SweepFrequencyResponse

Analysis
Whyitshouldbeinyour
diagnostictoolbox
KeithHill DobleEngineering

SFRAHistory
SweepFrequencyResponseAnalysis(SFRA)
hasbeenaroundforseveralyears
Oncerequiredhighlyspecializedpersonnelto
performandinterprettheresults
Powerfulandsensitivetooltoevaluatethe
mechanicalintegrity

SFRAHistory
1960:Lech&Tyminski,Poland
1966:L&TPublishResults
1969:Smith,UK(ERAtesting)
1976:FirstDoblePaper:
A.G.Richenbacker FrequencyDomainAnalysis
ofResponsesfromL.V.I.TestingofPower
Transformers

SFRAHistory
1978:Dick&Erven,Canada,IEEEfirst
SFRAapproach:TransformerDiagnostic
TestingbyFRA
1980s:CEGB,UK SFRAFieldTrials
1990s:ApplicationDevelopment
1990s:LVICommercialized
2000:SFRACommercialized(Doble)

SFRA
Doesnotevaluatetheconditionofthe
insulation
PowerFactorandlowvoltageTTRcanbe
acceptablebutcanstillhaveproblems
Mechanicalaspectsoftheunit
Identifyproblemsafterafault
Greattooltohelpidentifyshipping
damage

SFRA
SFRAtestshaveidentifiedproblems
Displacementanddeformationof
windings
Coredeformation
Shortedturns
Magnetizedcore
Othermechanicalproblems

WhentoPerformSFRA
AtthefactoryBenchmark
Whenshippingcarrierschange
Arrivalonsite
Thisisnotoverkill
Duringroutinetestingifnobenchmark
Afterafault(benchmarkisveryhelpful)

TransformerFire

DamagedBushingandBus

TestsaftertheFire
SFRAtestsdidnotrevealaproblem
(comparedtobenchmarksignatures)
Damagewascausedbyafireinthebus
duct
Transformerreturnedtoservice

SFRAData
Interpretationcanbechallengingifnot
impossibleforthenoviceorvery
experiencedifbenchmarkdataisnot
available

SFRAData

SFRAData
Performingthetestnotdifficultasthe
setupisverysimple
Softwareandtemplatesareoften
suppliedbythemanufacturers

SFRASetup

VariousTemplates

VariousTemplates

AnotherToolintheToolbox

ComponentCharacteristics
Threebasiccomponents
Resistors
Capacitors
Inductors

Eachcomponenthasadifferent
responsebeingcloselyelatedtotheir
geometry;bothinternalandexternalto
othercomponents

Resistor

The resistance of a resistor is not frequency

Inductor

At a low frequency the inductor will act as a short

ResponseofanInductor
0 dB down at low
frequency means it looks
like a dead short

Larger inductances start to

Inductive roll off

Capacitor

A capacitor will be the opposite of an inductor as the

ResponseofaCapacitor
Knee point depends
on the size of the
capacitor

Low frequency
response is like
an open circuit

Capacitive climb
back

0 dB down at
high frequency
like a dead short

RealTransformer

ParallelRLCCircuit
Constant dBs v. frequency

Dead short at
low frequency

Open circuit at
high frequency

Open circuit at
low frequency

Dead short at
high frequency

ResponseofaParallelRLCCircuit

TestSetup
Oncethenameplateandtemplateshavebeen
selectedthetestermustmakethe
connectionstotheapparatus.
Testershouldmakesurethatthetransformer
isisolatedfromallenergysourcesandshould
beawarethatstaticvoltagemaybepresentif
notproperlygrounded.

TestSetup
Thefirstresonancefrequencywillbeaffected
ifthetransformeristestedwithinsulating
fluidandwithoutinsulating).Removingoil
lowerscapacitancesandresonancesshiftto
higherfrequencies.(lowerresonantwithoil)
HowtheneutralpositionoftheLoadTap
Changer(LTC)wasreachedastherehavebeen
casesinwhichreachingneutralpositionfrom
theraiseorfromthelowertapsaffectedthe
signatures.

TestSetup
Amagnetizedcorecanaffectthesignatures
IfLTCorDETCarenotonneutralthe
signaturescanchange
Teston16Rifonlyonetestisperformed

TestSetup
Shortingofthesecondarywindingsforthe
shortcircuittestcanbeaffectedifthe
secondarywindingsarenotclosedItis
recommendedtojumperX1 X2,X2 X3and
X3 X1thusclosingthedeltaorwye
windings.
Makesurethatthetransformerbeingtestedis
grounded

TestSetup
Makesurethatthetestleadsgroundsare
properlyterminatedtoagoodground
Selectproperwindingconfigurationtemplate

TYPICAL
SFRA
SIGNATURES

WYE WithandWithoutOil

With Oil
Lower
resonant
frequencies

Without Oil
Higher
resonant
frequencies

Delta WithandWithoutOil

With Oil

Without Oil

Weseethesamechangeswhen
performingpowerfactortestson
transformerswithandwithoutoil

PowerFactorWithandWithoutOil
WithOil
WithoutOil
ma
Cap
mA
Cap
CH+CHL 49.53 13138 35.263 9353
CH
19.69 5223 14.503 3847
CHL
29.82 7909 20.76 5507
CL+CHL 93.85
CL
64.03
CHL
29.9

24894 65.835 17463

MagnetizedCore
Thesamephasewastestedonarrivalatsite.
Thelowfrequencyresponsesshowoffsetdue
tocoremagnetizationwhilethehigh
frequencyresponsesareunchanged.
Oftentimesamagnetizedcorecanbe
detectedbyperformingexcitationcurrent
tests.Severalmanufacturershavewinding
resistancetestsetsthathavetheabilitytode
magnetizethecore.

MagnetizedCore
Itisrecommendedtodemagnetizeandrepeat
thetesttoconfirmthatthecorewas
magnetizedandisthereasonfortheshift.

MagnetizedCore

INTERPRETATION
OF
SIGNATURES

InterpretationofSignatures
Mostusersofpowerfactorequipmentare
awarethatthepowerfactortestisareference
testandallowstheusertocompareresultsto
pasttestsandalsototestsresultsofidentical
equipment.Ifbenchmarkpowerfactorsor
wattslosslimitshavenotbeenestablishedit
canoftenbedifficulttodeterminethe
conditionoftheapparatusbeingtested

InterpretationofSignatures
ThisistrueforSFRAtestresultsasthe
signaturesbeforeandafteraneventcan
oftenhelptoidentifyaproblem.The
frequencyrangeinwhichthechangestake
placewillassistthetesterindetermining
whatcomponentorcomponentshave
changedsincethebenchmarktestswere
performed.

SmallchangeinfrequencyanddB
OpencircuitresultsshowONEphase
withashiftleftwhereitshouldnot

HoopBuckling
Suspectunitshowssame
increasedimpedance(more
dBsdown)foronephase the
sameonewiththeshiftlefton
opencircuitresults

Thisphaseappearstohave
symptomsofhoopbuckling

HoopBuckling

HoopBuckling
Transformerwasbent butnotbroken
Thistransformerwasreturnedtoservicefor
severalmonthsuntilareplacementcouldbe
located.Oncetheunitwasremovedfrom
servicetheinspectionofthewindings
revealedwindingdefamationthatis
classifiedashoopbuckling

FrequencyRegions
Thefollowingfourzonesareutilizedbyone
equipmentmanufacturerfortheanalysisof
SFRAresults.

FrequencyRegions

<10kHz

10kHzto
150kHz

150kHz
400kHz
400kHz
2MHz

FrequencyRegions
Thefollowingfourzonesareutilizedbyone
equipmentmanufacturerfortheanalysisof
SFRAresults.
Region1: <10kHz:Coredeformation,
opencircuits,shortedturns,residualmag
Region2: 10kHzto150kHz:BulkWinding
movementrelativetoeachother

FrequencyRegions
Region3: 150 400kHz:Deformation
withinmainandtapwindings
Region4: 400kHz 2MHzMovementof
mainandtapwindingleads

TypicalSignatures
Shortcircuit

Open circuit

IsthisaDeltaora
Wyesignature?

WyeWye_HVWindingLV_OC

H2H0
middlewinding

WyeWye_LVWindingHV_OC

WyeWye_HVWindingLV_SC

Startsat0

DeltaWye_HVWindingLV_OC

DeltaWye_LVWindingHV_OC

DeltaWye_HVWindingLV_SC

Startsat0

AutoTransformer_HVOC

AutoTransformer_LVOC

AutoTransformer_TertiaryOC

VariationwithLTC

16Lower

8 Lower

VariationwithDETC

TapA
TapE

VariationwithReversingSwitch

Onewinding neutralfrom
raise&lower

EffectofPoorGrounding
Badredlead
ground

Original

ANALYSIS
OF
DATA

TRANSFORMER
1

NormalResponse

WhyistheSignatureDifferent?

ShortedTurns

ShortedTurns

TRANSFORMER
2

PowerFactorResults
Test
1
2
3
4
5
6
7
8

Eng
High
High
High
Cal.1
Low
Low
Low
Cal.2

kV
10
10
10
2
2
2
2

mA
113.75
15.782
97.979
97.968
133.91
35.931
97.975
97.979

Watts
4.716
0.802
3.99
3.914
4.972
1.046
3.940
3.926

PF
0.41
0.51
0.41
0.40
0.37
0.29
0.40
0.40

ExcitationData

Normallyexpectthe2highstobewithin5%

 LowvoltageTTRandwindingresistancewere
consideredacceptablebythetester

Sincetheexcitationisoftenaffectedbya
magnetizedcoreyoumayquestionplacing
thistransformerintoservice.
Wouldyouplacethistransformerinto
service?

 SFRAtestingwasrecommendedforthisunitdueto
theverywidevarianceinexcitationcurrentresults

SFRA HVOpenCircuitTests

H3 H2AppearstobeAbnormal

HVTestswithLVWindingShortCircuited

Signaturesappearnormal

SFRALVOpenCircuitTests

X3 X0AppearstobeAbnormalandistheSamePhaseasH3 H2

ShortedTurns

WhatNottoSendtoYourCustomer

References
SweepFrequencyResponseAnalysis
TransformerApplication
Authors:
CharlesSweetser,B.Sc.,M.Sc.
Dr.TonyMcGrail,B.Sc.M.Sc.,PhD.

References
SpecialthankstovariousDobleclientservice
engineerswhoovertheyearshaveprovideda
widevarietyofdataandtechnicalexpertisefor
thispresentation.
DobleInternalPowerPointTraining
G.MatthewKennedy,MattGarber,Mario
Locarno,LongPong,Dr.TonyMcGrail,Linda
Nowak,andCharlesSweetser

PleaseRemembertoComplete
YourEvaluationForm
ThankYou!