IEEE Format FRA Draft Guide D6

IEEE PC57.
149/D6 April 2009
IEEE PC57.149™/D6
Draft Trial-Use Guide for the Application and
Interpretation of Frequency Response Analysis
for Oil Immersed Transformers
Prepared by the Working Group for the Development of a Guide for Transformer Frequency Response
Working Group of the
IEEE Power Engineering Society Transformers Committee
Copyright © 2006 by the Institute of Electrical and Electronics Engineers, Inc.

Three Park Avenue
New York, New York 10016-5997, USA
All rights reserved.
This document is an unapproved draft of a proposed IEEE Standard. As such, this document is subject to
change. USE AT YOUR OWN RISK! Because this is an unapproved draft, this document must not be
utilized for any conformance/compliance purposes. Permission is hereby granted for IEEE Standards
Committee participants to reproduce this document for purposes of IEEE standardization activities only.
Prior to submitting this document to another standards development organization for standardization
activities, permission must first be obtained from the Manager, Standards Licensing and Contracts, IEEE
Standards Activities Department. Other entities seeking permission to reproduce this document, in whole or
in part, must obtain permission from the Manager, Standards Licensing and Contracts, IEEE Standards
Activities Department.
IEEE Standards Activities Department

Standards Licensing and Contracts
445 Hoes Lane, P.O. Box 1331
Piscataway, NJ 08855-1331, USA
Copyright © 2006 IEEE. All rights reserved.

This is an unapproved IEEE Standards Draft, subject to change.
IEEE C57.149/D6 April 2009
Abstract: This trial use guide is applicable to the measurement of Frequency Response
Analysis (FRA) of oil immersed power transformers. It is intended to provide the user with the
requirements and specifications for instrumentation, procedures for performing the tests,
techniques for analyzing the data, and recommendations for long-term storage of the data and
results.
Keywords: Frequency Response Analysis (FRA), magnitude, phase angle, attenuation,

frequency domain, transfer function, resonance, RLC network, Bode Plot, impedance,
admittance, deviation.
Copyright © 2006 IEEE. All rights reserved. ii

IEEE C57.149/D6 April 2009
Introduction
(This introduction is not part of IEEE PC57.149/D4, Draft Trial-Use Guide for the Application and
Interpretation of Frequency Response Analysis for Oil Immersed Transformers.)
<Select this text and type or paste introduction text>
Patents
Attention is called to the possibility that implementation of this trial-use guide may require use of subject
matter covered by patent rights. By publication of this trial-use guide, no position is taken with respect to
the existence or validity of any patent rights in connection therewith. The IEEE shall not be responsible for
identifying patents or patent applications for which a license may be required to implement an IEEE
standard or for conducting inquiries into the legal validity or scope of those patents that are brought to its
attention.
Participants
At the time this draft trial-use guide was completed, the Working Group for the Development of a Guide
for Transformer Frequency Response Working Group had the following membership:
Charles Sweetser, Chair

<Vice-chair Name>, Vice-chair
Raj Ahuja Eduardo Garcia Triomphant Ngnegue

Sokom An Ramsis S. Girgis Dennis Nmarlow
Peter Balma David Goodwin Bipin Patel
Danny Bates Richard Graham Paulette Payne
Bill Boettger Ernst Hanique Mark Perkins
Rich. Breytenbach Roger R. Hayes Leon Plaster
Jeffrey Britton John P. Herron Bertrand Poulin
Alvaro Cancino Tim Huff Subhash Sarkar
Donald Cash Rowland James H. Jin Sim
Mark Cheatham Matt Kennedy Henry Smith
Donald Chu Samer Khaleel Thomas Spitzer
Larry Coffeen Dong Kim Guiseppe Termine
John Crouse Alexander Kraetge Subhash Tuli
Joseph Cultrera Min Jea Lee Roger Verdolin
Alan Darwin Jim McBride David Vinson
Bob Degeneff Tony McGrail May Wang
Fred Elliot Peter J. McKemmy Barry Ward
Don Fallon Nigel P. McQuin Joe Watson
Jim Fyvie Harold Moore Peter Werelius
The following members of the balloting committee voted on this trial-use guide. Balloters may have voted
for approval, disapproval, or abstention.
(to be supplied by IEEE)
Copyright © 2006 IEEE. All rights reserved. iii

IEEE PC57.149/D6 April 2009
CONTENTS
1. Scope and Application ............................................................................... ¡Error! Marcador no definido.
1.1 Purpose ............................................................................................... ¡Error! Marcador no definido.

1.2 Definitions .......................................................................................... ¡Error! Marcador no definido.
2. FRA Measurement Overview ..................................................................................................................... 5
2.1 Use of FRA .......................................................................................................................................... 5

2.2 FRA Base Line Measurement .............................................................................................................. 6
2.3 FRA Diagnostic Application ............................................................................................................... 6
2.4 Recommended FRA measurement test parameters ............................................................................. 6
3. Making an FRA Measurement ................................................................................................................... 7
3.1 Test Procedures.................................................................................................................................... 7

3.2 Test Environment Preparation ............................................................................................................. 7
3.3 Test Object Preparation ....................................................................................................................... 7
3.4 Test Set ................................................................................................................................................ 8
3.5 Test Leads ............................................................................................................................................ 8
3.6 Measurement Types ............................................................................................................................. 9
3.7 LTC and DETC Positions ...................................................................................................................10
3.8 Test Connections ................................................................................................................................10
4. Test records ...............................................................................................................................................16
4.1 Introduction ........................................................................................ ¡Error! Marcador no definido.

4.2 Test Records ....................................................................................... ¡Error! Marcador no definido.
4.3 Conclusions ........................................................................................................................................18
5. Measurement Analysis and Interpretation .................................................................................................18
5.1 Introduction ........................................................................................................................................18

5.2 Trace Characteristics ..........................................................................................................................18
5.3 Trace Comparison...............................................................................................................................19
5.4 FRA Relationship to Other Transformer Diagnostics.........................................................................22
6. Appendix 1: FRA Theory ..........................................................................................................................32
Copyright © 2006 IEEE. All rights reserved. iv

Draft Trial-Use Guide for the Application and

Interpretation of Frequency Response Analysis
for Oil Immersed Transformers
1. Scope and Application

This trial use guide is applicable to the measurement of Frequency Response Analysis (FRA) of an oil
immersed power transformer. The document covers the requirements and specifications for
instrumentation, procedures for performing the tests, techniques for analyzing the data, and
recommendations for long-term storage of the data and results.
This trial use guide can be used in both field and factory applications.
1.1 Purpose
The FRA measurement is a comparative type test, where a transfer function measurement from one
winding circuit arrangement is compared to a previous measurement of the same winding arrangement.
This trial use guide will aid the user in the process to produce uniform and repeatable results that can be
successfully used and archived for analysis purposes for decades into the future.
1.2 Definitions
For purposes of this trial-use guide, the following terms and definitions apply. The Authoritative Dictionary
of IEEE Standard Terms [xxx] should be referenced for terms not defined in this clause
1.2.1 Frequency Response Analysis (FRA)
Frequency response analysis is a sensitive diagnostic technique for detecting changes in the electrical
characteristics of power transformer windings. Such changes can result from various types of electrical or
mechanical stresses (shipping damage, seismic forces, loss of clamping pressure, short circuit forces, etc.).
The test is non-destructive and non-intrusive and can be used either as a stand alone tool to detect winding
damage, or as a diagnostic tool to pinpoint damages detected in other tests (e.g. insulation power factor,
dissolved gas analysis, or short circuit impedance tests). FRA consists of measuring the admittance (current
divided by voltage) of the capacitive and inductive elements comprising the transformer windings. The
measurement is performed over a wide range of frequencies and the results are compared with a reference
“signature” or “fingerprint” of the winding to make a diagnosis.
Copyright © 2006 IEEE. All rights reserved. 1

1.2.2 FRA Magnitude
In a sinusoidal steady state analysis, the network function can be expressed as:
H(jω) = M(w)ejϕ(ω)
Where,
M(ω) = | H(j ω)| = | I(j ω)/ V(jω)|
A plot of this function (M(ω)) vs. frequency ω is called the FRA magnitude.
1.2.3 FRA Phase Angle
ϕ(ω) is the phase angle. A plot of this phase angle function vs. frequency ω, is called the FRA phase angle
function.
1.2.4 Sweep Frequency Method
In the sweep frequency method for making FRA measurements, the wide range of required frequencies is
generated via a sweep of individual sinusoidal signals injected into one terminal. The magnitude of the
excitation source remains constant for all frequencies used for the test.
1.2.5 Impulse Voltage Method
In the impulse voltage method for making FRA measurements, the wide range of required frequencies is
generated via one or more voltage impulses injected into one terminal. If more than one impulse is used,
the wave shapes are different so as to provide a more uniform spectral density for calculating the results.
1.2.6 Open-Circuit Self Admittance
Open-circuit configurations are used to describe measurements made on a winding with all other windings
complete and floating.
1.2.7 Short-Circuit Self Admittance
Short-circuit Self Admittance are used to describe measurements made on a winding with one or more
windings shorted.
1.2.8 Winding Self Admittance
When a transformer is subjected to winding self admittance testing, the leads are configured in such a
manner that four terminals are used. These four terminals can be divided into two unique pairs, one pair for
the input and the other pair for the output. These terminals can be modeled in a two-terminal pair or a two-
port network configuration. Figure 1 illustrates a two-port network.

I1 I2
V1 V2
Linear Circuit
Figure 1 – Two-Port Network
 I 1  Y11 Y12  V1 

 I  = Y ⋅
Y22  V2 
 2   21
In the symmetric [Y] matrix, Yii is the self-admittance of the node i (the sum of all the admittances
connected to node i). In practice this is the admittance measured by applying a voltage to one end of a
winding and measuring the current through the other end of the winding.
1.2.9 Inter-winding Admittance
Yij is the inter-winding or coupling admittance between nodes i and j. In practice this is the admittance
measured by applying a voltage to one end of a winding and measuring the current through one of the
terminals of another winding.
1.2.10 FRA Resonance Frequency
The term FRA Resonance Frequency is generally used to describe admittance maxima appearing in the
magnitude of the frequency response function of a transformer, accompanied by a zero value appearing in
the phase angle of the frequency response function. In practice, a power transformer is represented by a
complex, distributed RLC circuit, which may include several FRA Resonance Frequencies over a given
frequency range. Admittance maxima occur at frequencies where the inductive and capacitive reactive
impedance elements comprising the equivalent circuit are equal in magnitude, thereby resulting in zero net
reactive impedance as viewed from the terminals of a series RLC circuit, or alternatively as an infinite net
reactive impedance as viewed from the terminals of a parallel RLC circuit. At an FRA Resonance
Frequency, the complex admittance of the test circuit becomes a purely real value, representing the
equivalent resistance of the circuit as viewed from the test terminals. The number of FRA Resonance
Frequencies occurring over a given frequency range depends on the design and construction of the
transformer.
1.2.11 Test Specimen
Test specimen is the particular winding or winding segment being tested
1.2.12 Equipment Ground

This is the safety ground connection to the conductive, non current-carrying parts of electrical equipment.
For a transformer, it usually consists of one or more ground connections between the tank and station
grounds. It may also include connections to neutral terminals.
1.2.13 Measurement Ground
Is the reference connection for the FRA measurement, typically the ground connection between voltage
measurement cables and the current measuring cables.
1.2.14 Noise
These are unwanted disturbances that may be superimposed upon a useful (desired) signal. Noise tends to
obscure the information content of the useful signal. Random noise is the part of the noise signal that is
generated in an unpredictable fashion. Noise can be classified into two categories; white (or broadband)
noise and 1/f (or low frequency) noise. It can be introduced into a measurement through one of the
following vehicles:
• Generated by the measurement instrument and coupled directly.

• Stray electrostatic and electromagnetic fields.
• Connection characteristics of the leads..
Common noise sources encountered in FRA measurements may include power frequency and harmonic
noise, power line carrier, broadcast and communication signals, atmospheric disturbances and electrical
equipment disturbances (e.g. arcing or sparking).
1.2.15 Mechanical Movement
Detecting mechanical movement damage to transformer windings is one of the main interests of FRA test
measurement. Mechanical movement refers to the actual movement of transformer parts (coils, core, leads,
or accessories) with respect to each other or to ground in such a manner as to change the internal
inductances or capacitances of the test specimen. This may be caused by seismic or shipping forces or by
in-service conditions such as through-faults, load currents, mechanical breakdown of components, or
failures.
1.2.16 Amplitude Displacement
Amplitude displacement is the ratio between the amplitude function of a previous FRA “fingerprint”
measurement (e.g. baseline measurement at the factory, at an earlier date in the substation or before a short-
circuit test) and a new measurement (e.g. after transformer relocation, after suspected damage or after
short-circuit test). The ratio can also be between amplitude measurements on two different phases of the
same transformer or between a transformer and a duplicate or near-duplicate transformer.
1.2.17 Phase Displacement
Phase displacement is the ratio between the phase function of a previous FRA “fingerprint” measurement
(e.g. baseline measurement at the factory, at an earlier date in the substation or before a short-circuit test)
and a new measurement (e.g. after transformer relocation, after suspected damage or after short-circuit
test). The ratio can also be between phase measurements on two different phases of the same transformer or
between a transformer and a duplicate or near-duplicate transformer.
1.2.18 Frequency Displacement
Frequency displacement is the frequency shift of the recognizable areas of the FRA wave shape, most
notably the resonant frequency points, between the amplitude or phase measurement of the test specimen
and the comparison measurement.

1.2.19 Base Line Measurement
A Baseline Measurement is done to provide a set of FRA waveforms for future comparative purposes for
investigative or diagnostic analysis. A FRA baseline measurement can be produced in the factory when the
transformer has been filled with oil and dressed as part of factory commissioning tests, or at an earlier date
in the substation. Baseline or previous measurements should be repeatable. If internal movement or change
does not occur within a transformer, comparisons of matched traces should overlay well.
1.2.20 Minor Deviation
A change in amplitude, phase angle, or frequency displacement that is considered to be within the normal
deviation for a test configuration.
1.2.21 Significant Deviation
A change in amplitude, phase angle, or frequency displacement that is considered to be outside the normal
deviation for a test configuration. This may warrant further investigation or be considered as diagnostic
evidence of change in the internal configuration of a transformer.
2. FRA Measurement Overview

The FRA measurement provides diagnostic information, in the form of a transfer function, related to the
RLC network of the specimen under test. The RLC network is intimately related to the physical geometry
and construction of the test specimen.
Physical changes within the test specimen alter the RLC network, and in turn can alter the transfer function.
The transfer function behavior can reveal a wide range of mechanical or electrical changes in the test
specimen. Different transformer failure modes may have different effects on the transfer function, or may
not have a recognizable affect on the transfer function at all.
FRA can often detect gross transformer defects, as can other electrical tests. However, a primary benefit of
FRA is the potential for detection of minor deviations that might be related to the mechanical or electrical
integrity of the transformer, that are not apparent with other electrical tests.
2.1 Use of FRA
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. 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)
• System Modeling Purposes
There are two distinct categories for application of FRA measurement: baseline measurement and
diagnostic measurement. In both cases the procedures and precautions used to generate a good
measurement are the same. However, there is a difference in motivation for the tests in each category.

2.2 FRA Base Line Measurement
The base line FRA measurement may be done either in the factory or the field, and it provides information
that can be used for some future need. There are several distinct reasons to generate base line FRA
measurements:
• Quality assurance
• Short-circuit testing
• Customer Specification
• Transportation diagnostics prior to relocation and commissioning
• To provide a standard of comparison for future diagnostic FRA measurements
Important factors to consider when performing baseline FRA measurements include determining the
necessary tests and connections that might later be needed for diagnostic purposes, documentation of
methods and connections, archiving data, verification of results, and repeatability of the methods and
results. This trial-use guide provides assistance in each of these areas.
2.3 FRA Diagnostic Application
There are also several distinct reasons to generate diagnostic FRA measurements within a factory or field
environment:
• After factory short-circuit testing

• Relocation and commissioning validation
• Post incident: lightning, external through-fault, internal short circuit, seismic event etc
• Routine diagnostic purposes
• Condition assessment of older transformers
• Evaluation of used or spare transformers
Important factors to consider when performing diagnostic FRA measurements include matching the set up
and instrumentation parameters used for the baseline measurements. When baseline data is not available,
then data on duplicate transformers or other identical phases of a three-phase transformer may be used.
Typical data from other transformers of the same type may also be helpful for comparison. Special methods
or preparation may be needed in certain field applications due to aging of the equipment and connections,
field applied treatment to bushings, modification to the transformer since the baseline measurements were
made, or problems in making good ground connections due to field painted surfaces. This trial-use guide
provides assistance in these areas.
2.4 Recommended FRA Measurement Test Parameters
Test equipment must produce a frequency response measurement with the following characteristics:
• The test shall be made over a frequency range so as to be able to diagnose problems in the core,
clamping structure, windings and interconnections. The measurement of the ‘total’ frequency
response of the transformer is preferred.
• Successive measurements must have adequate resolution to give unambiguous diagnosis.
The test equipment should have the following attributes:
• Calibrated to an acceptable standard.

• The output power of the excitation source should provide adequate power over the entire
frequency range to allow for consistent measurement of the transfer function across the frequency
range.
• The test set shall be capable of measuring sufficient dynamic range, over the frequency range in
order to accommodate all transformer test objects.(> 60 dB)
• The test system (set and leads) should provide a known and constant characteristics impedance.
• A three lead system, signal, reference and test, should be used to reduce effect of leads in the
measurement.
• Test leads should be coaxial cables of the same length, within 1 cm, and less then 30 m (100 ft)
long. Shielded test leads should have the ability to be grounded at either end.
• Both the Magnitude and Phase of the calculated transfer function should be presented.
3. Making an FRA Measurement
3.1 Test Procedures
As with any electrical test, making a frequency response measurement should be done in a safe and
controlled manner irrespective of test location. Considerations for electrical safety in testing apply not only
to personnel, but also to the transformer and test equipment. Prior to testing, involved personnel should
discuss the test procedure and environment for ensuring that the work to be performed and any safety
precautions are clearly understood. Other safety aspects are covered in industry standards, company or
local regulations and manufacturer’s instruction manual.
3.2 Test Environment Preparation
• Any transformer under test should be completely isolated from any high voltage source or power
system source.
• The transformer tank should be grounded.
• During the test, there should be strict adherence to local safety regulations and guidelines.
3.3 Test Object Preparation
It is recommended that the transformer be in as close to ‘in service’ condition as possible.
All external bushing connections should be disconnected. This includes phase connections, neutral
connections, and tertiary grounds.
Whenever possible, all test lead connections should be made directly to the bushing terminals. Any extra
conductor length that is included in the test circuit path will influence the FRA test result. Short lengths of
bus bar attached to the transformer will not appreciably influence the measurement, as long as the test leads
are connected directly to the bushing terminals after the attached bus bar, so that the bus bar is not part of
the test circuit.
In instances where it is impossible to connect directly to a transformer bushing and a short length of bus bar
must be included as part of the test circuit, it is possible to perform frequency response measurements with
a short section of bus bar attached. This will affect the results but may be acceptable as a test technique
where it is impossible to exclude such short lengths from the circuit. Examples include rigid connections in
confined work spaces. It is important to note the state of the transformer under test so as to provide a

consistent method of testing. Where a transformer in the field has been tested previously with small lengths
of bus bar attached, it should be tested in the same way subsequently, if a comparison to historical data is
necessary. Analysis of results must take in to account possible variations that may be caused by
connections and their supports. As a general guideline, external bus bar connections should be avoided.
Special consideration should be given to safety when testing a transformer without oil so that excessive
voltages are not applied or induced in a combustible environment. The results of frequency response
measurements differ as a consequence of removing the oil. Testing with oil is the most common and
preferred method for frequency response analysis.
Where a de-energized tap changer is fitted, it is a decision for the transformer owners as to whether they
wish to operate the de-energized tap changer.
3.4 Test Set
The test set should be grounded according to the recommendations of the test equipment manufacturer, or
to the same point as the transformer under test, in the absence of equipment manufacturer’s
recommendations. Generally, the transformer tank header shall be considered as reference potential for the
FRA measurement. It should be noted that in all FRA measurements, the grounding techniques will have a
significant effect on test results. Grounding techniques, including selection of ground conductors as well as
their routings, should therefore be precise, repeatable, and documented.
The test equipment should always be within the recommended calibration interval. When possible prior to
use, a self-check of the operation of the test equipment using a standard test object with a known FRA
response may be employed as a means of assuring correct operation of the equipment. This check is
especially valuable for checking FRA test equipment, since there is generally no intuitive way of knowing
if the test equipment is giving correct results when making field measurements.
3.5 Test Leads
Three leads should be used:
• Excitation “source”
• Specimen Input “reference”
• Specimen Output “measure”
These should ideally be the same lengths. As a minimum, the “reference” and “measure” leads should be
identical.
Test leads should be checked for continuity and integrity before use. The best means for checking lead
integrity is to perform the FRA self-check using a standard test object, as discussed in Section 3.4.
Where shorting leads are used as part of a test set up between bushing terminals, these should be insulated
from ground, and be as short as possible. The impedance of these leads will influence the test results.
Therefore, when the test procedure requires shorting of terminals, selection of shorting conductors as well
as their routings should be precise, and repeatable, and documented.
Where local recommendations and/or guidelines require test grounds be applied to separate windings not
under test, these grounds should be as short as possible and connected to the same ground as the
transformer. It should be recognized that while the FRA response in not invalidated by the presence of
additional winding grounds, the response with these grounds in place may be unique, and should not be
compared with previous FRA test results obtained without the grounds installed. For the test to yield
maximum value, every effort should be made to configure the test object exactly as recommended by the
test equipment manufacturer. If necessary, requests may be made to the appropriate authority whenever it is

deemed necessary to temporarily lift ground connections to separate windings, as long as the transformer is
fully isolated from other power sources, and no hazards to safety are generated by the proximity of the
transformer terminals to other energized substation equipment. In all cases, special permission must be
received from the appropriate authority to deviate from any local recommendations and/or guidelines.
General lead connection diagrams are shown below in Figure 2, which provide examples of a typical test
setup.
Figure 1 — General Lead Connection Diagram
3.6 Measurement Types
Open Circuit Self Admittance (OC) – An open circuit self admittance measurement is made from one end
of a winding to another with all other terminals floating. The Open Circuit test can be applied to both single
phase and three phase transformers. Open Circuit tests generally fall into 5 winding categories: High
Voltage, Low Voltage, Tertiary, Series, and Common. The Series and Common categories are applied to
auto-transformers.
Open Circuit Self Admittance tests are primarily influenced by the core properties at or around the
fundamental power frequency. The Open Circuit tests can be used in conjunction with exciting current tests
in determining failure modes that affect the magnetic circuit of the transformer.
Short Circuit Self Admittance (SC) - The short circuit self admittance measurement is made from one end
of a high voltage winding to another while the low voltage winding is shorted. For repeatability purposes, it
is recommended that all low voltage windings are shorted on three phase transformers to create a three
phase equivalent short circuit model. This ensures all three phase are similarly shorted to give a consistent
impedance. Any available neutral connections should not be included in the shorting process.
The Short Circuit Self Admittance test isolates the winding impedance from the core effects properties at or
around the fundamental power frequency. The Short Circuit results should produce similar and comparable
diagnostic information as seen in both leakage reactance and DC winding resistance measurements.
Inter-Winding (IW) – An Inter-Winding measurement is made from one end of a winding and measuring
the signal through one of the terminals of another winding, with all other terminals floating. Inter-Winding
measurements are capacitive in nature. These measurements exhibit a high impedance at low frequencies
(<100 Hz); the impedance generally decreases as frequency increases. This would include, for example, H1
to X1 on a double wound transformer, or H1 to Y1 on an autotransformer with a tertiary. Note that H1 to
X1 on an autotransformer is not an Inter-Winding measurement but an open circuit measurement on the
series winding.

Transfer Admittance (TA) - A Transfer Admittance measurement is made from one end of a winding to
another with the other ends of the windings under test grounded. All other winding terminals not under test
should remain floating. The Transfer Admittance test most closely resembles the turns ratio test properties
at or around the fundamental power frequency.
3.7 LTC and DETC Positions
The tap positions shall be noted on the test report for each test. Generally tests on windings with a LTC
should be in the tap combination that places all sections of the tap windings in the circuit. As a convention,
it is recommended that the LTC be in the extreme raise position.
However, when tests are performed at the LTC neutral tap position, the previous tap position should also be
recorded, as the resulting winding connection through the reversing switch will influence the test result.
Consistency with previous test applications is key for subsequent analysis. It is recommended that neutral
tap position measurements be made after arriving from the raised tap positions.
It is recommended that the DETC be in the position that is dictated by in service conditions.
Transformers in service occasionally have problems due to DETC movement; it is not recommended that
the DETC position be altered for a FRA test. The exception is in factory tests on a new transformer, where
it can be assumed that the DETC is in satisfactory condition.
LTC and DETC SUMMARY
• LTC – Extreme Raise

• DETC – As Found
3.8 Test Connections
The test connections described here do not include repeat tests for different tap positions. Alternative test
sequences suggested below shall be selected based on the recommendations of the test equipment
manufacturer, and the type of test voltage applied. Bushings not under test, including neutrals, shall be
ungrounded, unless grounding is required under local recommendations and/or guidelines.
All windings should be tested as shown below in the test connection tables. The tables are as follows:
TABLE 1 - TWO WINDING TRANSFORMERS

TABLE 2 - AUTOTRANSFORMER W/O TERTIARY
TABLE 3 - AUTOTRANSFORMER WITH TERTIARY
TABLE 4 - AUTOTRANSFORMER WITH BURIED TERTIARY
TABLE 5 - THREE WINDING TRANSFORMER PART 1
TABLE 6 - THREE WINDING TRANSFORMER PART 2
NOTE: Test connection variations on the basis test sequence and polarity are acceptable if specified by the
end user.

TABLE 1 - TWO WINDING TRANSFORMERS - 15 TESTS
Test Type Test # 3φ ∆-Υ 3φ Υ-∆ 3φ ∆-∆ 3φ Υ-Υ 1φ

Group 2 Group 2 Group 1 Group 1
θ ⇒ 30° θ ⇒ 30° θ ⇒ 0° θ ⇒ 0°
LAG LAG
HV Open Circuit (OC) Test 1 H1-H3 H1-H0 H1-H3 H1-H0 H1-H2
Self Admittance Test 2 H2-H1 H2-H0 H2-H1 H2-H0 (H1-H0)
Terminals Floating Test 3 H3-H2 H3-H0 H3-H2 H3-H0
LV Open Circuit (OC) Test 4 X1-X0 X1-X2 X1-X3 X1-X0 X1-X2
Self Admittance Test 5 X2-X0 X2-X3 X2-X1 X2-X0 (X1-X0)
Terminals Floating Test 6 X3-X0 X3-X1 X3-X2 X3-X0
Short Circuit (SC) Test 7 H1-H3 H1-H0 H1-H3 H1-H0 H1-H0
Self Admittance Test 8 H2-H1 H2-H0 H2-H1 H2-H0 Short
Short [X1-X2-X3]* Test 9 H3-H2 H3-H0 H3-H2 H3-H0 [X1-X2]*
Inter-Winding (IW) Test 10 H1-X1 H1-X1 H1-X1 H1-X1 H1-X1
Inter-Winding Admittance Test 11 H2-X2 H2-X2 H2-X2 H2-X2
All Terminals Float Test 12 H3-X3 H3-X3 H3-X3 H3-X3
Transfer Admittance (TA) Test 13 H1-X1 H1-X1 H1-X1 H1-X1 H1-X1
High (H) to Low (L) Test 14 H2-X2 H2-X2 H2-X2 H2-X2 Ground
Ground (H- and X-) Test 15 H3-X3 H3-X3 H3-X3 H3-X3 [H2, X2]
*Indicates short circuit tests: terminals are shorted together. The neutral is not included for 3φ Wye
connections, but may be included for 1φ test connections.
TABLE 2 - AUTOTRANSFORMER W/O TERTIARY - 12 TESTS
Test Type Test # 3φ 1φ

Series Winding (OC) Test 1 H1-X1 H1-X1
All Other Terminals Floating Test 2 H2-X2
Test 3 H3-X3
Common Winding (OC) Test 4 X1-H0X0 X1-H0X0
All Other Terminals Floating Test 5 X2-H0X0
Test 6 X3-H0X0
Short Circuit (SC) Test 7 H1-H0X0 H1-H0X0
High (H) to Low (L) Test 8 H2-H0X0 Short
Short [X1-X2-X3]* Test 9 H3-H0X0 [X1-H0X0]*
Transfer Admittance (TA) Test 10 H1-X1 H1-X1
High (H) to Low (L) Test 11 H2-X2 Ground
Ground (H0X0) Test 12 H3-X3 [H0X0]
* Indicates short circuit tests: terminals are shorted together. The neutral is not included for 3φ Wye

TABLE 3 - AUTOTRANSFORMER WITH TERTIARY - 33 TESTS

Test 3 H3-X3
Test 6 X3-H0X0
Tertiary Winding (OC) Test 7 Y1-Y3 Y1-Y2
All Other Terminals Floating Test 8 Y2-Y1 (Y1-Y0)
Test 9 Y3-Y2
High (H) to Tertiary (Y) Test 14 H2-H0X0 Short
Short [Y1-Y2-Y3]* Test 15 H3-H0X0 [Y1-Y2]*
Short Circuit (SC) Test 16 X1-H0X0 X1-H0X0
Low (L) to Tertiary (Y) Test 17 X2-H0X0 Short
Short [Y1-Y2-Y3]* Test 18 X3-H0X0 [Y1-Y2]*
Inter-Winding (IW) Test 19 H1-Y1 H1-Y1
High (H) to Tertiary (Y) Test 20 H2-Y2
All Terminals Float Test 21 H3-Y3
Inter-Winding (IW) Test 22 X1-Y1 X1-Y1
Low (L) to Tertiary (Y) Test 23 X2-Y2
All Terminals Float Test 24 X3-Y3
Transfer Admittance (TA) Test 28 H1-Y1 H1-Y1
High (H) to Tertiary (Y) Test 29 H2-Y2 Ground
Ground (H0X0 and Y-) Test 30 H3-Y3 [H0X0, Y2]
Transfer Admittance (TA) Test 31 X1-Y1 X1-Y1
Low (L) to Tertiary (Y) Test 32 X2-Y2 Ground
Ground (H0X0 and Y-) Test 33 X3-Y3 [H0X0, Y2]

TABLE 4 - AUTOTRANSFORMER WITH BURIED TERTIARY - 18 TESTS

Test 3 H3-X3
Test 6 X3-H0X0
Inter-Winding (IW) Test 10 H1-Y1 H1-Y1
High (H) to Tertiary (Y) Test 11 H2-Y1
All Terminals Float Test 12 H3-Y1
Inter-Winding (IW) Test 13 X1-Y1 X1-Y1
Low (L) to Tertiary (Y) Test 14 X2-Y1
All Terminals Float Test 15 X3-Y1

TABLE 5 - THREE WINDING TRANSFORMER PART 1: - 36 TESTS

Test Type Test # 3φ 3φ 3φ 3φ 1φ
∆-∆-∆ ∆-∆-Υ ∆-Υ-∆ ∆-Υ-Υ
θ ⇒ 0° θ ⇒ 30° θ ⇒ 30° θ ⇒ 30°
LAG LAG LAG
HV Open Circuit (OC) Test 1 H1-H3 H1-H3 H1-H3 H1-H3 H1-H2
All Other Terminals Floating Test 2 H2-H1 H2-H1 H2-H1 H2-H1 (H1-H0)
Test 3 H3-H2 H3-H2 H3-H2 H3-H2
LV Open Circuit (OC) Test 4 X1-X3 X1-X3 X1-X0 X1-X0 X1-X2
All Other Terminals Floating Test 5 X2-X1 X2-X1 X2-X0 X2-X0 (X1-X0)
Test 6 X3-X2 X3-X2 X3-X0 X3-X0
Tert Open Circuit (OC) Test 7 Y1-Y3 Y1-Y0 Y1-Y3 Y1-Y0 Y1-Y2
All Other Terminals Floating Test 8 Y2-Y1 Y2-Y0 Y2-Y1 Y2-Y0 (Y1-Y0)
Test 9 Y3-Y2 Y3-Y0 Y3-Y2 Y3-Y0
High (H) to Low (L) Test 11 H2-H1 H2-H1 H2-H1 H2-H1 Short
Short [X1-X2-X3]* Test 12 H3-H2 H3-H2 H3-H2 H3-H2 [X1-X2]*
High (H) to Tertiary (T) Test 14 H2-H1 H2-H1 H2-H1 H2-H1 Short
Short [Y1-Y2-Y3]* Test 15 H3-H2 H3-H2 H3-H2 H3-H2 [Y1-Y2]*
Short Circuit (SC) Test 16 X1-X3 X1-X3 X1-X0 X1-X0 X1-X0
Low (L) to Tertiary (T) Test 17 X2-X1 X2-X1 X2-X0 X2-X0 Short
Short [Y1-Y2-Y3]* Test 18 X3-X2 X3-X2 X3-X0 X3-X0 [Y1-Y2]*
Inter-Winding (IW) Test 19 H1-X1 H1-X1 H1-X1 H1-X1 H1-X1
High (H) to Low (L) Test 20 H2-X2 H2-X2 H2-X2 H2-X2
Inter-Winding (IW) Test 22 H1-Y1 H1-Y1 H1-Y1 H1-Y1 H1-Y1
High (H) to Tertiary (T) Test 23 H2-Y2 H2-Y2 H2-Y2 H2-Y2
All Terminals Float Test 24 H3-Y3 H3-Y3 H3-Y3 H3-Y3
Inter-Winding (IW) Test 25 X1-Y1 X1-Y1 X1-Y1 X1-Y1 X1-Y1
Low (L) to Tertiary (T) Test 26 X2-Y2 X2-Y2 X2-Y2 X2-Y2
All Terminals Float Test 27 X3-Y3 X3-Y3 X3-Y3 X3-Y3
Transfer Admittance (TA) Test 28 H1-X1 H1-X1 H1-X1 H1-X1 H1-X1
High (H) to Low (L) Test 29 H2-X2 H2-X2 H2-X2 H2-X2 Ground
Ground (H- and X-) Test 30 H3-X3 H3-X3 H3-X3 H3-X3 [H2, X2]
Transfer Admittance (TA) Test 31 H1-Y1 H1-Y1 H1-Y1 H1-Y1 H1-Y1
High (H) to Tertiary (T) Test 32 H2-Y2 H2-Y2 H2-Y2 H2-Y2 Ground
Ground (H- and Y-) Test 33 H3-Y3 H3-Y3 H3-Y3 H3-Y3 [H2, Y2]
Transfer Admittance (TA) Test 34 X1-Y1 X1-Y1 X1-Y1 X1-Y1 X1-Y1
Low (L) to Tertiary (T) Test 35 X2-Y2 X2-Y2 X2-Y2 X2-Y2 Ground
Ground (X- and Y-) Test 36 X3-Y3 X3-Y3 X3-Y3 X3-Y3 [X2, Y2]

TABLE 6 - THREE WINDING TRANSFORMER PART 2 - 36 TESTS

Test Type Test # 3φ 3φ 3φ 3φ
Υ-Υ-Υ Υ-Υ-∆ Υ-∆-Υ Υ-∆-∆
θ ⇒ 0° θ ⇒ 30° θ ⇒ 30° θ ⇒ 30°
LAG LAG LAG
HV Open Circuit (OC) Test 1 H1-H0 H1-H0 H1-H0 H1-H0
All Other Terminals Floating Test 2 H2-H0 H2-H0 H2-H0 H2-H0
Test 3 H3-H0 H3-H0 H3-H0 H3-H0
LV Open Circuit (OC) Test 4 X1-X0 X1-X0 X1-X3 X1-X3
All Other Terminals Floating Test 5 X2-H0 X2-H0 X2-X1 X2-X1
Test 6 X3-H0 X3-H0 X3-X2 X3-X2
Tert Open Circuit (OC) Test 7 Y1-Y0 Y1-Y3 Y1-Y0 Y1-Y3
All Other Terminals Floating Test 8 Y2-Y0 Y2-Y1 Y2-Y0 Y2-Y1
Test 9 Y3-Y0 Y3-Y2 Y3-Y0 Y3-Y2
Short Circuit (SC) Test 10 H1-H0 H1-H0 H1-H0 H1-H0
High (H) to Low (L) Test 11 H2-H0 H2-H0 H2-H0 H2-H0
Short [X1-X2-X3]* Test 12 H3-H0 H3-H0 H3-H0 H3-H0
Short Circuit (SC) Test 13 H1-H0 H1-H0 H1-H0 H1-H0
High (H) to Tertiary (T) Test 14 H2-H0 H2-H0 H2-H0 H2-H0
Short [Y1-Y2-Y3]* Test 15 H3-H0 H3-H0 H3-H0 H3-H0
Short Circuit (SC) Test 16 X1-X0 X1-X0 X1-X3 X1-X3
Low (L) to Tertiary (T) Test 17 X2-X0 X2-X0 X2-X1 X2-X1
Short [Y1-Y2-Y3]* Test 18 X3-X0 X3-X0 X3-X2 X3-X2
Inter-Winding (IW) Test 19 H1-X1 H1-X1 H1-X1 H1-X1
Inter-Winding (IW) Test 22 H1-Y1 H1-Y1 H1-Y1 H1-Y1
All Terminals Float Test 24 H3-Y3 H3-Y3 H3-Y3 H3-Y3
Inter-Winding (IW) Test 25 X1-Y1 X1-Y1 X1-Y1 X1-Y1
All Terminals Float Test 27 X3-Y3 X3-Y3 X3-Y3 X3-Y3
Transfer Admittance (TA) Test 28 H1-X1 H1-X1 H1-X1 H1-X1
Ground (H- and X-) Test 30 H3-X3 H3-X3 H3-X3 H3-X3
Transfer Admittance (TA) Test 31 H1-Y1 H1-Y1 H1-Y1 H1-Y1
Ground (H- and Y-) Test 33 H3-Y3 H3-Y3 H3-Y3 H3-Y3
Transfer Admittance (TA) Test 34 X1-Y1 X1-Y1 X1-Y1 X1-Y1
Ground (X- and Y-) Test 36 X3-Y3 X3-Y3 X3-Y3 X3-Y3
* Indicates short circuit tests: terminals are shorted together. The neutral is not included for Wye
connections.
Where transformer winding configurations are required which are not covered in the above tables, please
refer to the transformer nameplate. The configuration vectors will determine the test procedure.

4. Test records
4.1 Introduction
The Frequency Response Measurement is one that provides a wealth of information regarding the internal
geometry of a transformer.
To gain value from the measurement, it is important that measurements are made in a consistent manner.
Unlike, say, power factor measurements where it is possible to make consistent measurements based on an
analysis of the transformer design and some basic electrical engineering ‘first principals’, FRA requires
measurements to be made consistent with previous measurements or with those on similar units.
Accordingly, it is important that any test parameter that could affect test results is recorded in a clear,
concise, format such that the FRA signatures can be reconstructed without the use of proprietary software.
4.2 Test Records
A number of different elements of a particular test need to be recorded – these ensure that the same test set
up is possible on subsequent tests, and with respect to tests on other transformers.
The details given here are separated out into separate elements of a test set up, but should not be taken as a
prescription of how data should be stored.
In each area, the suggested minimum information required is marked as ‘required’; some other data can be
inferred from asset management systems within the transformer owner’s organization.
4.2.1 Data Format
Data should be viewable in most standard applications to allow flexible analysis of the results, comparison
between results from different test systems, and inclusion in reports. Because FRA records can be stored
for many years before being recalled for comparison, it is recommended to avoid any file format associated
with current applications or instruments, as these can change with time or might not be supported 15 or 20
years after they have been stored.
4.2.2 Data Elements
The object of a test is a transformer. Clearly, nameplate data for the transformer must be recorded, but there
are other factors that may vary between tests, which may impact results. These include:
a) Transformer: what is tested, nameplate data

b) Location: for example, a transformer may be tested before and after relocation
c) Test Equipment: especially if we are comparing data from a number of different OEM sets
d) Test Organization: to detect any systematic errors, which may result from poor procedures
e) Test Setup: state of transformer, oil level, bushings, tap position, temperature, etc
f) Test Results: date, time, frequency, magnitude and phase
4.2.3 Transformer Data
This is the ‘static’ data which does not vary between tests on a particular transformer:
i. Manufacturer – Required

ii. Year of manufacture – Required

iii. Serial Number – Required
iv. MVA under various cooling regimes (ONAN, OFAF, ODAF, etc) – Required
v. Voltage for: HV, LV1, LV2, Tertiary – Required
vi. Current
vii. Special ID (transformer ID) – Required
viii. Number of Windings and Configuration – Required
ix. Number of Phases
x. Impedance – up to four fields, e.g. HV-LV1, HV-LV2, HV-Tertiary; LV-Tertiary
xi. Transformer Type – gsu, trans, dist, furnace, etc
xii. Transformer configuration –auto, double wound, etc. – Required
xiii. Transformer construction – core form, shell form, winding type, etc
xiv. LTC Fields: Manufacturer, Serial Number, Manufacture Year, Range, Notes
xv. DETC Fields: Manufacturer, Serial Number, Manufacture Year, Range, Notes
xvi. Free form data entry fields for user specific details
4.2.4 Location Data
A small amount of data is required to make sure that we can record where a transformer is – especially if it
is going to move between locations on one site.
i. Location – Required
ii. Special Id – Required
iii. Free form data entry fields for user specific details
4.2.5 Test Equipment Data
Recording the test data equipment is key to ensuring that good equipment is used for the measurement.
i. Equipment Name and Model number – Required

ii. Manufacturer
iii. Test Equipment Serial Number – Required
iv. Calibration Date – Required
v. Measurement Impedance
vi. Free form data entry fields for user specific details
4.2.6 Test Organization Data
This records who did the testing.
i. Company – Required
ii. Operator – Required
iii. Company Id
iv. Free form data entry fields for user specific details
4.2.7 Test Set Up Data
These are elements that may vary between tests on a given transformer on a given day.
i. LTC Position, including previous position if at neutral – Required

ii. DETC Position – Required
iii. Oil temperature - Required
iv. Oil level
v. Oil Status (whether immersed or not) – Required
vi. Bushings grounded – Required
vii. Bushings connected – Required

viii. External circuits connected to bushings (length of bus, etc.)

ix. Measurement Type (Open Circuit, Short Circuit)
x. Signal & Reference Leads location – Required
xi. Reason for Test: – Required
xii. Free form data entry fields for user specific details
xiii. Date and time of trace – Required
4.3 Conclusions
To extract value from a test, data needs to be recorded regarding a number of test elements.
Good records mean that a particular trace will be useful in future.
5. Measurement Analysis and Interpretation
5.1 Introduction
This section looks at the extraction of value from an FRA measurement. Two basic analysis strategies are
deployed in this section: identifying expected trace characteristics, and comparing a trace with reference
results. The primary goal of FRA analysis is to determine the physical condition of the transformer, thus
ensuring internal component shave not moved as a result of transportation, sever insulation degradation and
faults.
5.2 Trace Characteristics
Since transformer designs and applications vary, the FRA results inherit diverse properties and
characteristics. However, a FRA trace, over specified frequency ranges, has a degree of predictability; low
frequency core effects, main winding effects, and short circuit responses. These expectations can be used to
identify basic problems that may exist within a transformer. In addition, different winding configurations
generally exhibit distinct patterns due to their relationship between phases and the core. With and
understanding of the relationship between frequency range and transformer configuration, physical
deformations can be narrowed to specific sections of the transformer.
The different measurement types produce different characteristic trace wave shapes, but the expected
changes in the traces are generally similar. Some trace characteristics indicate more significant winding
deformation and insulation degradation than others. One of the most significant change indicators is the
presence of an additional resonance peak(s) in the magnitude trace or the loss of an existing resonance
peak(s). The next most significant change indicator is usually a shift in frequency for an existing resonance
peak(s). These are usually examples of winding deformation or insulation degradation. Another trace
characteristic indicator is when a trace continues to increase in magnitude difference from the original
trace, while maintaining a similar shape, as frequency increases. This can be an indicator of winding
looseness, especially for frequencies above 500 kHz. Most significant trace magnitude findings are also
accompanied by significant trace phase differences.
FRA test type greatly affects the expected trace characteristics. Specific characteristic can be identified
depending on the test type.
5.2.1 Open Circuit Test

The open circuit test is the most common FRA test performed. Similar to an excitation current test, each
winding is measured individually while all other terminals float. This test is most influenced by the effects
of the core, main windings, tap winding and tap leads. Depending on the winding configuration, the low
frequency section (core) will usually take on a distinct shape.
High voltage windings display the most distinguishable pattern in the low frequency region. Wye
connected windings often show a single distinct or two closely placed null resonance between 20Hz and
5kHz. Delta connected windings will usually show two further spaced apart resonance in this same
frequency range. Low or tertiary voltage windings tend to follow the same pattern, but can be heavily
influenced by the high voltage winding.
The mid frequency regions is dominated by the network impedances created by the winding stack itself.
There could be either a complex series or simple series of resonances created.
Above 500 kHz, the results will continue to be influenced by the main winding’s impedance network, but
tends to be more sensitive to smaller geometric sections such as tap windings and tap leads. ADD winding
looseness and insulation degradation
5.2.2 Short Circuit Self-Admittance
Short circuit tests are designed to allow for the inspection of the winding without the influence of the core.
The transformer core’s reluctance circuit influences the lowest frequencies. To analyze the bulk inductance
of the winding without the influence of the core, the secondary windings are short circuited. This
configuration is similar to a short circuit impedance test or leakage reactance test.
The short circuit test’s trace characteristic is similar to the open circuit test with the exception of the low
frequency region. At these low frequencies, the increased impedance associated with the core reluctance is
removed. This results in an overall increase in the FRA trace’s magnitude in this region as well a more
inductive phase angle. The shape in the low frequency is dominated by the first order roll-off characteristic
of the windings main inductance. If the results are examined on a Bode plot, a predicable inductive roll-off
section can often be identified as is expected with first-order systems.
5.2.3 Inter-Winding
Inter-winding tests measure the network impedance between two voltage class windings. Given that this
measurement does not offer a direct connection, as the two windings are isolated, this trace shows an
increasing magnitude as a function of frequency. At the lowest frequencies, the measurement is highly
capacitive as would be expected by the open circuit configuration between the two test terminals. As the
frequency is increased a series of resonances will be formed with an overall trend of increasing magnitude.
These resonances are formed by the network impedances formed between the two voltage class windings
and then by the inter-winding resonances.
5.2.4 Transfer Voltage
The transfer voltage test is designed to measure the voltage ratio between two voltage class windings. In
this test configuration the two vector matching terminals are measured across while grounding the
associated leg terminal on each side. This configuration is similar to a single phase turns ratio test.
The trace is dominated by a flat magnitude response at the lowest frequencies. The magnitude in the low
frequency region is equal to the voltage ratio formed by the voltage class ratio between the two windings
under test. At the higher frequencies, a series of resonances are formed that are generally not examined as
it’s not the primary focus of this test.
5.3 Trace Comparison

Trace comparison is the primary method for the analysis of FRA results. Comparisons can be made against
the baselines and previous data, sister unit results, and across phases. Traces can also be examined for
expected patterns as mentioned in section 5.2; this type of comparison requires more experience but can
prevent erroneous data from being collected. Assuming the test equipment is repeatable and there is not
residual magnetism, the initial expectation is that any data comparison should result in near perfect
overlays.
Various tools and algorithms are available for comparing and analyzing FRA measurements.
5.3.1 Plot Inspection
Plot inspection involves plotting the FRA transfer function and examining the results. This method of
analysis is the most common and relies upon a reasonable level of skill. Specialized computer software or
spreadsheets are commonly used to present data. Common practice is display the plot in decibels (dB)
versus logarithmic frequency. This type of plot is also known as a Bode plot. Bode plots allow easier
inspection of resonances as it simplifies the display of exponentially changing impedances and are used
cross-industry for the analysis of transfer functions. Plots can also be graphed as impedance or admittance
versus frequency; either linear or logarithmic scales can be applied.
The plots are inspected for expected general patterns due to transformer configuration and comparing the
results with other traces. The other traces can be different phases, sister units or baseline results. This
method relies upon careful scrutiny of the trace resonances and magnitude/phase deviations. Please refer to
section 5.2 for a discussion on frequency range to component correlation and expected patterns. Deviations
noted due to comparison with another unit or phase could indicate physical deformation of a transformer’s
component.
5.3.2 Difference Plotting
Difference plotting was one of the first methods of analysis used for transformer FRA tests. This type of
analysis takes two traces for comparison and subtracts one from the other. A difference traces is the result.
Places in the difference plot that deviate from zero (i.e. no variation between the two traces) could indicate
problems and would require greater scrutiny. Resonance shifts or deviation in FRA magnitude would show
the greatest variation on the plot. This method does not necessarily assist ascertaining the cause of the
problem, but can assist in identifying suspect frequency regions. Difference plotting is not use on phase
plots due to the sensitivity of the phase measurement.
5.3.3 Cross Correlation Coefficients
Cross correlation coefficients, or simply correlation, provide an indication of similarity between two traces.
Another common term for cross correlation is the Pearson’s Correlation Coefficient or Pearson’s Product
Moment Correlation. Correlation allows for the expression of how random (or similar) a range of numbers
is. Correlation has a value between -1 and 1. If the value is 1.0, this indicates complete randomness
between the two traces under inspection, The closer the correlation is to zero (0), the more similar the two
traces are. If the value is -1.0 this indicates a complete inverse relationship. The goal for analysis is to get a
number as close to zero as possible, indicating similar linear relationships between the two traces.
Individual frequency ranges can be analyzed for correlation by changing the frequency bounds of the
correlation algorithm. Cross correlation is a complex algorithm, and is often calculated with the help of
specialized FRA software or commercially available mathematical programs. Cross correlation coefficients
are calculated by determine the covariance of the two traces under test. Covariance is the expressed as σxy.
Covariance calculates how much two traces deviate from each other. This is expressed in equations (1) and
(2) below. Equation (3) shows how two traces’, X and Y, deviations are integrated over the entire
frequency spectrum. In the case of isolating certain regions, the bounds would be changed from plus or
minus infinity to the spectrum under question.

Finally the correlation constant ρ is calculated by taking the covariance of the two traces under question
and then dividing them by the product of each traces standard deviation, thus normalizing the constant for
easy inspection.
Experience indicates that various levels of expected comparison exist. The level of comparison can be
categorized by the following:
5.3.4 Baseline Data
Baseline or pervious data should be repeatable. If internal movement or change does not occur within test
specimen, the matched traces should overlay well. Matched traces are defined as FRA results obtain from
the same test terminals and transformer configurations. An example would be two scans collected from the
same winding, such as H1-H3, on different test dates.
Data is collected before and after transformer relocation is expected to overlay well. Any variance is such
comparisons indicate a problem. One exception, caused by the magnetic circuit, occurs at low frequencies
and should be considered during evaluation. Magnetization and temperature change can cause the lower
frequencies of the trace to be slightly offset in certain cases. If residual magnetism is known to be present,
less scrutiny should be placed on this comparison.
As stated in Section 3, it should be noted that the LTC and DETC position influences the results. If the test
results are obtained in different tap positions, expect variation.
5.3.5 Similar Units
Similar unit results are expected to compare well. Genuine similar units show very little difference between
matched scans. Genuine similar units are considered to be identical in manufacturer, design and
construction. Care must be taken to asses if a similar unit is genuine. Even if a transformer appears
identical in all aspects, it’s possible that the two units may be constructed with some differences. This is
most prevalent when a utility purchases two identical transformers several months or years apart. The
manufacture may have made modifications to the design that does not change the functionality of the
transformer, but could lead to variations in the traces.
All tests on similar units were conducted with the LTC and DETC in the same position. The low frequency
and mid frequency regions governed by the core’s magnetic circuit and main windings should have the
closest similarity. There may be variations in the higher frequencies do to subtle changes in tap lead
placement and routing. The more sister units that can be compared, a better feel for the overall stability of
their traces can be ascertained. As such, it’s beneficial to plot several sister units on the same graph for
inspection, such as three single phase transformers installations.
5.3.6 Phase

Phase comparisons are the most difficult and require the greatest amount of experience to conduct properly.
In addition, some transformer designs do not have completely symmetrical designs. In these cases, there
will always be some phase to phase variations. Because of this, it’s paramount that benchmark data be
obtained as early in a transformer’s life as possible. In most cases, different phases will overlays with
reasonable similarity
The center phase, especially in core type transformers, exhibits the most deviation when comparing all
three phases. Often, the two outer phases compare well. Different flux paths seen by each phase contribute
to the observed differences. The affects of the core are expected at the lower frequencies; however the core
influence can overlap into the higher frequency range.
The actual windings of a three phase transformer are almost identical, but the connection scheme between
phases is very different. As an example, the phases of a wye winding are all at different distances from the
neutral; LTC connections fall into the same category. Thus, since the winding are not equilaterally spaced,
the varying lead length entering and leaving the windings influence the individual transfer function of each
winding. Overall phase symmetry appears to be a function of the overall physical size and complexity.
However, when no baseline test exists, and there is no genuine sister unit available to compare against, it is
still possible to determine obvious failure modes by cross phase comparison of the same voltage windings.
5.4 FRA Relationship to Other Transformer Diagnostics
The FRA results depending on the test connections can be used to confirm the results of other diagnostic
tests. These tests include:
• Single Phase Exciting Current

• Turns Ratio
• Short Circuit Impedance (Leakage Reactance)
• DC Winding Resistance
5.4.1 Single Phase Exciting Current
Single phase exciting Current test results can be compared the FRA’s low frequency region in the open
circuit test that is applied to the HV winding. The open circuit test is heavily influenced by the core
properties at or around the fundamental power frequency. The FRA results are expected to produce a
pattern similar to the Single Phase Exciting Currents test. When examining the magnitude plots in the low
frequency region, their decibel or ohm levels should mimic the pattern exhibited by the same connection
excitation current test at the excitation frequency.
5.4.2 Turns Ratio
The Transfer Voltage test most closely resembles the turns-ratio test properties at or around the
fundamental power frequency. The transfer function unit, dB or Ohms, can be interpreted to match the
turns ratio results. Several frequencies at or around the fundamental power frequency should be averaged to
estimate the turns ratio value.
5.4.3 Turns Ratio
The Transfer Voltage test most closely resembles the turns-ratio test properties at or around the
fundamental power frequency. The transfer function unit, dB or Ohms, can be interpreted to match the
turns ratio results. Several frequencies at or around the fundamental power frequency should be averaged to
produce the turns ratio value.
5.4.4 DC Winding Resistance

If the Short Circuit test produces a horizontal response at frequencies less than 30 Hz, then the FRA results
can be compared the DC Winding Resistance results. Any differences between phases at these low
frequencies should be checked with a DC Winding Resistance test.
5.5 Failure Modes
In general, the FRA test is sensitive to failure modes that cause geometric change(s) within a transformer.
In fact, the popularity of the FRA test has been driven by the desire to detect mechanical failures within a
transformer. Failure modes are not exclusive to geometric variations within a transformer and can include
variation in the core’s magnetic circuit and contact resistance.
FRA test variations can be cause by a single type of failure or a combination of two or more. Failure due to
faults creates high over-currents through the transformer. As a result, the transformer experiences strong
and often violent electromagnetic forces. These violent events can often lead to compounded failure modes.
These compounded events can complicate the FRA analysis but often helps better understand the condition
of the transformer. Below are listed known failure modes and an explanation of each.
• Radial “Hoop Buckling” Deformation of Winding

• Axial Winding Elongation “Telescoping”
• Overall- Bulk & Localized Movement
• Core Defects
• Contact Resistance
• Winding Turn-to-Turn Short Circuit
• Open Circuited Winding
• Loss of Clamping Pressure
• Residual Magnetization
5.5.1 Radial Winding Deformation – “Hoop Buckling”
Radial winding deformation or “Hoop Buckling” is a winding compressive failure that is characterized by a
pronounced change to the windings radial geometry. This type of failure can result form the high current
electromagnetic forces caused by high over current faults. The winding is subjected to high radial
compressive forces and will end up “buckling” the winding along the entire length. The forces are
concentrated toward the inner winding. Radial winding deformation occurs in two forms, free and forced.
Radial winding deformation affects the FRA measurements as follows:
Table 1 – Radial Winding Deformation

Frequency Range Radial Winding Deformation
Assuming, no other failure modes exist:
20 Hz – 10 kHz Open Circuit Tests:

This region (core region) is generally unaffected during radial winding deformation.
Short Circuit Tests:
Results in an increase in impedance. The FRA trace for the affected phase generally
exhibits slight attenuation within the inductive roll-off portion.
5 kHz – 100 kHz Open Circuit and Short Circuit Tests:
The bulk winding range can shift or produce new resonance peaks and valleys
depending of the severity of the deformation. However, this change is minimal and
difficult identify. The changes will be greater on the affect winding, but it is still
possible to have the effects transferred to the opposing winding. The response in the
bulk region should be used as secondary evidence to support the analysis.
50 kHz – 1 MHz Open Circuit and Short Circuit Tests:
Radial winding deformation is most obvious in this range. It can shift or produce

new resonance peaks and valleys depending of the severity of the deformation. The
changes will be greater on the affect winding, but it is still possible to have the
effects transferred to the opposing winding.
> 1 MHz Open Circuit and Short Circuit Tests:
This range is generally unaffected in this range. However, severe deformation can
extend into this range.
Typical “Radial” results are shown below:
Figure 1 – Radial Movement Response from LV Open Circuit Test
Figure 2 – Radial Movement Response from HV Short Circuit Test
5.5.2 Axial Winding Elongation – “Telescoping”
Axial winding movement includes two types of winding geometric changes. The winding is stretched or
“telescoped” then constricts due to a reduction in the windings radius. The geometric variations induced by
this type of failure are complex and can lead to multiple resonances shifting across a broad frequency
range.
Table 2 – Axial Winding Deformation
Frequency Range Axial Winding Deformation

This region (core region) is generally unaffected during axial winding deformation.
Results in a change in impedance. The FRA trace for the affected winding causes a
difference between phases or previous results in the inductive roll-off portion.
Axial winding deformation is most obvious in this range. The bulk winding range

can shift or produce new resonance peaks and valleys depending of the severity of
the deformation. The changes will be greater on the affect winding, but it is still
possible to have the effects transferred to the opposing winding.
Axial winding deformation can shift or produce new resonance peaks and valleys
depending of the severity of the deformation. The changes will be greater on the
affect winding, but it is still possible to have the effects transferred to the opposing
winding.
The response to axial winding deformation is unpredictable.
Typical “Axial” results are shown below:
Figure 3 – Axial Movement Response from Tertiary Open Circuit Test
Figure 4 – Axial Movement Response from HV Short Circuit Test
5.5.3 Overall- Bulk Movement
These related failure modes are separated by the qualitative severity in the overall movement of windings
or sections of windings. Considered a general movement of the winding, the causes could be due to a
variety of reasons. Generally, this type of failure is used to describe the movement of the coils due to
physical shock as a result of high current forces or transportation. Physical movement of the transformer
could be due to shipping or seismic activity.
Table 3 – Bulk Winding Deformation

Frequency Range Bulk Winding Deformation

This region (core region) is generally unaffected during bulk winding movement.

This region is generally unaffected during bulk winding movement. All phases
should be similar.
Bulk winding movement is most obvious in this range. Newly created resonance
peaks or valleys are the key indicator. The bulk winding range can shift or produce
new resonance peaks and valleys depending of the magnitude of the Movement. The
changes will be greater on the affect phase.
Generally this range remains unaffected. However, changes to the CL capacitance
can cause resonance shifts in the upper portion of this range.
Changes to the CL capacitance can cause resonance shifts.
5.5.4 Core Defects
Core defects failures cause changes to the core’s magnetic circuit. Core defects can include burnt core
laminations, shorted core laminations, multiple/unintentional core grounds, lost core ground and joint
dislocations.
Table 4 – Core Defects

Frequency Range Core Defects

These types of failures will affect the lower frequency regions generally below 10
kHz. Core defects often change the primary core resonance shape. Less weight
should be placed on shifting, because identifying core defects can sometimes be
masked by the effects of core residual magnetization. If the open circuit core appears
to be loaded, (looking closer to a short circuit response), this could indicated a core
defect.
This region is generally unaffected during bulk winding movement. All phases
should be similar.
This range can shift or produce new resonance peaks and valleys.
Generally this range remains unaffected. However, if the fault is due to a core
ground issue, changes to the CL capacitance can cause resonance shifts in the upper
portion of this range.
If the fault is due to a core ground issue, changes to the CL capacitance can cause
resonance shifts.
Typical “Core Defect” results are shown below:

Figure 7 – Core Defect Response from LV Open Circuit Test
Figure 8 – Loss of Core Ground from LV Open Circuit Test
5.5.5 Contact Resistance
Though not necessarily a classical failure mode, high contact resistance readings can be detected by FRA
testing. Any metal to metal mating surface that connects the bushings to the windings, LTC or DETC can
lead to higher impedances through the test circuit applied. The end result can cause changes in both the low
and highest frequencies. Poor contact resistance can be cause by connections that have worked themselves
loose, corrosion, contact build-up or burning.
Table 5 – Contact Resistance

Frequency Range Contact Resistance

This region (core region) is generally unaffected in the presence of contact
resistance.
The results will not compare well against previous data or amongst phases. The
affected winding is generally offset.
This range can shift or produce new resonance peaks and valleys. The changes will
be greater on the affect phase.

Typical “Contact Resistance” results are shown below:
Figure 9 – Contact Resistance Response from LV Open Circuit Test
Figure 10 – Contact Resistance Response from HV Short Circuit Test
5.5.6 Winding Turn-to-Turn Short Circuit
Turn to turn faults is arguably one of the easier failure modes that can be identified by the FRA test. Turn
to turn short circuits can occur between two neighboring turns or between phases. The short can be either a
low impedance solid short or high resistance leakage path.
Table 6 – Winding Turn-to-Turn Short Circuit

Frequency Range Winding Turn-to-Turn Short Circuit

The short circuit failure mode removes the effect of the core’s reluctance from the
open circuit FRA results. The FRA open circuit trace assumes a similar behavior as
short circuit test. The affected winding will show the greatest change. This failure
mode will also affect the FRA responses from all other windings, but not as much.

Typical “Turn-to-Turn Short” results are shown below:
Figure 11 – Turn-to-Turn Short Response from HV Open Circuit Test
Figure 12 – Turn-to-Turn Short Response from HV Short Circuit Test
5.5.7 Open Circuited Winding
An open circuit can be cause by connections that come loose or coils that become burned through due to a
catastrophic thermal failure. The end result is very high impedances being inserted into FRA tests along the
same coil. It is common that the transfer function will drop across a wide spectrum. For complete open
circuits, the results will often be lost in the noise floor of the measurement.
Table 7 – Open Circuit Winding

Frequency Range Open Circuit Winding

The primary core resonance shape changes to account faulty winding scheme.
The open circuit winding influence is most obvious in this range. Newly created
predominate resonance peaks or valleys are the key indicator. This range can shift or
produce new resonance peaks and valleys. The changes will be greater on the affect

phase.
Typical “Open Circuit” results are shown below:
Figure 13 – Open Circuit Response from HV Open Circuit Test
Figure 14 – Open Circuit Response from LV Open Circuit Test
Figure 15 – Open Circuit Response from HV Short Circuit Test
5.5.8 Loss of Clamping Pressure

Over time the yoke and clamping structure of the core winding can become loose. This can lead to the
gradual spreading of the turn to turn distances axially along a winding.
5.5.9 Residual Magnetization
Though not necessarily a failure, residual magnetization within the core must be identified, so not be mis-
interpreted. Residual magnetization is cause by a section of core steal holding on the magnetic polarity due
to an applied DC bias. DC winding resistance testing, switching operations, and geomagnetic phenomena
are source of residual magnetism. Residual magnetization can be identified by the shifting of the low
frequency core resonance (zero) to the right from demagnetized results. Residual magnetization can be
removed by demagnetizing the core, and should be conducted if there is concern about the condition of the
core.
5.6 Modeling
Practically, there are two methods to obtain the frequency response characteristic of a transformer winding.
The first is to make physical measurements on the physical transformer winding structure.
The second is to model the transformer analytically for example with an RLC network, and mathematically
compute the desired response curves. The advantage of this is that all locations are mathematically
accessible and it is possible to simulate “fault” conditions without physically damaging a transformer.
The challenge is to build a valid and accurate analytic model. As discussed in the FRA theory section, it is
necessary to represent the complete three phase transformer in order to simulate an accurate FRA response.
A complete model of the windings can be constructed using a series of n-stage ladder networks such as
shown below to cover the required frequency range.
The inductance matrix would include the core and the leakage inductances, whereas the capacitance matrix
would include the shunt and series capacitances. Conductor loss and dielectric loss are represented as
resistance, which is frequency-dependent. The effect of any three-phase connections is recreated by further
extension of the model. The end result can be a very large model to be solved in the frequency domain,
requiring lengthy computation, but with modern computers this is no longer a problem. It has been shown,
however, that it is feasible to make certain simplifications to the model and its component matrices without
loss of accuracy in the critical regions of the response curve [1].

It is obviously important to understand the difference between the physical device and the mathematical
model that is used. However, if an existing transformer is modelled that has had sufficient FRA
measurements carried out both during and after its construction; it is possible to validate the model with a
good degree of confidence [2].
It is now possible to use the validated model to simulate particular physical changes, such as winding
damage or movement, in order to see the changes to the frequency response. Sensitivity studies, effects of
different connections, etc, can now be carried out to aid the use of the FRA diagnosis tool and improve the
ability to make confident interpretation of response curve changes.
[1] Jie Li, Charalambos Charalambous and Z.D. Wang, “Interpretation of FRA Results Using Low
Frequency Transformer Modelling” , the XV International Symposium on High Voltage Engineering,
University of Ljubljana, Elektroinštitut Milan Vidmar, Ljubljana, Slovenia, August 27-31, 2007
[2] J.A.S.B. Jayasinghe, Z.D. Wang, A.W. Darwin and P.N. Jarman, "Practical Issues in Making FRA
Measurements on Power Transformers", the 14th International Symposium on High Voltage Engineering
(ISH-2005), paper T07-96, 25-29 August 2005, Beijing, China.
6. Appendix 1: FRA Theory

Frequency response analysis (FRA) is a technique that is used to diagnose the condition, or more
importantly the change of mechanical condition, of a transformer by analyzing the transformer winding’s
frequency characteristic. FRA provides internal diagnostic information using non-intrusive procedures.
The technique consists of measuring the admittance (the current at the response end of the winding divided
by the voltage at the other input end of the winding) of the transformer winding largely comprised of
capacitive and inductive elements. The measurement is performed over a wide range of frequencies and the
result is compared with a reference “signature” or “fingerprint” result of the winding to make a diagnosis.
The measurements are generally made across the two terminals of a winding to derive the winding end-to-
end frequency response in the form of amplitude against frequency as given by the equation below
V&2 ( f )
A( f ) = 20 log 10 , dB
V& ( f )
1
& &
where V1 is the source voltage applied at one end of winding and V2 represents the measured quantity at
&
the other end of the winding. In theory, the second measurement V2 is a replica of the current response at
one end of winding to the injected voltage at the other end of the winding. Therefore the winding end-to-
end response can be regarded as the “virtual” transfer admittance of the winding.
The measured response is usually shown graphically by plotting the logarithmic amplitude ratio of the
output voltage to input voltage in dB (y scale) against the frequency (x scale). The frequency scale can be
logarithmic or linear. Both are used, although the logarithmic often shows the complete frequency range
more clearly. The linear scale is useful for looking at discrete frequency bands and to compare small
differences at particular frequencies. Typical responses of the LV windings of a transformer [1] are shown
below using both logarithmic and linear scales.

As can be seen, the Frequency Response of a transformer winding (often called the FRA response curve) is
quite complex and consists of decreasing and increasing magnitude (in dB) with respect to frequency. The
various resonances (maxima) and anti-resonances (minima) are determined by the electrical characteristics
of the transformer winding. These characteristics can be represented by the transformer equivalent circuit
and would include the elements of resistance, inductance and capacitance. The inductance and capacitance
values in this equivalent circuit are determined winding structure & geometry, and insulation structure &
clearance, and the resistance is contributed by conductive loss and dielectric loss.
In order to better understand and interpret possible changes to the FRA response, it is useful to understand
which elements are responsible for the various regions and shapes of the response curve. This can be
realized by examining the individual responses of the various elements and their combinations.
Taking the simplest representation of a single winding as a lumped circuit elements network, and initially
ignoring the inductive and capacitive couplings between windings, the end-to-end frequency response of a
winding is dependent on how the elements in this network behave together at different frequencies.
At low frequencies, a transformer winding behaves as an inductive element, and the end-to-end FRA
response follows a falling magnitude trend across the frequency range with a linearly decreasing slope of
approximately –20dB per decade. As the inductance is increased, the magnitude is decreased. Power
transformers with higher voltage and larger power rating usually have larger negative response magnitudes
at low frequencies. Effectively there are two inductance components affecting the frequency response; one
is the core magnetizing inductance, and the other is the leakage inductance of the winding. Each affects the
response in a different frequency region. The former affects the FRA response in the lower frequency
region up to ~100Hz while the latter influences the FRA response at higher frequencies.
At high frequencies, a transformer winding behaves as a capacitive element, and the end-to-end FRA
response follows a rising magnitude trend across the frequency range with a linearly increasing slope of
approximately 20dB per decade. As the capacitance is increased, the magnitude is increased. Power
transformers having both higher voltage and larger power rating usually have smaller negative response
magnitudes at high frequencies.
The combination of winding inductance and winding series capacitance results in paralleled inductance and
capacitance. LC in parallel will produce parallel anti-resonance at a certain frequency, blocking the signal
at that particular frequency. This consequently produces a local anti-resonance in the magnitude response at
that particularly frequency.
In the case that the winding series capacitance is relatively small (a single layer winding or a plain disc
winding as an example), the shunt capacitance becomes significant. The combination of winding
inductance and shunt capacitance will result in inductance and capacitance in series and produce series
resonance. The simplest representation of LC in series is a T-connection where the shunt capacitance is
connected in the middle of the two halves of the winding inductance. The end-to-end FRA response of a T–
connected LC network shows a series resonance at a certain frequency, amplifying the signal at that
particular frequency. This consequently produces a resonance at that particularly frequency.

The effect of resistance on the end-to-end FRA response is to attenuate (smooth) the sharpness of the
resonances and the anti-resonances. Conductor loss and dielectric loss are represented as resistance, which
is frequency-dependent. In the equivalent circuit of a transformer winding, these are either connected in
series with the inductance or connected in parallel with the capacitance.
In summary, the basic features of FRA response are shown alongside.
However, even in a single winding, these basic LC components are produced by mutual coupling between
turns and parts of a winding, effectively resulting in a network of multiple lumped parameters. The
diagrams below show firstly, the inclusion of mutual coupling and secondly, the inclusion of losses.
a) n-stage lumped network with mutual coupling; b) n-stage lumped ladder network with losses
Thus the winding equivalent circuit now comprises of multiple distributed parameters, and the frequencies
of the resonances are determined primarily by the winding leakage inductance L, the shunt capacitance Cg,
and the winding series capacitance Cs. There are two extreme cases.
In the one case, when the winding series capacitance is small enough to be neglected (where Cs tends to
zero), there will be resonances with even frequency intervals. On the other hand, when the winding series
capacitance is extremely large (where Cg tends to zero), there will be no resonance at all. With low Cs, the
response begins with a flat magnitude trend and resonances at intervals of frequencies determined by the L,
Cg and Cs values and then followed by a decreasing inductive trend. An anti-resonance appears at a
frequency determined by L and Cs, which is followed by the increasing capacitive trend. As Cs is
increased, some of the resonances diminish and the anti-resonance appears at a lower frequency.
The relative proportions of the series capacitance (Cs) and the ground capacitance (Cg) are thus significant
in determining the FRA response for a specific winding structure. They determine not only which winding
type has the higher magnitude, but also the shapes and the position of the resonances and anti-resonances
and whether these appear at lower or higher frequencies.

In a “practical” transformer, there are at least two windings per phase and the interaction between these two
windings needs to be considered. In general, this interaction can be described by inductive and/or
capacitive coupling. There are also interactions between windings of different phases due to their electrical
connections, such as for a delta, or the sharing of the same neutrals. Windings not under test and not
electrically connected to the tested winding, will also have an impact on the frequency response through
mutual inductive and capacitive couplings.
Summary of Effects
Knowledge of the above effects and the response “shapes” for different windings and combinations of
windings is useful in making diagnostic interpretation of changes to the FRA response curves.
The figure below [1] shows typical frequency responses for the HV windings of an autotransformer taken
from 10Hz up to 10MHz and shown on the logarithmic scale. The frequency range can be divided into 4
regions depending on the dominant influence of the various electrical properties of the transformer with
frequency.
The frequency regions described here are typical but not exact, and they would be slightly varied depending
on transformer design and arrangement.
10
0
Winding
-10
Core structure
-20 influence influence
Magnitude, dB
-30
-40
-50
-60 Earthing
Interaction leads
-70 A phase between influence
B phase windings
-80
C phase
-90
1 2 3 4 5 6 7
10 10 10 10 10 10 10
Frequency, Hz
Region 1 - Core Effect

At low frequency (below 2 kHz), the frequency response begins with the decreasing magnitude of -
20dB/decade. This is due to the magnetizing inductance of the core. This is then followed by a minimum
which occurs due to the series resonance between the magnetizing inductance of the core and the total
capacitance of the transformer. In this example, the B phase is clearly different from the other two phases
due to the different magnetic reluctance path of the middle leg of a three-limb core.
Region 2 - Interaction Between Windings

Between 2-20 kHz, the frequency response is influenced by the interaction between the windings as well as
by how the windings and neutrals are connected and terminated (open/closed delta; floating/grounded).
Region 3 - Effect of Winding Structure of the Winding under Test

From 20 kHz up to 1MHz, the winding structure will heavily influence the frequency response. In this
example, the FRA plot for this frequency range has less resonances and anti-resonance and a mainly
capacitive raising trend, due to the high series capacitance of the HV winding.
Region 4 - Effect of Leads of Taps and Earthing leads

Beyond the frequency of 1MHz, the trend of the frequency response is irregular and complex, influenced
by the tap leads and the measurement earthing leads.
[1] D.M. Sofian, “Transformer FRA Interpretation for Detection of Winding Movement”, PhD thesis, the
University of Manchester, July 2007.
End of Paper


IEEE Format FRA Draft Guide D6

Uploaded by

Copyright:

Available Formats

IEEE Format FRA Draft Guide D6

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IEEE Format FRA Draft Guide D6

Uploaded by

Copyright:

Available Formats

IEEE PC57.

149/D6 April 2009

Working Group of the

IEEE Power Engineering Society Transformers Committee

Copyright © 2006 by the Institute of Electrical and Electronics Engineers, Inc.

IEEE Standards Activities Department

Copyright © 2006 IEEE. All rights reserved.

Keywords: Frequency Response Analysis (FRA), magnitude, phase angle, attenuation,

Copyright © 2006 IEEE. All rights reserved. ii

<Select this text and type or paste introduction text>

Charles Sweetser, Chair

Raj Ahuja Eduardo Garcia Triomphant Ngnegue

(to be supplied by IEEE)

Copyright © 2006 IEEE. All rights reserved. iii

1. Scope and Application ............................................................................... ¡Error! Marcador no definido.

1.1 Purpose ............................................................................................... ¡Error! Marcador no definido.

2. FRA Measurement Overview ..................................................................................................................... 5

2.1 Use of FRA .......................................................................................................................................... 5

3. Making an FRA Measurement ................................................................................................................... 7

3.1 Test Procedures.................................................................................................................................... 7

4. Test records ...............................................................................................................................................16

4.1 Introduction ........................................................................................ ¡Error! Marcador no definido.

5. Measurement Analysis and Interpretation .................................................................................................18

5.1 Introduction ........................................................................................................................................18

6. Appendix 1: FRA Theory ..........................................................................................................................32

Copyright © 2006 IEEE. All rights reserved. iv

Draft Trial-Use Guide for the Application and

1. Scope and Application

1.2.1 Frequency Response Analysis (FRA)

Copyright © 2006 IEEE. All rights reserved. 1

1.2.2 FRA Magnitude

M(ω) = | H(j ω)| = | I(j ω)/ V(jω)|

1.2.3 FRA Phase Angle

1.2.4 Sweep Frequency Method

1.2.5 Impulse Voltage Method

1.2.6 Open-Circuit Self Admittance

1.2.7 Short-Circuit Self Admittance

1.2.8 Winding Self Admittance

Copyright © 2006 IEEE. All rights reserved. 2

 I 1  Y11 Y12  V1 

1.2.9 Inter-winding Admittance

1.2.10 FRA Resonance Frequency

1.2.11 Test Specimen

Test specimen is the particular winding or winding segment being tested

1.2.12 Equipment Ground

Copyright © 2006 IEEE. All rights reserved. 3

1.2.13 Measurement Ground

• Generated by the measurement instrument and coupled directly.

1.2.15 Mechanical Movement

1.2.16 Amplitude Displacement

1.2.17 Phase Displacement

1.2.18 Frequency Displacement

Copyright © 2006 IEEE. All rights reserved. 4

1.2.19 Base Line Measurement

1.2.20 Minor Deviation

1.2.21 Significant Deviation

2. FRA Measurement Overview

2.1 Use of FRA

• Factory short-circuit testing

Copyright © 2006 IEEE. All rights reserved. 5