omicron sfra

8/15/2019 Omicron SFRA

1/83

Sweep Frequency Response

Analysis

Why

it

should

be

in

your

diagnostic toolbox

Keith Hill – Doble Engineering


2/83

SFRA History

Sweep Frequency Response Analysis (SFRA)

has been

around

for

several

years

Once required highly specialized personnel to

perform

and

interpret

the

resultsPowerful and sensitive tool to evaluate the

mechanical integrity


3/83

SFRA History

1960:

Lech

&

Tyminski,

Poland

1966:

L&T

Publish

Results

1969: Smith, UK (ERA testing)

1976:

First

Doble

Paper:

A.G. Richenbacker “Frequency Domain Analysis

of

Responses

from

L.V.I.

Testing

of

Power

Transformers”


4/83

SFRA History

1978: Dick & Erven, Canada, IEEE first

SFRA

approach:

“Transformer

Diagnostic Testing by FRA”

1980’s:

CEGB,

UK

– SFRA

Field

Trials1990’s:Application

Development

1990’s:

LVI

Commercialized2000: SFRA Commercialized (Doble)


5/83

SFRA

Does not evaluate the condition of the

insulationPower Factor and low voltage TTR can be

acceptable

but

can

still

have

problemsMechanical aspects of the unit

Identify

problems

after

a

faultGreat tool to help identify shipping

damage


6/83


7/83

When to Perform SFRA

At the factory Benchmark

When shipping

carriers

change

Arrival on site

This is not “over‐kill”

During routine testing if no benchmark

After a fault (benchmark is very helpful)


8/83

Transformer Fire


9/83

Damaged Bushing and Bus


10/83

Tests after the Fire

SFRA tests did not reveal a problem

(compared to

benchmark

signatures)

Damage was caused by a fire in the bus

duct

Transformer returned to service


11/83

SFRA Data

Interpretation can be challenging if not

impossible for

the

novice

or

very

experienced if benchmark data is not

available


12/83

SFRA Data


13/83

SFRA Data

Performing the test not difficult as the

setup is

very

simple

Software and templates are often

supplied by

the

manufacturers


14/83

SFRA Setup


15/83

Various Templates


16/83

Various Templates


17/83

Another Tool in the Toolbox


18/83

Component Characteristics

Three basic components

Resistors Capacitors

Inductors

Each component has a different response being closely elated to their

geometry; both

internal

and

external

to

other components


19/83

Resistor

The resistance of a resistor is not frequency

dependent as the resistance will remain the samewith a change in frequency. There will be a straight

line from a low frequency to a higher frequency


20/83

Inductor

At a low frequency the inductor will act as a shortand the inductance will increase as the frequency

increases.


21/83

Response of an Inductor

0 dB down at low

frequency means it looks

like a dead short

Larger inductances start to

roll off at lower frequencies

Inductive roll off


22/83

Capacitor

A capacitor will be the opposite of an inductor as the

capacitor will increase with an increase in frequency


23/83

Response of a Capacitor

Low frequency

response is like

an open circuit

Capacitive climb

back

Knee point depends

on the size of the

capacitor 0 dB down athigh frequency

like a dead short


24/83

“Real Transformer”


25/83

Parallel RLC Circuit

Constant dBs v. frequency

Dead short at

low frequency

Open circuit at

high frequency

Open circuit at

low frequency

Dead short at

high frequency


26/83

Response of a Parallel RLC Circuit


27/83

Test Setup

Once the nameplate and templates have been

selected the

tester

must

make

the

connections to the apparatus.

Tester should make sure that the transformer

is isolated from all energy sources and should

be aware

that

static

voltage

may

be

present

if

not properly grounded.


28/83

Test Setup

The first resonance frequency will be affected if the transformer is tested with insulating fluid

and

without

insulating).

Removing

oil

lowers capacitances and resonances shift to higher frequencies. (lower resonant with oil)

How the neutral position of the Load Tap

Changer

(LTC)

was

reached

as

there

have

been

cases in which reaching neutral position from the raise or from the lower taps affected the

signatures.


29/83

Test Setup

A magnetized core can affect the signatures

If LTC or DETC are not on neutral the

signatures

can

change

Test

on

16R

if

only

one

test

is

performed


30/83

Test Setup

Shorting of the secondary windings for the short circuit test can be affected if the

secondary windings

are

not

“closed”

It

is

recommended to jumper X1 – X2, X2 – X3 and X3 – X1 thus closing the delta or wye

windings.

Make

sure

that

the

transformer

being

tested

is

grounded


31/83

Test Setup

Make sure that the test leads grounds are

properly terminated

to

a “good”

ground

Select

proper

winding

configuration

template


32/83

TYPICAL

SFRA

SIGNATURES


33/83

WYE ‐ With and Without Oil

With Oil

Lower

resonant

frequencies

Without OilHigher

resonant

frequencies


34/83

Delta – With and Without Oil

With Oil

Without Oil


35/83

We

see

the

same

changes

when

performing power factor tests on

transformers with

and

without

oil


36/83

Power Factor With and Without Oil

ma Cap mA

CapCH+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 24894 65.835 17463

CL 64.03 16984 45.144 11974

CHL 29.9 7905 20.759 5506

With

Oil Without

Oil


37/83

Magnetized Core

The same phase was tested on arrival at site.

The

low

frequency

responses

show

offset

due

to core magnetization while the high

frequency responses are unchanged.

Often times a magnetized core can be detected by performing excitation current

tests.

Several manufacturers

have

winding

resistance test sets that have the ability to de‐

magnetize the core.


38/83

Magnetized Core

It is recommended to demagnetize and repeat

the

test

to

confirm

that

the

core

was

magnetized and is the reason for the shift.


39/83

Magnetized Core


40/83

INTERPRETATION

OF

SIGNATURES

i f Si


41/83

Interpretation of Signatures

Most users of power factor equipment are

aware that

the

power

factor

test

is

a reference

test and allows the user to compare results to

past tests and also to tests results of identical

equipment.

If “benchmark”

power

factors

or

watts loss limits have not been established it

can often

be

difficult

to

determine

the

condition of the apparatus being tested

I i f Si


42/83

Interpretation of Signatures

This is true for SFRA test results as the

signatures before

and

after

an

event

can

often help to identify a problem. The

frequency

range

in

which

the

changes

take

place will assist the tester in determining

what component or components have

changed since the benchmark tests were

performed.

Small change in frequency and dB


43/83

Small change in frequency and dB

Open circuit results show ONE phase

with a shift left where it should not

H B kli


44/83

Hoop Buckling

This phase appears to have

symptoms of hoop buckling

Suspect unit shows same

increased impedance (more

dB’s down) for one phase ‐ the

same

one

with

the

shift

left

on

open circuit results

H B kli


45/83

Hoop Buckling

H B kli


46/83

Hoop Buckling

“Transformer was bent ‐ but not broken”

This transformer was returned to service for

several months

until

a replacement

could

be

located. Once the unit was removed from

service the inspection of the windings

revealed winding

defamation

that

is

classified as “hoop buckling”

Frequency Regions


47/83

Frequency Regions

The following four zones are utilized by one

equipment

manufacturer

for

the

analysis

of

SFRA results.


48/83

Frequency Regions


49/83

Frequency Regions

The following four zones are utilized by one equipment manufacturer for the analysis of SFRA

results.

Region 1:


50/83

Frequency Regions

Region 3: 150 ‐ 400kHz: Deformation

within

main

and

tap

windings

Region 4: 400kHz ‐ 2MHz Movement of

main and tap winding leads

Typical Signatures


51/83

Typical Signatures

Short circuit

Open circuit

Is this a Delta or a

Wye

signature?

Wye Wye HV Winding LV OC


52/83

Wye‐Wye_HV Winding‐LV_OC

H2‐H0

middle winding

Wye‐Wye LV Winding‐HV OC


53/83

Wye‐Wye_LV Winding‐HV_OC


54/83

Delta‐Wye HV Winding‐LV OC


55/83

Delta Wye_HV Winding LV_OC

Delta‐Wye LV Winding‐HV OC


56/83

Delta Wye_LV Winding HV_OC

Delta‐Wye HV Winding‐LV SC


57/83

Delta Wye_HV Winding LV_SC

Starts at “0”

Auto Transformer HV‐OC


58/83

Auto Transformer_HV OC

Auto Transformer LV‐OC


59/83

Auto Transformer_LV OC

Auto Transformer Tertiary‐OC


60/83

Auto Transformer_Tertiary OC

Variation with LTC


61/83

Variation with LTC

8 Lower

16 Lower

Variation with DETC


62/83

Tap E

Tap A

Variation with Reversing Switch


63/83

g

One winding

– neutral

from

raise & lower

Effect of Poor Grounding


64/83

g

Bad red

lead

ground

Original


65/83

ANALYSIS

OF

DATA


66/83

TRANSFORMER

1

Normal Response


67/83

p

Why is the Signature Different?


68/83

Shorted Turns

Shorted Turns


69/83


70/83

TRANSFORMER

2

Power Factor Results


71/83

Test Eng kV mA

Watts PF

1 High 10 113.75 4.716 0.41

2 High 10 15.782 0.802 0.513 High 10 97.979 3.99 0.41

4 Cal.1 97.968 3.914 0.40

5 Low 2 133.91 4.972 0.37

6 Low 2 35.931 1.046 0.29

7 Low 2 97.975 3.940 0.40

8 Cal.2 2 97.979 3.926 0.40

Excitation Data


72/83

Normally expect the 2 highs to be within 5%


73/83

Low voltage TTR and winding resistance were

considered acceptable

by

the

tester.

Since the excitation is often affected by a


74/83

magnetized

core

you

may

question

placing

this transformer into service.

Would you place this transformer into

service?


75/83

SFRA testing was recommended for this unit due to

the very wide variance in excitation current results

SFRA ‐ HV Open Circuit Tests


76/83

H3 – H2 Appears to be Abnormal

HV Tests with LV Winding Short Circuited


77/83

Signatures appear normal

SFRA ‐LV Open Circuit Tests


78/83

X3 – X0

Appears

to

be

Abnormal

and

is

the

Same

Phase

as

H3

– H2

Shorted Turns


79/83

What Not to Send to Your Customer


80/83

References


81/83

Sweep

Frequency

Response

Analysis

Transformer

Application

Authors:

Charles Sweetser, B.Sc., M.Sc.

Dr. Tony

McGrail,

B.Sc.

M.Sc.,

PhD.

References


82/83

Special thanks

to

various

Doble

client

service

engineers who over the years have provided a wide variety of data and technical expertise for

this presentation.

Doble

Internal

Power

Point

Training

G. Matthew

Kennedy,

Matt

Garber,

Mario

Locarno, Long Pong, Dr. Tony McGrail, Linda Nowak, and Charles Sweetser


83/83

Please Remember to Complete

Your Evaluation

Form

Thank You!

omicron sfra

Documents