Regional Technical Seminar

Factory Testing
Vivek Bhatt
[email protected]

Prolec GE Waukesha / Proprietary and Confidential

 Reasons for Testing

 Compliance with user specifications

 Compliance with applicable industry standards

 Assessment of quality and reliability

 Verification of design calculations

Prolec GE Waukesha / Proprietary and Confidential

 • C57.12.00 and C57.12.90 are two most commonly used transformer IEEE standards
• Both these standards got released in early 2022, with revision date of 2021.

Prolec GE Waukesha / Proprietary and Confidential

 IEEE C57.12.00-2021 Table 17
Routine Tests
Routine tests shall be made on every transformer to verify that the product meets the design specifications

Design Tests
Design tests shall be made to determine the adequacy of the design of a particular type, style, or model of
transformer or its component parts. Test data from previous similar designs may be used for current designs,
where appropriate. Once made, the tests need not be repeated unless the design is changed to modify
performance.

Other Tests
Other tests are identified in product specifications and may be specified by the purchaser in addition to
routine tests

Prolec GE Waukesha / Proprietary and Confidential

 Class I and Class II
Power Transformers

Nominal System Top Nameplate

Voltage (kV) Rating (KVA)
Any
< 69 kV
Class I = 69 kV
<10,000 – 1ϕ
<15,000 – 3ϕ
Any
≥115 kV
Class II ≥69 kV < 115kV
≥10,000 – 1ϕ
≥15,000 – 3ϕ

Prolec GE Waukesha / Proprietary and Confidential

 Preliminary Testing

Prolec GE Waukesha / Proprietary and Confidential

 Preliminary Testing
Tests Class I Class II

Voltage Ratio Routine Routine

Insulation Power factor Routine Routine

Insulation Resistance Routine Routine

Waukesha Waukesha
1Φ Excitation test
Routine Routine
Waukesha Waukesha
CT Ratio & Polarity
Routine Routine
Control Wiring Checks
Routine Routine
& Hi-pot
Waukesha
Auxiliary Losses Routine
Routine

Prolec GE Waukesha / Proprietary and Confidential

 Preliminary Tests
Voltage Ratio Test
• Performed with ratio-meter (TTR) based on voltage comparison
principle to check that windings are wound with correct turns
including tapped turns
• Low voltage is applied to HV winding and voltage measured across
LV/other winding is fed back to ratio-meter which displays the
applied/measured voltage ratio (= turns ratio)
• Turns ratio is compared with voltage ratio requirement to meet
tolerance of +/- 0.5%

CT Ratio and Polarity Test

• Verify Polarity ( & also Ratio ) and wiring to control box

Leak Test
• 10 PSI for 10 hours minimum
• Typically Tested during Manufacturing before release to Test

Prolec GE Waukesha / Proprietary and Confidential

 Preliminary Tests (cont.)
Insulation Power Factor C57.12.90 Sec. 10.10

Between
Windings

To Ground

• Test voltage is typically 10kV

• Power Factor is affected by temperature;
Recommended 100 to 400 C
• No IEEE Limit for PF, Max 0.5% good for
most units

Prolec GE Waukesha / Proprietary and Confidential

 Preliminary Tests (cont.)
Single Phase Excitation Test
• Test typically performed on HV terminal and tested at 10kV
• Test is performed one phase at a time and currents are compared
• For three phase transformers, two phases are expected to have
similar and higher current compared to third; current measured on
phase wound on center limb on three legged core will have lower
current due to lower magnetic reluctance

Winding Insulation Resistance C57.12.90 Sec. 10.11

• Typically tested at 1/2.5/5 kV and held for 1 minute before taking reading
• Test performed high voltage to low voltage and ground and low voltage to
high voltage and ground
• Acceptable values varies with design, voltage class and cooling medium -
typically is greater than 500MOhms

Prolec GE Waukesha / Proprietary and Confidential

 Performance Tests

Prolec GE Waukesha / Proprietary and Confidential

 Performance Characteristic Tests
Tests Class I Class II

No Load Losses Routine Routine

% Excitation Tests Routine Routine

Load Losses Routine Routine

Positive Sequence
Routine Routine
Impedance

Zero Sequence
Special Routine
Impedance

Winding Resistances Routine Routine

Sound Test Other Other

Prolec GE Waukesha / Proprietary and Confidential

 No-Load Test Connection – C57.12.90 Section: 8
No Load Loss and Excitation Current
• Core Loss ~ Hysteresis Loss, Eddy Current Loss
• Hysteresis Loss ~ Flux Density & Grade of Steel
• Eddy Current Loss ~ Frequency, Temperature

Test Circuit
• Transformer is excited from either TV/LV or HV side at 60
Hz with a variable voltage sinusoidal source
• All other terminals are left open
• Applied voltage is slowly increased to test voltage 90%,
100%, 110%
• Need high precision measurement System
• Losses corrected to 20oC
Prolec GE Waukesha / Proprietary and Confidential
 Load Losses and % Impedance

Load Loss C57.12.90 Sec. 9

• Load Losses are the losses of TRANSFORMER DUE TO LOAD CURRENT
• Load Loss = I2R loss + Eddy loss + Stray loss

• Eddy losses depend on conductor thickness and width and leakage

flux distribution

• Stray loss depends on % impedance, winding dimensions and clearance to tank and clamps

Impedance
• % Impedance = VOLTAGE FOR RATED CURRENT X 100
RATED VOLTAGE

Prolec GE Waukesha / Proprietary and Confidential

 Load Loss Test Connection
Test Circuit
Transformer is excited, preferably from HV side at 60 Hz with a variable voltage sinusoidal source. LV
terminals are shorted. Applied voltage is slowly increased to feed the rated test current in the windings.

Measurement

With the help of a precision loss measurement system load current, voltage and losses are measured:

Measured loss = I² R Loss at ambient + stray loss

I² R Loss at 85ºC = I² R Loss at ambient*(234.5+85)/(234.5 + ambient)
Stray Loss at 85ºC = Stray Loss at ambient*(234.5+ambient)/(234.5 + 85)

Prolec GE Waukesha / Proprietary and Confidential

 Load Loss Test Connection (cont.)

Prolec GE Waukesha / Proprietary and Confidential

 Impulse Testing – C57.12.90 Section:10.3
• Lightning Impulse Class II – Routine, Class I – Other
‒ Reduced Wave RFW (50 – 70% of Full Wave)
‒ Full Wave *
‒ Two (2) Chopped Waves
‒ Full Wave
‒ Full Wave *

• Transformer Neutrals
‒ 1 RFW
‒ 2 FW
‒ 1 FW*

*Added in 2015 Standard

Prolec GE Waukesha / Proprietary and Confidential

 Impulse Test
Lighting Impulse
• Front Time – 1.2 microseconds +/- 30% Tolerance (1.67 Times the time between 30% and 90% voltage)
• Tail Time – 50 microseconds +/- 20% (Time to 50% peak voltage)
• Chop Time > = 3 microseconds for > = 150 KV
> = 2 microseconds for < 150 KV

Prolec GE Waukesha / Proprietary and Confidential

 Waveform Comparisons – RFW & FW Overlay

Voltage Waveform Current Waveform

Prolec GE Waukesha / Proprietary and Confidential

 Impulse Generator (cont.)

Multiplier Circuit
 RC circuit with circuit Inductance
 Multiplier circuit introduced by
Prof. Marx
 Impulse capacitor Cs are charged in
parallel and discharged in series after
firing the switching gaps F
 Front Time T1 is determined by Rd
whereas time to half value is
determined by Re

Prolec GE Waukesha / Proprietary and Confidential

 Voltage and Current Wave Shape Comparison
(Reduce and Full Wave)
Chop wave failure Failure between LTC Leads

Turn to turn failure within

Cable to cable failure winding

Prolec GE Waukesha / Proprietary and Confidential

 Impulse Failure Waveforms

Prolec GE Waukesha / Proprietary and Confidential

 Switching Impulse Test
Switching Impulse Test C57.12.90 Sec. 10.2
• Time to peak value > 100 microseconds

• Time for 90 % of peak Value > 200 microseconds

• Time to first zero on tail of the wave >1000 microseconds

Test Circuit
• Test for each HV Line terminal

• Ground Neutral terminal for all Wye connection

• Ground other end of all Delta windings

• All Line terminals to be kept open except test terminal

Prolec GE Waukesha / Proprietary and Confidential

 Low Frequency Dielectric Test

Applied Voltage Test

• Transformer Connections
• Test Levels

Induced Voltage Test

• Transformer Connections
• Test Levels – Class I & Class II
• Partial Discharge

Prolec GE Waukesha / Proprietary and Confidential

 Induced Voltage Test vs. Applied Voltage Test
Test Voltage
• For Delta connected windings, applied test voltage level corresponds
to NSV

For example:
Equivalent applied test voltage for 230kV ( 750,825,900 BIL) is 345kV

• For Wye connected windings, the applied test voltage is limited to the
BIL of Neutral

For example:
If line end BIL is 550kV and neutral end BIL is 150kV, then equivalent
applied test voltage is limited to 50kV (equivalent for 150 BIL)

Prolec GE Waukesha / Proprietary and Confidential

 Induced Voltage Test – C57.12.90 Section:
10.7 to 10.9
Test Connection
• Three phase voltage is applied to LV terminals at frequency ≥2 times rated frequency; all other line
terminals are left open, Neutral and Tank is grounded

Test Voltage & Duration

Class I Transformers
• Test voltage is equivalent to twice the volts/turn and line end is raised to achieve equivalent power
frequency test voltage across phases

• Test duration is 7200 Hz; if test frequency is 180 Hz then test duration = 7200/180 = 40 seconds

• Test is considered to be passed if no collapse of voltage occurs or no audible internal sound is present

Prolec GE Waukesha / Proprietary and Confidential

 Induced Voltage Test (cont.)

Test Voltage & Duration

Class II Transformers
• Enhancement level – 173% maximum tap voltage for 7200 Hz
• One hour test voltage – 150% for 1 hour
• Partial discharge limits < = 250 pC

Prolec GE Waukesha / Proprietary and Confidential

 C57.12.00 – Table 4

Prolec GE Waukesha / Proprietary and Confidential

 Temperature Distribution Model
Winding
Top Hot Spot

Hot Spot
Gradient
R

Location
V
Ambient Oil Winding

W Avg. Winding
T D COM SER Avg. Gradient
V G

L Oil Rise 65C 65C

T Avg. Winding Rise
Hot Spot Winding Rise
C
80C
Bot.

24 Hour Max Avg. Ambient = 30C Temperature

(Max Ambient = 40C)

Prolec GE Waukesha / Proprietary and Confidential

 Temperature Rise Test – C57.12.90 Section: 11
• Measurements during temperature rise test
‒ Top oil temperature
‒ Ambient temperatures
‒ Top and bottom radiator temperature
‒ Hot winding resistance at shut down
• Top Oil Rise = Top oil temperature – Average ambient
• Mean oil rise = Top oil temperature – Average of top & bottom header temp
• Average winding rise
= { (Hot Resistance/Cold Resistance) X (234.5+ambient temp) } – Ambient
• Gradient = Average winding rise – Mean oil rise
• Hot spot Rise = Top oil rise + Hot spot gradient
• Hot spot gradient = Gradient (1 + k); k = hot-spot factoralculate

Prolec GE Waukesha / Proprietary and Confidential

 Average Winding Rise

Prolec GE Waukesha / Proprietary and Confidential

 DGA
• Before Test , After Dielectric, Before/After Temp Rise test, After all Tests
• Gassing rate depend on many factors - Winding temperature rise, Ambient
Temperature, Duration of test, Design characteristics like current and flux density
• Expect significant difference lab to lab .. Requires ASTM D3612 Method C

Gas Generation During

Limits per C57.130
Temp Rise Test
PPM/ Hour
Hydrogen H2 <1
Carbon Monoxide CO <2
Carbon dioxide CO2 < 18
Methane CH4
Ethane C2H6 < 0.4
Ethylene C2H4
Acetylene C2H2 0

Prolec GE Waukesha / Proprietary and Confidential

 Sound Test – C57.12.90 Section: 13
• Core audible sound: This sound component originates in the transformer core
• Load audible sound: This sound component is primarily produced by vibrations of the windings and tank walls
when the transformer is loaded.
• When a transformer is highly loaded, load sound can be a significant contributor to the total sound of the
transformer ,especially for low no-load noise medium and large power transformers.
• Cooling system audible sound: typically consists of broadband fan noise, plus discrete tones (of low levels) at the
fan blade passage frequency and its harmonics.
• The sum of core and cooling system sound components is typically referred to as the no-load noise of a
transformer.
• The total audible sound of the transformer, however, is the sum of all three components, 2015 standard outlines
measurement methods for Load sound and calculation to arrive Total sound.
• Sound levels are specified in NEMA-TR1 and that is only No-Load Sound Level
• Load Sound is not significant for smaller transformers ( < 100 MVA ) unless No load sound required is
below NEMA

Prolec GE Waukesha / Proprietary and Confidential

 Sound Test – Measurement
• Measurements are generally taken on a weighted scale as per NEMA standard
• Location of measurements start at drain plug and around the transformer at approximate 3
foot intervals, 12 inches away from transformer tank/radiators as applicable at 1/3 and 2/3
heights for transformers over 9 feet
• With fans running – readings are taken 6 feet distance

Prolec GE Waukesha / Proprietary and Confidential

 Other Tests
• Bushing Cap & PF – C1/C2
• Core excitation test – typically 110% for 24 Hours
• Leakage Reactance test
• Front of Wave Impulse testing
• Frequency response analysis (FRA)
• Fast Front Switching Impulse ( FFSI )
• LTC Tests
• Operate LTC at No Load Voltage
• Operate LTC under Load
• Dynamic Resistances
• DGA from LTC

Prolec GE Waukesha / Proprietary and Confidential

 Questions

Prolec GE Waukesha / Proprietary and Confidential