wind power transformer design

Wind Power Transformer Design By Philip J Hopkinson, PE

3/29/2014 1

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

IEEE T&D Chicago 04/17/2014


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 2 IEEE T&D Chicago 04/17/2014


3/29/2014 3

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 4

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 5

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 6

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 7

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


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 8 IEEE T&D Chicago 04/17/2014


3/29/2014 9

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 Grounded Clamp

HV

L

L



3/29/2014 10

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 Grounded Clamp

HV

L

L

HV = 34500 LV = 600 Gr Y

kVA = 1850 Frequency = 60 Hz

C1-C5 calculated In EXCEL



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 11

Vi Outer Core Loop Inside Surface

804 Volts

LV Start is Grounded

0 VoltsC2 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

HV



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 12

0.030" thick static shield

0.020" thick paper

Outside Core Leg

Core Ground

Grounded Electrostatic Shield



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 13 IEEE T&D Chicago 04/17/2014


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 14 IEEE T&D Chicago 04/17/2014


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 15 IEEE T&D Chicago 04/17/2014

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 16 IEEE T&D Chicago 04/17/2014

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

3/29/2014 17

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 18

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 19 IEEE T&D Chicago 04/17/2014


3/29/2014 20

Vacuum breakers compact, clean, efficient, generally safe.. But Transformers often in trouble!



3/29/2014 21

Shielded cables amplify resonances with reflected waves and voltage doubling

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



3/29/2014 22

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

Island Park Substation Unit #2

Line 1 to Gr. Voltage

With 20:1 Attenuators & Arc Gap@ 5.5"

-120

-100

-80

-60

-40

-20

0

20

40

60

-200 0 200 400 600 800 1000

Microseconds

kV

1. Circuit breaker

contacts open and

transient recovery

voltage (TRV) rises

by Ldi/dt to (-)100

kV in~90 µsec.

2. Contacts reignite back into

conduction, current rises to

peak and decays to chopping

level, inducing oscillatory

transient. Voltage rises by

+145 kV in <1 µsec., then

oscillates to zero.

3. Current decays and is chopped out

of conduction and voltage oscillates to

zero.

5. Contacts

open

sufficiently to

prevent

reignition

current and

interruption is

completed

4. Second TRV

voltage rises to (-)80

kV, then breaker

reignites, raising

voltage by +120 kV

in <1 µsec., etc.

Load-break Switching often fails transformers Current chopping unavoidable at light currents < 6 amps Reignition transients likely at low power factor Circuit damping key to solving problems

The higher the system energy efficiency the more likely are winding failures



3/29/2014 23

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 24

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 25

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 26

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 27



of 1000 times 3. 2 XN Current daily for 8 hrs. in

130 C bath followed by 16 hours cool down de-energized.


CuCu Average Resistance

000.0E+0

100.0E-6

200.0E-6

300.0E-6

400.0E-6

500.0E-6

600.0E-6

700.0E-6

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

Days

Resis

tan

ce in

Mic

ro

-O

hm

s

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 28



of 1000 times 3. 2 XN Current daily for 8 hrs. in

130 C bath followed by 16 hours cool down de-energized.


AgCu Average Voltage Drop

000.0E+0

20.0E-3

40.0E-3

60.0E-3

80.0E-3

100.0E-3

120.0E-3

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

Days

Vo

ltag

e D

ro

p in

Vo

lts

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 29



of 1000 times 3. 2 XN Current daily for 8 hrs.

in 130 C bath followed by 16 hours cool down de-energized.


AgAg Average Resistance

000.0E+0

50.0E-6

100.0E-6

150.0E-6

200.0E-6

250.0E-6

300.0E-6

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

Days

Resis

tan

ce in

Mic

ro

-O

hm

s

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 30

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 31

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 32

Arc-Flash Concerns

Desire third cabinet 1. Gauges 2. Load-break switch

Handle 3. Schrader valve 4. Sampling port



3/29/2014 33

Arc-Flash Concerns

Desire HV Cabinet Door to open first



3/29/2014 34

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 35


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 36


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 37

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


wind power transformer design

Documents