundamental harmonic decreases.

Figure 18.2.1: Magnetizing current Figure 18.2.2: Harmonics of the

magnetizing current
102 Electrical Testing Of Power Transformers

18.3 Measuring circuit

Additionally to the measuring circuit, which is exactly the same as for the measurement of no-load
losses and no-load current (see sections: 5 and 6), a harmonic-analyzer (Fluke 41B) is used in one of
the ammeter circuits. The possibility of voltage distortion should be excluded. Therefore the generators
and matching transformers have to be large enough, obtaining a linear magnetizing characteristic and a
voltage free of distortion.

18.4 Measuring procedure

The measuring is conducted at 90%, 100% and 110% of the transformers rated voltage for each of the
three phases. The voltage is increased gradually from zero to the full values. Switching on directly
would create transient inrushes including DC components, which could saturate the iron core and
increase the measuring uncertainty.
 Measurement of the harmonics of the no-load current 103

18.5 Appendix: Measure arrangement of customized vector group

Test circuit

Measuring instruments

Fluke 41B

Powermeter/ Power harmonics analyzer

Fundamental to 13th Harmonic (ampere):
Accuracy1:
±(3% of rdg + 3 digits)

13th to 31st Harmonic (ampere):

13th: ±(3% of rdg + 3 digits)…31st: ±(8% of rdg + 3 digits

Fundamental (phase):
± 2 degrees

2nd to 31st Harmonic (phase):

2nd: ± 5 degrees…31st: ± 20 degrees

1
referred to harmonics measurement
 Frequency response analysis measurement (FRA) 105

19 Frequency response analysis measurement (FRA)

19.1 Purpose and Standards

Frequency response analysis is a major advance in transformer and reactor condition analysis, allowing
you to “see” inside transformers without costly de-tanking. Since the FRA test is used to detect
mechanical movement or damage in a transformer, it is most appropriately used after some event or
condition that has the possibility of causing mechanical movement or electrical damage to the
transformer assembly (e.g. earth quake). Some of the typical scenarios where FRA - measurements
may be used include:

Factory short-circuit testing

Installation or relocation
After a significant through-fault event
As part of routine diagnostic measurement protocol
After a transformer alarm (i.e. sudden pressure, gas detector, Buchholz)
After a major change in on-line diagnostic condition
(i.e. a sudden increase in combustible gas, etc.)
After a change in electrical test conditions (i.e. a change in winding capacitance)

Standard Section/Clause Type of test

60076-18
“…when a frequency response measurement is required either on-site or
IEC in the factory either when the test object is new or at a later stage. This
standard is applicable to power transformers, reactors, phase shifting
transformers and similar equipment”
Special test
C57.149
IEEE “Guide for the Application and Interpretation of Frequency Response
Analysis for Oil-Immersed Transformers”

VDE 0532-76-18
Table 19.1.1: Associated Standard

19.2 General

The most common cause of failure in the 20-400 MVA transformer group is general ageing of insulating
material. It is therefore desirable to periodically check the mechanical condition of transformers during
their service life, particularly for older units and after significant short-circuit events, to provide an early
warning of impending failure. Such a capability is perhaps just as important as the ability to diagnose
suspected short-circuit failures.
106 Electrical Testing Of Power Transformers

Conventional condition monitoring techniques such as dissolved gas analysis (DGA) are unlikely to
detect such damage until it develops into a dielectric or thermal fault, so a specialist technique is clearly
required for the monitoring and assessment of mechanical condition. Thus the most important
monitoring technique is one that can give the information about insulating condition for ageing or
insulating deterioration products.

Resonance circuits
The image on the right-hand side shows the
complex electrical relationship which exists
between the transformer’s windings, its core,
the tank, the oil, the insulation, and the tank
wall. At the first approximation a transformer
windings can be represented by a complex
ladder network with series inductance and
capacitance as well as the parallel capacitance
to ground. These circumstances lead to
Resonance circuits (RLC - circuits), Series- as
well as Parallel resonance circuits with resonant
properties.
Figure 19.2.1: Illustration of resonance circuits
The transfer function of such networks,
inside the transformer
calculated by the FRA – measurement, shows
a number of poles at the resonance frequency
of the local L and C circuits. A breakdown between the turns or coils of the transformer under test
corresponds to a short circuit of one or more of these local LC - networks. This will result in shifting the
resonant pole to another frequency or the creation of a new pole, which can be interpreted as an
indication of failure in transformer’s structure.

19.3 Measuring circuit

Generally there are four different measuring configurations, which are explained briefly in the following.
The measurements are conducted on each of the transformer HV-windings (A,B,C) with the principle
tapping position.

End-to-end measurement
A frequency response measurement made on a single coil
(phase winding) with the source lead ( ) connected to one end
and the response lead ( ) connected to the other end.

End-to-end (YNd1)

Capacitive inter-winding measurement

A frequency response measurement made on two neighbouring
coils (windings of the same phase) with the source lead ( )
connected to one end of a winding, the response lead ( )
connected to one end of another winding and with the other
winding ends floating.
Capacitive inter-winding (YNd1)
 Frequency response analysis measurement (FRA) 107

Inductive inter-winding measurement

A frequency response measurement made on two neighbouring
coils (windings of the same phase) with the source lead ( )
connected to one end of the higher voltage winding, the
response lead ( ) connected to one end of the other winding
and with the other ends of both windings grounded.
Inductive inter-winding (YNd1)

End-to-end short-circuit measurement

A frequency response measurement made on a single coil
(phase winding) with the source and lead ( ) connected to
one end, the response lead ( ) connected to the other end.
Another windings of the same phase are short-circuited.

End-to-end short-circuit (YNd1)

19.4 Measuring procedure

The Frequency Response Analysis is a test measurement made to the winding structures of a power
transformer, performed by an instrument injecting a voltage signal into the top of the selected winding
structure and measuring the voltage signal appearing at the bottom of the same or neighbouring
selected winding structure, calculating the transfer function. This is performed for each winding
structure of the transformer. Values of impedance and phase angle are measured over the given
frequency range and are available as logarithmic graphs, for each of the tested winding structures. This
graph is a representation of the frequency response of the windings at different frequencies (example in
Figure 19.4.1).

Any transformer under test should be completely isolated from any high-voltage source or power
system source. The transformer tank should be grounded. Two of the bushings will be used. The
measurement (most accepted: End-to-end measurement) will be conducted for all windings in
maximum tapping position. The stabilizing winding (if existing) remains and will not be earthed.

Frequency range Measuring instrument

The test is made over the entire frequency range
up to 2 MHz, so as to be able to diagnose TETTEX FRA 5310
problems in the core, clamping structure,
windings and interconnections. Successive Frequency
frequency measurements are equally spaced so Response Analyser
as to obtain adequate resolution in the higher
frequencies so as to give unambiguous
Accuracy: ± 0,1dB
diagnosis. The highest measuring frequency
depends on the highest voltage level of the Frequency
10Hz – 10MHz
transformer. For an example of a Frequency Range:
resonance analysis for one winding see the Sweep Logarithmic or
following figure. Mode: Linear

> 72,5 kV at least 1 MHz

< 72,5 kV at least 2 MHz
108 Electrical Testing Of Power Transformers

effective in determination of following conditions

Low Frequency Range Core problems within the transformer

(0 – 50 kHz) Presence of shorted turns.

Axial collapse
Medium Frequency Range
Localized winding movement
(50 – 500 kHz)
Winding asymmetry

High Frequency Range Movement of main and tap winding leads

(500 – 2000 kHz) LTC and DETC connections

Figure 19.4.1: Example of a Frequency resonance analysis

 FDS measurement for moisture estimation 109

20 FDS measurement for moisture estimation

20.1 Purpose and Standards

Traditional oil sampling methods requires the use of an equilibrium diagram for evaluating moisture in
transformers and can result in errors in the assessment. The application of conventional equilibrium
charts causes inaccurate results due to the uncertainties during the sampling and water-in-oil
measurement. Together with very long time constants for equilibrium processes, it leads to a very poor
accuracy. The dielectric response method, in contrast, is a very reliable method providing a high degree
of accuracy in assessing the moisture content in the paper insulation.

Moisture in the solid part of power transformer insulation (paper, pressboard) is one of the most critical
condition parameters. Water enters transformers from the atmosphere (breathing, leaky seals) and
during installation and repair. Aging of the oil-paper insulation also increases the moisture level.
Typically, the solid part of the insulation structures holds most part of the water, i.e. 2000 times more
than the oil. Measurement of the water content in oil-paper insulation is therefore a helpful tool for
making an assessment of the ageing of the cellulose and a key factor to ensure transformer’s reliability
and longevity.

Standard Section/Clause Type of test

60422
“…supervision and maintenance of the quality of the insulating
oil in electrical equipment.”
IEC Special test
60296
“It applies to oil intended for use in transformers, switchgear and
similar electrical equipment in which oil is required for insulation
and heat transfer.”
Table 20.1.1: Associated Standards

20.2 General

Today transformers are not automatically replaced, if they have reached the end of their life span, but
left in service as long as possible. In contradiction to the past, power transformers are operated
nowadays at or above rated power. This accelerates the ageing process of the inner insulation,
particularly of the insulation paper, which cannot be easily replaced.

Moisture entering in oil-paper insulations can cause dangerous effects:

Decreases the dielectric withstand strength

Accelerates cellulose aging (the so-called: de-polymerization)
Causes the emission of gas bubbles at high temperatures and may lead to a sudden electrical
breakdown.
110 Electrical Testing Of Power Transformers

It is important to have reliable tools for measuring the water content in the oilpaper insulation. This way
unexpected breakdowns can be avoided and the maintenance, repair or replacement can be scheduled
in time. Those faults are often indicated by the oil analysis, which is a proven and meaningful tool.
Dielectric spectroscopy methods are used to determine the moisture in the solid insulation of power
transformers.

1. Polarisation and depolarisation current (PDC)

Dielectric spectroscopy in the time domain.

2. Frequency domain spectroscopy (FDS)

Measurement of the capacitance and the dissipation factor over a wide
frequency range ( Hz to kHz)

Although the results of PDC and FDS methods are comparable and can be transformed from the time
domain into the frequency domain and reversely, both methods have advantages and disadvantages. A
new approach combines both methods. The FDS measurement is replaced by the PDC method in the
low frequency range and the results are transformed into the frequency domain, whereas the FDS is
used for higher frequencies, which can be done rather quickly.

The measuring instrument OMICRON Dirana makes use of this principle and shortens the
measurement time to a minimum. Two input channels for simultaneous measurement of two insulation
gaps make it even faster. The software takes also the conductivity of the oil into account. This makes
the results for aged transformers much more reliable compared to the standard model curves which
were used in the past. The analysis of the gas in oil is a well-proven method of analysis but must be
complemented by efforts to locate any faults. This way important maintenance can be performed in time
to avoid a sudden total failure. The fault location can be successfully performed using modern type test
equipment for resistance, winding ratio, short circuit impedance, capacitances and dissipation factor
(tan ), FRA and PD measurements.

20.3 Measuring circuit

The dielectric response is measured by a OMICRON

three-terminal measurement that Power output Dirana
includes the output voltage (Power
output), the measurement current (CH A) CH A
and a guard (CH B) to prevent
disturbances due to current paths caused
by dirty bushings or electromagnetic CH B
fields and to short-circuit the winding-
capacitance, which is necessary for the
return-voltage-analysis. During the test LV HV
the tank is generally earthed.

Test object

Figure 20.2.1: Principle measuring circuit

 FDS measurement for moisture estimation 111

Measuring instrument

OMICRON Dirana

FDS/PDC Dielectric Response Analyzer

Measuring range: 10 µHz...5kHz

Measuring voltage/current: 200 / 50

Measuring time: 2mHz…1kHz: approx. 15min
100 Hz…1kHz: < 3h
10 Hz…1kHz: < 6h

20.4 Measuring procedure

The OMICRON instrument derives the moisture content in paper or pressboard from properties such as
polarization current, complex capacitance, and dissipation factor. Each of these parameters is strongly
affected by moisture.

The dissipation factor plotted against frequency shows a typical S-shaped curve (see example in Figure
20.4.1). With increasing moisture content, temperature or aging the curve shifts towards the higher
frequencies. Moisture influences the low and the high frequency areas. The linear, middle section of the
curve with the steep gradient reflects oil conductivity.

This moisture determination is based on a comparison of the transformers dielectric response to a

modeled dielectric response. A fitted algorithm compares the measured data with the model data and
calculates the geometry data, the moisture content as well as the oil conductivity. Only the oil
temperature needs to be entered.

Measurement steps
For e.g. a two-winding transformer which has been disconnected from the network:

Apply the test voltage to the HV windings

Current measurement via LV windings
Connect the guard to the bushing flange (ground)

Unlike the conventional equilibrium method, the measurement can be taken right away. There is no
requirement to allow the transformer to cool, or wait until moisture equilibrium between paper and oil
has been reached.
112 Electrical Testing Of Power Transformers

Assessing the results

The IEC Standard categorizes
moisture saturation of more than 6 %
as “moderately wet“, which is
equivalent to a moisture content of
approximately 2.2 %. At this level the
dangerous effects caused by water
can affect the insulation. Based upon
this, corrective action should be taken
for e.g., a drying process.

Moisture content relates water mass

to the material mass, whereas
moisture saturation relates water
mass to the maximum water mass a
material can adsorb. Figure 20.4.1: Example of a moisture measurement
(curve B with increased moisture)
 Measurement of insulation capacitances & loss factor (tan ) 113

21 Measurement of insulation capacitances & loss factor

(tan )

21.1 Purpose and Standards

Similar to the Insulation resistance measurement, the determination of the loss factor (tan ) allows
certain conclusions about the condition of the transformer insulation and serves as a reference value for
later measurements, especially for comparisons during operation at later stage.

Standard Section/Clause Type of test

Routine test
60076-1
> 72,5 kV
Clause 10.1.3: Power transformers – Part 1: General,
IEC
“Measurement of the dissipation factor (tan ) of the
Special test
insulation resistance capacitances”
72,5 kV

Routine test
for class II
transformers
C57.12.90
IEEE
Clause 10.10 “Insulation power-factor tests”
Other test
for class I
transformers
Table 21.1.1: Associated Standards

21.2 General

These special tests include the determination of the winding capacitance with respect to ground and
also the loss factor. The loss factor is defined (by IEC; also known as power factor in IEEE Standards)
as the ratio between the absorbed active power and the absolute value of the reactive power, which
corresponds to tan . In the ideal insulator, the angle would be 90°C as it is purely capacitive and non-
conducting. However in real insulators, there will be some leakage current and resistive losses through
the dielectric. There is no relationship between the loss factor and the transformers withstand (dielectric
test). Moreover its dependence on temperature is substantial and erratic. The various insulation
materials and liquids used in the transformer result in large variations of loss factor as well. That`s why
it indicates to information about the condition of the oil.
114 Electrical Testing Of Power Transformers

21.3 Measuring circuit

The loss factor will be generally measured in special bridge circuits, based on comparing the
capacitance and balancing them. The measuring instrument Tettex MiDAS 2880 does this automatically.
External influences must be reduced by a coaxial cable. In the figures below examples of the measuring
circuit are given.

Tettex MiDAS 2880

Mobile Insulation Diagnosis & Analysing System

Accuracy:
± 0,5% of readings ± 0,01% of resolution
(Dissipation Factor)
Accuracy:
± 0,3% of readings ± 0,3 pF
(Capacitance)

Figure 21.3.2: Test connection for HV side Figure 21.3.1: Test connection for LV side
against tank and LV side against tank and HV side

21.4 Measuring procedure

A 10kV voltage at 50Hz is applied to the test object. Depending on the number of windings and whether
there exists a stabilizing winding, the number of measurements will differ. Usually measurements will be
taken of:

low-voltage side against the tank and high-voltage side

high-voltage side against the tank and low-voltage side
low-voltage and high-voltage sides against the tank

The winding of the test object that is not measured and the tank are in each case grounded via the
measuring equipment. If a stabilizing winding exist more measuring combinations are applicable. The
gathered values are relevant for later measurement as reference values.
 Measurement of excitation current with 400V 115

22 Measurement of excitation current with 400V

22.1 General

This test will be performed with portable instruments (common digital multimeter), carried out at the
jobsite before the transformer will be connected to the power supply. The excitation current, measured
with a 400V supply in the field, allows an evaluation of the transformer’s condition and serves as a final
check. Because the value is already known before dispatch, possible rough damage (internal
winding/core faults) caused by shipping can be detected quickly this way. That’s why the measured
excitation current, which has a fixed value, is also understood as a “fingerprint” for the respective
transformer.

Note: Usually known as “no-load measurement”, the gathered results can be used to obtained further
basic parameters like transformer turns ratio or polarity.

22.2 Measuring procedure

The excitation current is the current flowing into the high voltage winding with the low voltage side open.
This current should be proportional to the no-load acceptance test, regarding the use of test voltages
different from nominal values. Therefor an alternating voltage of 400V will be applied to the high voltage
winding of the transformer in each of its phases (low voltage side open).

Care have to be taken that the core is fully unsaturated before the measuring the excitation current.
Otherwise a distorted value will be measured, leading to false conclusion about transformer’s condition.
To avoid saturation of the core, a slightly higher voltage than system voltage will be applied to the
transformer terminals and will be then reduced to zero. The core should be free of any residual
magnetism after that.
 Calibration of the Winding Temperature Indicator (WTI) 117

23 Calibration of the Winding Temperature Indicator (WTI)

23.1 General

The Winding Temperature Indicator (WTI) is supposed to indicate the hottest spot in the winding. An
alternative are fiber optic temperature sensors can be imbedded directly into the winding (14.6 Hot-spot
measurement) and are more accurate, which may be worth it, since exact values for hot-spot
temperatures are essential when higher loading is required. They can be used as monitoring device or
turn on additional cooling or activate alarms, as top oil thermometers do.

The WTI, also known as transformer temperature transmitter, can

only be used in conjunction with a pointer thermometer, which is
used to present the hot-spot temperature in the transformer. The
measured value is captured by a liquid filled temperature sensor
and is then transmitted via a spiral spring which is linked with zero
backlashes to a pointer spindle.
The temperature transmitter heats the sensors of the pointer
thermometer by simulating the winding temperature. The
temperature gradient between the winding and the coolant is
dependent on the current in the winding of the transformer. To this MESSKO pointer thermometer
winding current the secondary current of the bushing current
transformer is proportional, which is used to supply the heating
resistor of the temperature transmitter. This causes an increase
corresponding to the transformer loads in the actually measured oil
temperature. When it comes up to monitoring issues, the
measured value can also be transferred to a digital display, via an
installed Pt100 resistor. The temperature gradient at rated load,
which is a known data from the temperature rise test, have to be
set by DIP switches and calibrated before putting into operation.

Bushing CT input

MESSKO Transformer
temperature transmitter (WTI)

setting temperature gradient

for electrical connections

for temperature sensor

of dial thermometer

Inside temperature transmitter

118 Electrical Testing Of Power Transformers

23.2 Calibration

As mentioned, the adjustment value for the calibration is calculated by means of the temperature rise
gradient and the bushing current transformer’s rated current, which are known data of the transformer
at rated power. By injecting the rated CT current and by means of a specific curve (thermal replica or
formula) of the heating resistance, the Winding temperature indicator can be calibrated before
installation.

Steps

connect current supply

inject a current corresponding to the rated power
(regard BCT ratio) of the transformer (usually
about 2 A)
maintain the current for a couple of minutes and
check the rise of the indicator
temperature rise gradient has to correspond to Variable current source for calibration
rated current of bushing CT (see specific curve)

Connection diagram for calibration