Wind Power Transformer Design

By Philip J Hopkinson, PE

Wind Power
Transformer at base of
Tower

Critical link to
successful Wind
Turbine connection to
the grid

Subject of Many
complaints
1. Gassing
2. Winding Failures
3. Contact coking
4. Arc-Flash
5. Low liquid level
3/29/2014 IEEE T&D Chicago 04/17/2014 1
 Wind Power Transformer Design
By Philip J Hopkinson, PE

1. Consideration focused on Generator step-up

Transformers from 600 volt to 34,500 volt
2. Inputs from field experience
3. Standards work within IEC TC 14 and IEEE
4. Arc-Flash Safety considerations from NFPA 70E 2009
5. Work with Manufacturers of Transformers
6. Analytical studies
7. Attempt to develop meaningful universal standard
requirements for transformers rated up to 10 MVA.
8. IEEE Transformer Working Group P60076-16
9. IEC Document 60076-16
10. Currently 100 members on IEEE Working Group

3/29/2014 IEEE T&D Chicago 04/17/2014 2

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Typical Wind Turbine

transformer

1. 3-Phase Pad mounted

2. 600 volt low voltage Gr Y
3. 34,500 volt high voltage
4. Commonly Wye high voltage
5. Sometimes Delta high voltage
6. Most with 5-leg wound cores
7. Some with 3-leg stacked cores
8. Nearly 100% liquid filled
9. Nearly all with sheet low voltage
10. All with wire high voltages

3/29/2014 IEEE T&D Chicago 04/17/2014 3

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Typical Wind Turbine transformer

concerns

1. High Hydrogen gas

2. High voltage winding failures
3. Safety of HV load-break switch
4. Carbonized HV Switches and Tap
changers
5. Low liquid levels in cold
temperatures

3/29/2014 IEEE T&D Chicago 04/17/2014 4

 Wind Power Transformer Design
By Philip J Hopkinson, PE

High Hydrogen Gas

1. Present in Up to 50% of
transformer populations
2. Hydrogen as high as 20,000 ppm
3. Small amounts Ethane and
Ethylene and Methane
4. Windings apparently OK
5. Most transformers returned to
factories pass Routine Tests,
including Impulse.

3/29/2014 IEEE T&D Chicago 04/17/2014 5

 Wind Power Transformer Design
By Philip J Hopkinson, PE

High Hydrogen Gas always partial

discharge related

1. But windings not involved

2. Leads not involved in most cases
3. Hot spots in windings OK
4. Some gassing improves with time
5. If not windings or leads then what
else?

3/29/2014 IEEE T&D Chicago 04/17/2014 6

 Wind Power Transformer Design
By Philip J Hopkinson, PE

What about the iron core?

1. Magnetic energy reservoir

2. Voltage generator from changing
flux
3. Many laminations of thin steel
4. Very thin inter-laminar insulation
5. High inter-laminar capacitance
6. Susceptible to static charge

3/29/2014 IEEE T&D Chicago 04/17/2014 7

 Transformer Core Grounds in Wound Cores
By Philip J Hopkinson, PE

1. The issue is gassing and Partial Discharge

2. Dielectric Breakdown in windings not the problem
3. Gassing and PD worst at 34.5 kV
4. Gassing not an issue at 5 kV
5. Gassing at 15 kV an annoyance only
6. Gassing coming from Core
7. Solution via grounding or shielding

3/29/2014 IEEE T&D Chicago 04/17/2014 8

 Transformer Core Grounds in Wound Cores
By Philip J Hopkinson, PE

Ground Lead

Vi Outer Core Vi Inner Core Vi Inner Core Vi Outer Core

C5 C2 C1 C3 C4 C1 C1 C4 C4 C1 C1 C4 C3 C1 C2 C3 C5

C5 C5 C5 C5

LV L Grounded Clamp

HV

3/29/2014 IEEE T&D Chicago 04/17/2014 9

 Transformer Core Grounds in Wound Cores
By Philip J Hopkinson, PE

Ground Lead

Vi Outer Core Vi Inner Core Vi Inner Core Vi Outer Core

C5 C2 C1 C3 C4 C1 C1 C4 C4 C1 C1 C4 C3 C1 C2 C3 C5

C5 C5 C5 C5

LV L Grounded Clamp

HV
C1-C5 calculated
HV = 34500 kVA = 1850 In EXCEL
LV = 600 Gr Y Frequency = 60 Hz

3/29/2014 IEEE T&D Chicago 04/17/2014 10

 Transformer Core Grounds in Wound Cores
By Philip J Hopkinson, PE

Vi Outer Core Loop Inside Surface

804 Volts
LV Start is Grounded
HV
0 Volts
C2 C1
L-G Volts
19919
Core Ground
C3 0 Volts
Case I, Core Ground Connected to Frame
Voltage drop in Outer Core stacks = 804 Volts
1. Outside core grounds at 34.5 kV result in high core volts
2. Inside core grounds eliminate static charges and drop core
volts to volt/turn levels

3/29/2014 IEEE T&D Chicago 04/17/2014 11

 Transformer Core Grounds in Wound Cores
By Philip J Hopkinson, PE

0.030" thick static shield

0.020" thick paper

Outside Core Leg

Core Ground

Grounded Electrostatic Shield

1. Outside core grounds at 34.5 kV result in high core volts
2. Inside core grounds eliminate static charges and drop core
volts to volt/turn levels
3. Core shields equally effective

3/29/2014 IEEE T&D Chicago 04/17/2014 12

 Transformer Core Grounds in Wound Cores
By Philip J Hopkinson, PE

With Outside Core Ground and no core shield:

1. Hydrogen at 15 kV typically 100-300 ppm
2. Hydrogen at 34.5 kV typically 3,000-10,000 ppm
3. Hydrogen accompanied by small amounts of
a. Ethane
b. Ethylene
c. Methane

3/29/2014 IEEE T&D Chicago 04/17/2014 13

 Transformer Core Grounds in Wound Cores
By Philip J Hopkinson, PE

With Outside Core Ground and no core shield:

1. Partial discharge best detector
2. PD should be conducted per Class II transformers
a. Start at 50% of rated volts
b. Go to 100% of rated volts and record
c. Go to 110% of rated volts and record
d. Go to 150% of rated volts and hold for 1 hour
e. Drop back to 110% of rated volts and hold for at
least 10 minutes and record.
f. Drop back to 100% of rated volts and record.
g. Drop back until pd extinguishes below 100 pc
3. Transformer fails test if extinguish Pd (< 100 pc)
occurs at less than 110% of rated volts

3/29/2014 IEEE T&D Chicago 04/17/2014 14

 Transformer Core Grounds in Wound Cores
By Philip J Hopkinson, PE

Conclusions:
1. Wound Cores most responsible for High
Hydrogen with Outside Core Grounds
2. Absolute Inside Core Grounds OK
3. Shielded Cores OK
4. 3-Leg Stacked Cores generally immune

3/29/2014 IEEE T&D Chicago 04/17/2014 15

 Transformer Core Grounds
By Philip J Hopkinson, PE

Recommendations:
Specify Core Precautions in Standard
1. For 5 Leg Wound Cores
a. Absolute inside core grounds or
b. Shielding
2. 3 Leg Stacked Cores
3. 4 and 5 leg stacked cores May need some shielding
Future Action:
Working with IEEE Transformers Committee to specify
isolation of cores from medium voltage windings in
IEEE C57.12.00.

3/29/2014 IEEE T&D Chicago 04/17/2014 16

 High Voltage Winding Failures from Switching
By Philip J Hopkinson, PE

Load-break Switching often

fails transformers

1. Switching often not done at

transformer
2. Groups of ~15 transformers
switched by vacuum breakers
3. First and / or Last transformer
in Group most vulnerable
4. Current Chopping and
Reignition Transients are
failure-initiators
5. IEEE C57.142 Addresses Issues
6. Resistor-Capacitor Snubbers
needed

3/29/2014 IEEE T&D Chicago 04/17/2014 17

 High Voltage Winding Failures from Switching
By Philip J Hopkinson, PE

Common Denominators for

switching-induced problems

1. Switching done at Vacuum

breaker
2. Transformers connected to
vacuum breakers by shielded
cables
3. No arresters at transformers..
but it wouldn’t help
4. Switched currents < 6 amps
5. Current chopping and
reignition transients present

Winding failures in tap section

or at line ends

3/29/2014 IEEE T&D Chicago 04/17/2014 18

 High Voltage Winding Failures from Switching
By Philip J Hopkinson, PE

3/29/2014 IEEE T&D Chicago 04/17/2014 19

 High Voltage Winding Failures from Switching
By Philip J Hopkinson, PE

Vacuum breakers compact, clean, efficient, generally safe..

But Transformers often in trouble!
3/29/2014 IEEE T&D Chicago 04/17/2014 20
 High Voltage Winding Failures from Switching
By Philip J Hopkinson, PE

Shielded Cables integral part

of damaging voltage
transients

1. Act like Transmission

lines
2. Wave velocity half the
speed of light
3. Surge impedance
independent of length
4. Low energy dissipation
5. EPR generally superior
to XLPE due to higher
dissipation

Shielded cables amplify resonances with reflected waves

and voltage doubling

3/29/2014 IEEE T&D Chicago 04/17/2014 21

 High Voltage Winding Failures from Switching
By Philip J Hopkinson, PE

Ping test #5 at Tap 4-5 & 9.5 amps Feb. 1. 06

Island Park Substation Unit #2 Load-break Switching
Line 1 to Gr. Voltage often fails transformers
With 20:1 Attenuators & Arc Gap@ 5.5"
3. Current decays and is chopped out
60
40
of conduction and voltage oscillates to
zero.
Current chopping
20 unavoidable at light
0
-200 -20 0 200 400 600 800 1000 currents < 6 amps
kV

-40
-60
-80 Reignition transients likely
-100 at low power factor
-120
Microseconds

1. Circuit breaker 2. Contacts reignite back into 5. Contacts

Circuit damping key to
4. Second TRV
contacts open and conduction, current rises to
peak and decays to chopping
voltage rises to (-)80 open
sufficiently to
solving problems
transient recovery
level, inducing oscillatory
kV, then breaker prevent
voltage (TRV) rises
reignites, raising
by Ldi/dt to (-)100 transient. Voltage rises by reignition
voltage by +120 kV current and
kV in~90 µsec. +145 kV in <1 µsec., then
oscillates to zero. in <1 µsec., etc. interruption is
completed

The higher the system energy efficiency the more likely are winding failures
3/29/2014 IEEE T&D Chicago 04/17/2014 22
 High Voltage Winding Failures from Switching
By Philip J Hopkinson, PE

Resistor-Capacitor Snubbers
provide needed damping

1. Can be mounted in
transformer
2. Can also be mounted in
switchgear
3. Current chopping will still
happen
4. Reignition transients will
totally disappear
5. Transformer survives!

3/29/2014 IEEE T&D Chicago 04/17/2014 23

 High Voltage Winding Failures from Switching
By Philip J Hopkinson, PE

Conclusion

R-C Snubbers in Vacuum breaker cabinet can prevent switching failures

Recommendation

Include brief tutorial and recommendation in the new Wind Power

Document P60076-16

3/29/2014 IEEE T&D Chicago 04/17/2014 24

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Thermal Stability of
Electrical Contacts in
switches and tap changers

1. Desired stable life for

30+years
2. Many electrical contacts
susceptible to oxidation
3. Key Variables are:
a. Contact materials
b. Contact pressures
c. Type of fluid
d. Current amplitude and
variability
e. Temperature
f. Worsened by high
current harmonics
3/29/2014 IEEE T&D Chicago 04/17/2014 25
 Wind Power Transformer Design
By Philip J Hopkinson, PE

Thermal Stability of Electrical

Contacts in switches and tap
changers

1. Functional life test being

documented in IEEE PC57.157
2. 30 day test with acceleration
of 1000 times
3. 2 XN Current daily for 8 hrs.
in 130 C bath followed by 16
hours cool down de-
energized.
4. Test passed if:
a. Stable
b. Resistance change < 25%

3/29/2014 IEEE T&D Chicago 04/17/2014 26

 Wind Power Transformer Design
By Philip J Hopkinson, PE

CuCu Average Resistance Thermal Stability of Electrical

700.0E-6
Contacts in switches and tap
changers
600.0E-6

500.0E-6
1. Functional life test being
Resistance in Micro-Ohms

documented in IEEE PC57.157

400.0E-6 2. 30 day test with acceleration
of 1000 times
300.0E-6
3. 2 XN Current daily for 8 hrs. in
200.0E-6 130 C bath followed by 16
hours cool down de-
100.0E-6
energized.
000.0E+0 4. Test passed if:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
a. Stable
Days
b. Resistance change < 25%
CuCu (C147) oil (1-9-96/2-15-96) CuCu (C147) sil (1-9-96/2-15-96) CuCu (C147) sil (3-28-96/6-1-96)
CuCu (C110) sil (10-23-97/1-21-98) CuCu FR3 (12-15-04/2-21-05) CuCu FR3 (4-18-05/6-27-05)

3/29/2014 IEEE T&D Chicago 04/17/2014 27

 Wind Power Transformer Design
By Philip J Hopkinson, PE

AgCu Average Voltage Drop

Thermal Stability of Electrical
Contacts in switches and tap
120.0E-3
changers
100.0E-3
1. Functional life test being
80.0E-3 documented in IEEE PC57.157
Voltage Drop in Volts

2. 30 day test with acceleration

60.0E-3 of 1000 times
3. 2 XN Current daily for 8 hrs. in
40.0E-3 130 C bath followed by 16
hours cool down de-
20.0E-3
energized.
4. Test passed if:
000.0E+0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
a. Stable
Days b. Resistance change < 25%
AgCu sil (10-7-98/12-16-98) AgCu fr3 (8-6-04/10-13-04) AgCu oil (12-15-04/2-21-05) AgCu oil (4-18-05/6-27-05)

3/29/2014 IEEE T&D Chicago 04/17/2014 28

 Wind Power Transformer Design
By Philip J Hopkinson, PE

AgAg Average Resistance Thermal Stability of Electrical

300.0E-6
Contacts in switches and tap
changers
250.0E-6

1. Functional life test being

documented in IEEE PC57.157
Resistance in Micro-Ohms

200.0E-6

2. 30 day test with acceleration

150.0E-6 of 1000 times
3. 2 XN Current daily for 8 hrs.
100.0E-6
in 130 C bath followed by 16
hours cool down de-
50.0E-6
energized.
4. Test passed if:
000.0E+0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 a. Stable
Days b. Resistance change < 25%
AgAg (Soft) sil (3-28-96/6-1-96) AgAg (Hard) sil (6-4-96/11-5-96) AgAg (Line 1) fr3 (8-6-04/10-13-04)

3/29/2014 IEEE T&D Chicago 04/17/2014 29

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Thermal Stability of
Electrical Contacts in
switches and tap changers

1. Test quite meaningful

2. 40+ years proven to find
stable contacts
3. IEEE PC 57.12.157 Guide
should be helpful to the
industry
4. Tests should be performed
by each manufacturer for
each switch type
5. These conclusions to be
included in Wind Power
Transformer Standard
P60076-16
3/29/2014 IEEE T&D Chicago 04/17/2014 30
 Wind Power Transformer Design
By Philip J Hopkinson, PE

Arc-Flash Concerns

1. NFPA 70E reference for

2009
2. Both HV and LV Cabinets
have concerns with LV
worst
3. Suit-up needed to just
check gauges
4. These conclusions to be
included in Wind Power
Transformer Standard
P60076-16

3/29/2014 IEEE T&D Chicago 04/17/2014 31

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Arc-Flash Concerns

Desire third cabinet

1. Gauges
2. Load-break switch
Handle
3. Schrader valve
4. Sampling port

3/29/2014 IEEE T&D Chicago 04/17/2014 32

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Arc-Flash Concerns

Desire HV Cabinet
Door to open first

3/29/2014 IEEE T&D Chicago 04/17/2014 33

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Low Liquid Level

Concerns

1. Meter shows level

below minimum
2. No leak evidence
3. Cold temperatures
4. Transformer may
be de-energized
5. Critical switchgear
and other
accessories may
not be immersed
in the liquid

3/29/2014 IEEE T&D Chicago 04/17/2014 34

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Low Liquid Level

Concerns

1. Volumetric change
of mineral oil with
temperature
Delta V =0.0007*(C-1)

2. Assume tank filled

at +70 C
3. Assume coldest
temperature (-) 40 C
4. Temperature change
is (-) 110 C
5. Volume shrink = -
7.6%
3/29/2014 IEEE T&D Chicago 04/17/2014 35
 Wind Power Transformer Design
By Philip J Hopkinson, PE

Low Liquid Level

Concerns

Suppose original liquid

height at 50”

A 7.6% volume
reduction is (-) 4”

Recommendation

Specify added liquid

level to fully cover
switches and accessories

3/29/2014 IEEE T&D Chicago 04/17/2014 36

 Wind Power Transformer Design
By Philip J Hopkinson, PE

Summary

IEEE P60076-16 should

be an important
document for Wind
Power Transformers

1. Normal C57.12.00
considerations
2. Core Grounds and
shielding
3. Resistor –Capacitor
Snubbers
4. Stable electrical
contacts
5. Arc-Flash issues
6. Added liquid level

3/29/2014 IEEE T&D Chicago 04/17/2014 37