TUTORIAL

Transformer Failure Due to Circuit Breaker

Induced Switching Transients

David D. Shipp, PE Thomas J. Dionise, PE

Fellow, IEEE Senior Member, IEEE
Eaton Electrical Group Eaton Electrical Group
130 Commonwealth Dr. 130 Commonwealth Dr.
Warrendale, PA 15086 Warrendale, PA 15086
[email protected] [email protected]

2011 I&CPS CONFERENCE 2011

NEWPORT BEACH, CA

MAY 1 – 4, 2011
 TABLE OF CONTENTS

1.0 INTRODUCTION.....................................................................................1-1

2.0 CASE STUDY OVERVIEW.....................................................................2-1

3.0 EMTP MODELING & CASE STUDIES ...................................................3-1

4.0 MOTOR STARTING AUTOTRANSFORMER CASE STUDY.................4-1

5.0 PT FAILURES.........................................................................................5-1

6.0 SNUBBER PERFORMANCE MEASUREMENTS ..................................6-1

7.0 SUMMARY..............................................................................................7-1

8.0 REFERENCES........................................................................................8-1

i
1.0 INTRODUCTION

Switching transients associated with circuit breakers have been observed for many
years. Recently this phenomenon has been attributed to a significant number of
transformer failures involving primary circuit breaker switching. These transformer
failures had common contributing factors such as 1) primary vacuum or SF-6
breaker, 2) short cable or bus connection to transformer, and 3) application involving
dry-type or cast coil transformers and some liquid filled.

This tutorial will review these recent transformer failures due to primary circuit
breaker switching transients to show the severity of damage caused by the voltage
surge and discuss the common contributing factors. Next, switching transient
simulations in the electromagnetic transients program (EMTP) will give case studies
which illustrate how breaker characteristics of current chopping and re-strike
combine with critical circuit characteristics to cause transformer failure.

Design and installation considerations will be addressed, especially the challenges

of retrofitting a snubber to an existing facility with limited space. Finally, several
techniques and equipment that have proven to successfully mitigate the breaker
switching transients will be presented including surge arresters, surge capacitors,
snubbers and these in combination

1-1
2.0 CASE STUDY OVERVIEW

2-1
 Transformer Failure Due to Circuit Breaker
Induced Switching Transients
Part 1: Case Study Overview
David D. Shipp, PE Thomas J. Dionise, PE
Fellow, IEEE Senior Member, IEEE
Eaton Electrical Group Eaton Electrical Group
130 Commonwealth Dr. 130 Commonwealth Dr.
Warrendale, PA 15086 Warrendale, PA 15086

Case Study Overview:

Data Center Transformer Failure
Attributed to
Circuit Breaker Induced
Switching Transients
System Configuration, Failure Analysis
and RC Snubber Fix

R. McFadden – JB&B – NY, NY

D. Shipp – Eaton Electrical – Pgh, PA

Presentation Purpose/Summary

• Purpose #1 Safety:
– Alert Design & User Community
– Drive Solutions Through Market Interest
• Unique Case History
– Forensic Evidence
– High Speed Power Quality Capture
• Presentation Summary
– Simplified System Overview
– Forensic Review
– Power Quality Snapshots
– Installed RC Snubber Fix
• Presentation Summary: David Shipp, Eaton
– Phenomenon Detailed Explanation
– Science of RC Snubber Design
 Site Specifics

• Utility Service:
– 26kV
– Double Ended Loop Through
• Transformers
– (6) Total
– 26kV Primary, 480V Secondary
– VPI
– 3000kVA AA / 3990 kVA FA
– 150 kVbil
• Switching Device
– Vacuum Circuit Breaker
• Cable
– 35kV, 133% EPR Insulation, 1/3 Concentric Ground

Simplified System Configuration

26kv Utility “A” 26kv Utility “B”

M M

40’ 40’

T-A3 T-B3
CRACs CRACs
80’ Chillers Chillers
80’

T-A2 T-B2
UPS UPS
100’+ To Normal Side 100’+
Of 480vac ATS
T-A1 (Typical) T-B1
OTHER OTHER

Failure/Sequence of Events
• All Transformers Fully Tested:
• Pre-functional: Turns Ratio, Insulation Resistance, etc
• Functional:UPS Full Load Tests, UPS Transient Tests, Data Center
Room Validation Testing
• Final Pull The Plug Test (PTP)
• 4 Electricians “simultaneously” open (4) 26KV vacuum breakers to
simulate a general utility outage.
• All systems successfully transfer to standby generation but:
• Loud Pop heard in Substation Room B
• Relay for VCB feeding TB3 signaling trip
• Decision made to shutdown generator test and investigate issue in
“B” substation room
• 2 Electricians “simultaneously” close (2) 26KV vacuum breakers to
Substation Room “A”
• Transformer TA3 fails catastrophically.
Pre-PTP Condition
26kv Utility “A” 26kv Utility “B”

M M

40’ 40’

T-A3 T-B3
CRACs CRACs
80’ Chillers Chillers
80’

T-A2 T-B2
UPS UPS
100’+ 0.8MW 0.9MW 100’+

T-A1 T-B1
OTHER OTHER
1.5MW 1.3MW

Failure #1: De-Energization During PTP

26kv Utility “A” 26kv Utility “B” Relay
Signaling
Trip
M M

40’ 40’

T-A3 T-B3
CRACs CRACs
80’ Chillers Chillers
80’

T-A2 Audible Pop T-B2

UPS UPS
100’+ In Substation 100’+
Room
T-A1 T-B1
OTHER OTHER

Failure #2: Energization During PTP

26kv Utility “A” 26kv Utility “B”
Relay Trips
VCB M M

40’ 40’

T-A3 T-B3
CRACs CRACs
80’ Chillers Chillers
80’

T-A2 Catastrophic T-B2

UPS Failure UPS
100’+ 100’+

T-A1 T-B1
OTHER OTHER
Failure #2
Transformer Failure On Energization

Coil to Coil Tap

Burnt Off

Failure #2
Transformer Failure On Energization

Suspected Area
of Initial Flash

Coil to Coil Conductor

Burnt Off

Failure #2
Transformer Failure On Energization
Coil to Coil Tap
Burnt Off
Failure #2
Transformer Failure On Energization
Coil to Coil Tap
Burnt Off

Upward Twist
From Lower Blast

Failure #1
Transformer Failure On De-Energization
Flash/Burn Marks

Failure #1
Transformer Failure On De-Energization
Close up of
Flash/Burn Marks

Coil to Coil not

Winding to Winding

This Transformer was Returned

to the Factory and Passed all
Standard IEEE Tests!
Failure #1
Transformer Failure On De-Energization
15KV Cable
In 26KV
Transformer

No Coil to Coil
Cable Supports

Some Burn
Marks Match
Point of
Cable Contact

Failure #1
Transformer Failure On De-Energization
Some Burn
Marks Match
Vicinity of Cable

Cable is Undamaged

Unaffected Transformer

Factory Installed
Cable Support
Current Transient: Failure #2
Transformer Energization
Second Transformer
Inrush
AØ
CØ

BØ

Event Initiated
A-B Phase 19,000 amp, 3 Phase
5 Cycle CB Clearing Time

Current Transient: Failure #1

Transformer DeEnergization
8000amp A-B
Phase
Approx ¼ Cycle
BØ

AØ C Phase Shows
2nd Xfmr Load
& Event Duration

CØ

Breaker Induced
Switching Transients
 Case 1
Hydro Dam, MT 2005
• MV Vac Bkr Replacements Vendor “A”
• 13.8 kV
• 20 feet of cable
• 50 kV BIL (W) ASL Dry Type Txmrs
• Customer Energized before Vendor OK
• Txmr Failed
• No Surge Protection Applied

CASE 2
Cleveland Hospital 3/06
• Vacuum Breakers – Vendor “A” and Vendor
“C”
• 13.8 kV
• 95 kV BIL
• Dry Type Txmr
• 27 feet of Cable
• Bkr Manufacturer Paid to Repair Failed 2500
KVA Txmr
• Surge Protection Added Afterwards

CASE 3
RAILROAD SUBSTATION 11/06
• Vacuum Breaker – Vendor “A”
• 26.4 kV
• 150 kV BIL
• Generic Liquid Filled Rectifier Txmr
• 37 feet of Cable
 CASE 4
NJ DATA CENTER 12/06
• 26.4 kV – Vendor “B”
• 4 Txmrs Switched Under Light Load
• 2 Txmrs Failed-1 on Closing/1 on
Opening
• 40 Feet of Cable
• 2 other Txmrs Did Not Fail - 80 Feet of
Cable
• Arresters Were Present

CASE 5
OIL FIELD – AFRICA 6/07
• Vacuum Breaker – Vendor “D”
• 33 kV
• 7 Feet of Cable
• Dry Type Txmr in 36 Pulse VSD
• Arresters Were Applied
• Txmr Failed Upon Energization

CASE 9
Oil Drilling Ship – 6/2002
• Vendor “A” IEC Vac Bkrs in Vendor “D”
Swgr
• 11kV, 60 HZ
• Cast Coil Dry Type IEC VSD Propulsion
Txmr Failed (7500 kVA) – 75 kV BIL?
• < 30 feet of Cable
• Fed from Alternate Bus – Now 80 feet of
Cable
• No further Failures Reported
 CASE 11
Hospital – Slide 1
• 13. 8 kV System
• Vendor C Breakers/
Vendor F Txmrs
• 2 Txmrs Failed During
Commissioning
• Differential Relay
Tripped During motor
Starting – under
Highly Inductive Load

CASE 11
Hospital – Slide 2

CASE 11
Hospital – Slide 3
 CASE 12
Motor Starter Auto Txmr Failure
• 4160V
• 5000 HP
• Reduced Voltage
Auto Txmr Failed
• Arresters on Wye
Point
• Internal Resonance
• Layer Wound
• Failed Layer to Layer

CASE 12 Run Contactor Closes

3516 HZ

CASE 12 - Sweep Frequency Test

4500 HZ (Admittance)
CASE 12 - Run Contactor Closes
– 844 HZ With Snubber

SWITCHING TRANSIENTS DUE TO

VACUUM / SF6 BREAKERS
• Opening -- Current Chop
• Closing -- Prestrike/Re-ignition/Voltage
Escalation
• Vacuum Bkrs --Both Closing and
Opening
• SF6 -- Opening
• Air -- Generally Acceptable

Current Chop
 Current Chop
Current Chop in Vacuum is a Material Problem

Current Chop

• All Types of
Interrupters Chop
Current
• This is not a Unique
Feature of Vacuum

Switching Inductive Circuits

• Closing
• Opening
• Voltage Escalation
• Surge Suppression
Switching Inductive Loads
CLOSING

Switching Inductive Loads

OPENING

Switching Inductive Loads

VOLTAGE ESCALATION
SWITCHING TRANSIENT THEORY

• “Thou Shalt Not Change Current

Instantaneously in an Inductor”
• Conservation of Energy –
– You Cannot Create or destroy Energy –
You Can Only Change Its Form

ENERGY EQUATION

1 2 1
LI  CV 2
2 2
OR

L
V I
C

TOTAL VOLTAGE

• Vt = Venergy + Vdc + Vosc

• Venergy is from the Energy Equation

• Vdc = DC Off-set that Sometimes is
Present
• Vosc = the Oscillatory Ring Wave
 TRANSFORMER LIMITS

• Magnitude – BIL Ratings

• Dv/dt Limits
• Both MUST Be Met
• Dry Type Txmrs Particularly
Susceptable
• Liquid Filled Not Immune.
• Consider “Hammer Effect”

ANALYSIS

• When Open Bkr, Txmr is Left

Ungrounded
• Ring Wave is a Function of its Natural
Frequency

1
NF 
2 LC

CASE 3
Waveforms Without Snubbers
 CASE 3
Current – Without Snubbers

Ichop

CASE 3
Waveforms With Snubbers

CASE 4
Data Center 12/06
• 26.4 kV
• Vendor “B” Breakers
• 4 Bkrs Switched at Once
• 2 Dry Type Txmrs Failed (40 Ft of
Cable)
• 2 Txmrs Did not Fail (80 Ft of Cable)
• Unfaulted Txmr Winding Failed BIL
@162 kV ( Rated 150 kV)
 CASE 4
Data Center, NJ 12/06 - W/O Snubbers

v [kV] 08-Jan-07 14.53.47

200

150

100

50

-50

-100

-150

-200

0 5 10 15 20 25

(42) XFMA3C (48) XFMB3C t [ms]

Plot of the Transformer A3 and B3 primary voltage Phase C

-207 kV
Open circuit breaker with chopped current of 9 Amperes

CASE 4
Data Center, NJ 12/06-With Snubbers

v [kV] 09-Jan-07 10.50.25

40

30

20

10

-10

-20

-39 kV
-30

-40
0 5 10 15 20 25

(29) XFMA3B (35) XFMB3B t [ms]

Plot of the Transformer A3 and B3 primary voltage Phase B

Open circuit breaker with chopped current of 9 Amperes
Application of the R-C snubber circuit

CASE 6
Data Center 2, NJ 12/06
• 13.2 kV
• Vendor “B” Breakers
• 3 MVA Dry Type Txmrs
• 60 ft Cable – Required Snubbers
• 157 ft Cable – Required Snubbers
• No Problem at Startup
 CASE 6
13.2 kV SYSTEM – 119 kV @ 678 HZ

CASE 6
WITH SNUBBERS – 33.6 kV @ 236 HZ

CASE 7
Chemical Plant, NC 3/07
• 12.47 kV System
• 20+ Year Old Oil Filled Txmrs
• Vendor “A” Vacuum Bkrs retrofitted on
Primary
• 10 Feet of Cable
• No Problem at Startup
 CASE 7
425 kV - 12. 47 kV - 23 kHZ Ring Wave

CASE 7
Added Snubbers – 12.47 kV

CASE 8
DATA CENTER – Indiana 6/07
• 12.47 kV System
• Vendor “A” Breakers
• 270 Feet of Cable
• No Additional Surge Protection
Required
 CASE 8
-55 kV at 800 HZ

CASE 11 Hospital
186.9 kV / 2000 HZ

CASE 10 – Snubber
Reduced to 33.8 kV / 368 HZ
 MITIGATING TECHNIQUES

• Arresters
• Surge Capacitors
• Snubbers (RC)
• Hybrid / Combinations
• Liquid vs Dry Type
• Electronic Zero Crossing Switching

Snubber - 3 Phase Capacitor

Generally Solidly Grounded System
• Paper Mill - AL
• 13.8 kV

Snubbers – 2nd paper mill - beta site

 Snubber – Single Phase Capacitors
Low Resistance Grounded Systems
• Paper Mill
• Vendor “A” Swgr
• 13.8 kV

Snubber – Single Phase Capacitors

• Silicon “Chip” Plant

• Montana
• Very Specialized
Dry Type Txmrs
• 13.8 kV
• Cables < 100 feet
• Primary Fused
Switch AF Solution
• Vendor “A” Vac Bkrs

RC Snubber Installed – Case 4

Lightning
Arrestor

Fuse

Blown Fuse
Viewing Window

Resistor

Capacitor
RC Snubber Installed – Case 4

Resistor

Capacitor

RC Snubber Installed – Case 4

Fuse

Blown
Fuse
Indicator

SWITCHING TRANSIENT STUDY

• Quantifies Problem
• Predict Exposure / Risk
• Select Best / Most Cost Effective
Solution
• Do “What if” Cases
• Verify Results
 RECOMMENDATIONS

• Factor into Design Up-front

• Do Study – Results Are Bkr Manufacturer
Specific
• Use Protection Only When / Where Needed
(if not there, cannot fail)
• Fused or Unfused Snubbers??
• Loss of Fuse Detection??
• Fear Not! - Mitigating Techniques Have Been
Proven.
• Discrete Snubber Components??

Conclusion/Next Steps
• This is a System Problem
– Transformer, Cable, Switching Device, Proximity
– Statistical Event , Possible Undetected Failures
• Data Centers Fall into the Highest Risk Category
– High Power Density
– Close Proximities
– Frequent Switching
• Draft IEEE C57.142
– A Guide To Describe The Occurrence and Mitigation Of Switching Transients
Induced By Transformer And Switching Device Interaction
– Does not accurately warn users of all areas of concern
– This case did not meet the areas of concern noted in Draft C57.142
– Need to push for formalization of the standard with new lessons learned
• RC Snubbers
– Transformer Manufacturer appears best positioned to implement the solution
– Limited Cataloged Product (One Manufacturer)
– Not embraced by all transformer manufacturers
– Design Parameters of RC Snubber not clearly defined
• Lives, Property and Uptime are all at risk

QUESTIONS ?
3.0 EMTP MODELING & CASE STUDIES

3-1
Transformer Failure Due to Circuit Breaker
Induced Switching Transients
Part 2: EMTP Modeling and Case Studies
David D. Shipp, PE Thomas J. Dionise, PE
Fellow, IEEE Senior Member, IEEE
Eaton Electrical Group Eaton Electrical Group
130 Commonwealth Dr. 130 Commonwealth Dr.
Warrendale, PA 15086 Warrendale, PA 15086

Transient Analysis Tools

• Electromagnetic Transients Program (EMTP)

• Developed by Hermann Dommel brought to BPA
• Major contributors: Scott Meyer & Dr. Liu at BPA
• Alternative Transients Program (ATP)
• Alternative to commercialized EMTP
• Free to all if agree not to commercialize it
• EMTP-RV
• PSCAD EMT/DC

2
2

ATP

• ATP Draw interface

• 3-phase modeling
• Solution method - Trapezoidal integration
• Robust for wide variety of modeling needs
• Extensive library of models
• Many options and features

3
3

1
 Source Model

U RUTIL XUTIL UT
• “3-phase wye cosine”
• Resistance in ohms
• Inductance in mili-Henries UN
• Source Impedance VUTIL

• Z1 and X/R
• Z0 and X/R

SYSTEM SOURCE
AT 13.8 KV

4
4

Cable Model
C1 LCABLE RCABLE C2
• “Pi Model”
• Resistance in ohms
• Inductance in mili-Henries
• Cable charging in micro-Farads
• ½ charging on sending end
• ½ charning on receiving end C/2 C/2
• Multiple pi models in some cases
CABLE
13.8KV

5
5

Breaker Model

• “Switch”
• Time dependent switch
• Open or close at a specific time
• Opens at current zero unless specify otherwise
• Current dependent switch
• Needed for current chopping
• Define current at which to open switch
• Independent Poles (A, B and C independent)
• Data request
VCB
• Vacuum or SF6 breaker nameplate BKR
• Current chop (ask manufacturer)
• Transient Recovery Voltage Ratings – T2, E2 and RRV

6
6

2
Transformer Model
T1 LTRAN RTRAN N:1 T2
• “Three Phase Model”
• Resistance in ohms
• Inductance in mili-Henries
• Magnetizing current
• Winding capacitance
• CH and CL for dry-types
• CHL for oil-filled
• More detailed modeling
• Saturation
• Hysteresis CH CL
• Data Request RLRG
• %Z and X/R, MVA
• BIL
TRANSFORMER

7
7

Complete System Model

U RUTIL XUTIL UT C1 LCABLE RCABLE C2 T1 LTRAN RTRAN N:1 T2

UN
VUTIL

C/2 C/2 CH CL
RLRG

SYSTEM SOURCE VCB CABLE

TRANSFORMER
AT 13.8 KV BKR 13.8KV

8
8

Time Step

• Time Step – Integration Time Step

• Choice depends on frequency of expected transient
• Too large – miss high frequency effects
• Too small – excessive simulation times
• Nyquist Criteria
• Minimum sample rate = 2 x frequency
• Example 10-20KHz ring for VCB transient

9
9

3
Switching Transient Simulations

• Switching transient simulations

• EMTP and model requirements
• Case studies
• Data Center
• Ferry Propulsion
• Laddle Melt Furnace
• Successful techniques/solutions

10
10

Ferry Propulsion

Highlights
• 60K passengers per day
• 20M passengers per year
• 4160V distribution
• 3 x 2865kW diesel gens
• 2 x 1865KW forward propulsion motors
• 2 x 1865KW reverse propulsion

11
11

Problem

• Forward propulsion drives ferry

• Upon reaching dock, reverse propulsion stops ferry
• July 1, 2009, reverse propulsion was lost entering dock
• lost power and hit a pier at full speed
• one serious injury and nine minor injuries
• 750 to 800 passengers were aboard
• impact did not send any passengers overboard
• 1185KVA rectifier transformer failed on reverse propulsion
motor
• Evidence of insulation failure on first few turns of primary
winding
• Investigation into source of such failure Î VCB switching

12
12

4
Ferry - Oneline

Vacuum
Breaker

Short
Cable

Dry-Type
Transformer

Forward Reverse

13
13

Ferry – Electrical Highlights

• 3 x 2865kW diesel generators

• 4160V, 3-bus system
• Vacuum circuit breakers – 630A mains, ties,
feeders
• 2 x 1865kW motors (forward propulsion)
• 2 x 1865kW motors (reverse propulsion)
• 8 x 1185KVA dry-type transformers, 30kV BIL
• 3-winding rectifier transformers
• 12-pulse effective (6-pulse VFD per winding)
• Feeder cable lengths of 50 feet each
14
14

Worst Case Scenarios

• Feeder cable lengths of 50 feet

• Each 1185KVA transformer has one 4160V feeder
• Need to examine both feeders for each transformer
• “Worst Case” Switching Transient Simulations
• Close 4160V VCB to transf. with 50ft. cable (model validation)
• Close 4160V VCB to transf. with 50ft. cable (re-ignition)
• Open 4160V VCB to transf. with 50ft. cable (current chop)
• Repeat each case with Snubber

15
15

5
Measurements
Transient Overvoltages at Primary of Rectifier Transformer – VCB Closes

16
16

Study Cases

• Case 1 – Closing the 4.16 kV feeder breaker feeding the three

winding rectifier transformer with the distance of 50 feet.
Simulated to match the Siemens measurements to ensure that
the computer model is accurate. (Model Validation)
• Case 2 – Same as Case 1, except with the RC snubber.
• Case 3 - Closing the 4.16 kV feeder breaker feeding the three
winding rectifier transformer, then the 4.16 kV feeder breaker
opens during the inrush current. Shows the possibility of the
vacuum breaker re-ignition. (Re-ignition)
• Case 4 - Same as Case 3, except with the RC snubber.
• Case 5 - Open the 4.16 kV feeder breaker feeding the three
winding rectifier transformer under the light load condition.
(Current Chop)
• Case 6 – Same as Case 5, except with the RC snubber.

17
17

Case 1 – Model Validation

Transformer primary voltage

V max of 4.96kV < 30kV BIL

Oscillation of 20,203Hz > 1000Hz

18
18

6
Case 2 – Valid Model with Snubber
Transformer primary voltage

V max of 3.982kV < 30kV BIL

Oscillation of < 1000Hz

19
19

Case 3 – Re-ignition
Transient Recovery Voltage (TRV) at VCB

20
20

Case 4 – Re-ignition with Snubber

Transient Recovery Voltage (TRV) at VCB

21
21

7
Case 5 – Current Chop
Transformer primary voltage
V max of 31.9kV > 30kV BIL
Oscillation of 958Hz ~ 1000Hz

22
22

Case 6 – Current Chop with Snubber

Transformer primary voltage

V max of 8.9kV < 30kV BIL

Oscillation of 299Hz < 1000Hz

23
23

Summary of Current Chop Cases

24
24

8
Summary of Re-Ignition Cases

25
25

Recommendations

• Install snubber at primary of each 1185KVA

rectifier transformer
• 40ohm resistor
• Non-inductive
• Peak voltage – 6 kVpeak
• Peak energy – 2100 Joules
• Average power – 190 Watt
• 0.5uF capacitor
• Rated voltage - 4.16kV

26
26

Data Center
Highlights
• Tier III
• LEEDS Certified
• 12.5MVA Capacity
• 13.2KV Ring Bus

27
27

9
Data Center - Oneline

Vacuum
Breaker

Short
Cable

Cast-Coil
Transformer

28
28

Data Center – Electrical Highlights

• 2 x 24.9 kV lines from Factory Shoals and Buzzard

Roast
• 2 x 12.5 MVA transformers
• 13.2 kV “ring-bus” configuration
• MSA and MSB and generator bus GPS
• Vacuum circuit breakers – 600A mains & ties, 200A feeders
• 3 x 2250 KW generators
• 6 x 3750KVA cast coil transformers, 90kV BIL
• 3 x DE Subs CSA, CSB and CSC
• 3 x SE Subs MDSA, MDSB and MLB
• Feeder cable lengths vary from 109 to 249 Feet

29
29

Worst Case Scenario

• Feeder cable lengths vary from 109 to 249 Feet

• Each 3750KVA transformer has two 13.2kV feeders
• Need to examine both feeders for each transformer
• Shortest of all 13.2kV cable runs to 3750KVA
Transformers
• MSB to Transformer CSC – 109 feet
• MSA to Transformer CSC – 111 feet
• “Worst Case” Switching Transient Simulation
• Open 13.2 kV VCB at MSB to CSC with 109ft. cable
• Open 13.2 kV VCB at MSA to CSC with 111ft. cable

30
30

10
Worst Case – Study Case 13

• Study Case 13 - Open the 13.2 kV feeder breaker at

Bus MSB feeding the 3,750 kVA dry type transformer
CSC with the shorter distance of 109 feet. The result
for this case will represent the “worst-case” condition,
the other feeder from Bus MSA has a longer feeder
distance of 111 feet.
• Study Case 14 – Same as Case 13, except with the
application of the 0.25 μF surge capacitor.
• Study Case 15 – Same as Case 13, except with the
application of the snubber circuit with 30ohm and 0.25
μF surge capacitor.
31
31

Case 13 – no surge protection

Load current at 13.2kV VCB

Load current of 10A

Chopped current of 6A

32
32

Case 13 – no surge protection

Transformer CSC primary voltage

VC max of 65.3kV > 95kV BIL

VB max of 116kV > 95kV BIL
VA max of 123kV > 95kV BIL
Oscillation of 969Hz > 1000Hz

33
33

11
Case 14 – 0.25uF surge cap
Load current at 13.2kV VCB

Load current of 10A

Chopped current of 6A

34
34

Case 14 – 0.25uF surge cap

Transformer CSC primary voltage

VA max of 29.4kV < 95kV BIL

VB max of 19.1kV < 95kV BIL
VC max of 15.7kV < 95kV BIL
Oscillation of 215Hz < 1000Hz

35
35

Case 15 – snubber 30ohm and 0.25uF

Load current at 13.2kV VCB

Load current of 10A

Chopped current of 6A

36
36

12
Case 15 – snubber 30ohm and 0.25uF
Transformer CSC primary voltage

VB max of 28.6kV < 95kV BIL

VA max of 19.4kV < 95kV BIL
VC max of 15.9kV < 95kV BIL
Oscillation of 215Hz < 1000Hz

37
37

Case 15 – snubber 30ohm and 0.25uF

Snubber Current – important for duty on resistor and capacitor

IC peak of 7.7A
IB peak of 8.0A
IA peak of 6.8A

38
38

Comparison of Results

No
Snubber

0.25uF
surge cap

Snubber
R=30ohm
C=0.25uF

39
39

13
Measured Transients with Snubber

40
40

Recommendations

• Install snubber at primary of each 3750KVA cast coil

transformer 30ohm and 0.25uF
• Install surge caps 0.25uF at each emergency
generator
• Install surge arresters at the following locations:
• both incoming power transformers
• every distribution dry type transformer
• every generator
• line side of both main circuit breakers
• line side of the three generator circuit breakers
• load side of every feeder breaker and every tie circuit breaker

41
41

Laddle Melt Furnace

Highlights
• Retiring 3 x 38MW EAFs
• Adding 1 x 155MVA EAF
• Adding 2 x 138kV lines to new EAF
• Adding SVC at 34.5kV
• Adding 1 x 20MW LMF

42
42

14
LMF & EAF - Oneline

Vacuum
Breaker

Short
Bus

Existing Oil-Filled
EAF New Transformer
LMF
43
43

UTILITY
4713MVA 3PH SC
9.26 X/R
LMF 138KV

1600A

50/66/83MVA
135.3/26.4KV
7.5%Z
13OHM
SF6 27KV
Breaker SF-6 BREAKER
2000A

Short ALUMINUM
Bus IPS BUS
53FEET

Oil-Filled AUTO LTC

Auto-Regulating 56MVA
27-10KV
Transformer 3.3%Z Vacuum
VACUUM BREAKER
1200A Breaker

HEAVY DUTY
COPPER PIPE Short
28FEET Bus
LMF XFMR
50/56MVA
25/.53KV Oil-Filled
2.5%Z Furnace
LMF Transformer
New 20MW
LMF
44
44

LMF Circuit – Electrical Highlights

• 50MVA Power Transformer 135/26.4kV

• 27kV system
• SF6 circuit breaker – 2000A
• Bus length of 53 feet
• 56MVA autoregulating transformer, 200kV BIL
• Vacuum circuit breaker – 1200A
• Bus length of 28 feet
• 50MVA oil-filled LMF transformer, 200kV BIL
• 20MW LMF
45
45

15
Worst Case Scenarios

• Examine switching transients at both transformers

• “Worst Case” Switching Transients for Auto-Reg Transf
• Open SF6 breaker to transf. with 53ft. bus (6A current chop)
• 3 cycles after energizing Auto-Reg Tran, open SF6 bkr to transf.
with 53ft. bus (re-ignition) Í highly inductive current
• “Worst Case” Switching Transients for LMF Transf
• Open VCB to transf. with 28ft. bus (6A current chop)
• 3 cycles after energizing LMF Tran, open VCB to transf. with
28ft. bus (re-ignition) Í highly inductive current
• Repeat each case with Snubber

46
46

Case 1 – Open VCB

LMF Transformer primary voltage

V max of 386kV > 200kV BIL

Oscillation of 1217Hz > 1000Hz

47
47

Case 2 – Open VCB with Snubber

LMF Transformer primary voltage
V max of 56.4kV < 200kV BIL
Oscillation of 200Hz < 1000Hz

48
48

16
Case 3 – Open SF6 Breaker
Auto-regulating Transformer primary voltage

V max of 23kV < 200kV BIL

No oscillating frequency

49
49

Case 4 – Open SF6 Breaker with Snubber

Auto-regulating Transformer primary voltage

VB max of 54.7kV < 200kV BIL

Oscillation of 197Hz < 1000Hz

50
50

Case 5 – VCB Re-ignition

Transient Recovery Voltage for VCB

51
51

17
Case 5 – Highly Inductive Current
Inrush current to LMF transformer

6000A Peak
Inrush

VCB
Closes
VCB
Opens

52
52

Case 6 – VCB Re-ignition & Snubber

Transient Recovery Voltage for VCB

53
53

Case 7 – SF6 Breaker Re-ignition

Transient Recovery Voltage for Siemens SF6

54
54

18
 Case 7 – Highly Inductive Current
Inrush current to Auto-Regulating transformer

14000A Peak
Inrush

SF6 CB
Closes

SF6 CB
Opens

55
55

Case 8 – SF6 Breaker Re-ignition & Snubber

Transient Recovery Voltage for SF6 Breaker

56
56

Summary of Current Chop Cases

57
57

19
Summary of Re-Ignition Cases

58
58

Recommendations

• Install snubber (100ohm and 0.15uF) at primary of

each transformer
• 100ohm resistor
• Non-inductive
• Peak voltage – 38 kVpeak
• Peak energy – 17,500 Joules
• Average power – 1000 Watt
• 0.15uF, 34.5kV surge capacitor (not available)
• 1-pole, 24kV, 0.13uF
• 2-pole, 14.4kV, 0.5uF
• Series combination gives 0.103uF

59
59

20
4.0 MOTOR STARTING AUTOTRANSFORMER CASE STUDY

4-1
Medium Voltage Reduced Voltage Autotransformer
Starter Failures
– Explaining The Unexplained –

Arthur J. Smith III, Member IEEE Larry Farr, Senior Member IEEE
Vice President Principal Engineer
Waldemar S. Nelson and Co. Inc. Eaton Electrical
New Orleans, LA Asheville NC

Technical Advisor TC-17A IEC

Chair of CANENA TC-17A SC-3

Medium Voltage Reduced Voltage Autotransformer

Starter Failures
– Explaining The Unexplained –

TODAY’S AGENDA
• Background
• Investigation
• Testing 1988
• Findings
•Testing 2002-03
• Findings
• Solution
• Conclusions

Medium Voltage Reduced Voltage Autotransformer

Starter Failures
– Explaining The Unexplained –

• Background

1
 Background

• For the past century the Auto-

transformer or “Korndorfer”
Starter has been a standard in
the electrical industry.

Background

• In the late 1970s, unexplained

“High Voltage Stress” failures
started occurring.

2
 Background
Failures Were Occurring Worldwide
• North Sea Platform 12kV starters failed multiple
times. Problems solved with single phase
autotransformers.
• 20,000 hp-15kV starter in British Columbia failed
four times until surge arresters from the Zero tap to ground
were installed.
• South America 2400 volt failed multiple times.
• Gulf of Mexico – Multiple 5kV starter failures
• IEC 60470 Clause 6.102.7 Change-over Ability Test
• Off the West Coast of Africa – Multiple 5kV starter failures

Background
What Happened?
Zero Tap Circuit to Ground

3 Inches !!

Background
Failure Modes

High Voltage Flashover

Telescoping Coils
Layer to Layer

3
 Background
High Voltage Flashover
Tap to Tap Over the Surface

Background
High Voltage Flashover
Coil to Ground

Background
High Voltage Flashover
Layer to Layer Failure

4
 Background
High Voltage Flashover
Zero-tap Circuit to Ground

Background
Telescoping Coils
• S and R Conducting Simultaneously
• Flashover of S with R Closed

Background

Flashover of S with R Closing

5
Medium Voltage Reduced Voltage Autotransformer
Starter Failures
– Explaining The Unexplained –

• Background
• Investigation
• Testing 1988

Testing
Initial Testing

Testing
Initial Testing

6
 Testing
Initial Testing

Testing

Medium Voltage Reduced Voltage Autotransformer

Starter Failures
– Explaining The Unexplained –

• Background
• Investigation
• Testing 1988
• Findings

7
 Findings

Design Modifications
• Changed Design to Three Coil Three
Legged Autotransformer
• Changed Control Circuit to Include a
Transition on Current Below 125%
FLA
• HOWEVER, Failures Continued at
a Rate of Two to Three a Year !

Findings

The High Voltage Stress Failures

Occurred When:
• Transformers were on the 80% tap
• Transitions Were Forced at or
Near Locked Rotor Current

Findings

Remaining Failures Required a Closer Look

and Additional Testing

8
Medium Voltage Reduced Voltage Autotransformer
Starter Failures
– Explaining The Unexplained –

• Background
• Investigation
• Testing 1988
• Findings
• Testing 2002-03

Additional Testing

(February 2002 High Power Lab in Pennsylvania)

• However, The Sample Rate Was
only 50,000 Samples Per Seconds
and I suspect some level of
smoothing.

• Confirmed the Test Results From

1988 on Three Coil Three Legged
Transformers

Additional Testing

Dielectric Withstand
Two Layers of .005 Nomex 14kV rms for 60 Secs.

9
 Additional Testing
Partial Discharge

Additional Testing
November, December and January
Belmont, North Carolina

Additional Testing
A Closer Look

10
 Additional Testing
February
Arden, North Carolina

Medium Voltage Reduced Voltage Autotransformer

Starter Failures
– Explaining The Unexplained –

• Background
• Investigation
• Testing 1988
• Findings
• Testing 2002-03
• Findings

Findings

Issue Found !

11
 Findings
Current Flat During Voltage Escalation

Findings
When Reignition Occurs

Findings
Each Burst
70,000 Volts/Micro Sec

Each Burst
70,000 Volts/Micro Sec

12
Medium Voltage Reduced Voltage Autotransformer
Starter Failures
– Explaining The Unexplained –

• Background
• Investigation
• Testing 1988
• Findings
• Testing 2002-03
• Findings
• Solution

Solution

New Circuit for Autotransformer Starters

Solution
With Surge Arresters
12-13,000 Volts at 12-18 amps for 800 Micro Sec

13
Medium Voltage Reduced Voltage Autotransformer
Starter Failures
– Explaining The Unexplained –

• Background
• Investigation
• Testing 1988
• Findings
• Testing 2002-03
• Findings
• Solution
• Conclusions

Conclusions

• When autotransformer starters are forced

to transition before they reach near full
speed, high voltages are generated on the
0% taps
• When controllers are configured for
transition on current sensing below 125%
FLA, high voltage potential is reduced but
not eliminated.

Conclusions
• When 6 kV distribution surge arresters are
installed on a 4,160 volt circuit from the 0%
tap to ground, the voltage is clamped to 13
kV without the resultant high dv/dt across
the transformer coil regardless of transition
current
• Vacuum contactor autotransformer
starters and starter retrofits require
surge arresters even if no problems
were encountered with air break
contactors

14
 Conclusions
• The 15kV Autotransformer Starter in
British Columbia was set on the 55%
tap. The starter used MV Circuit
Breakers. Therefore all
autotransformer starters need surge
arrestors zero tap to ground or across
the VI,s.
• Care must be taken so the Metal Oxide
arrestors do not conduct near normal
operating voltages.

15
5.0 PT FAILURES

5-1
Transformer Failure Due to Circuit Breaker
Induced Switching Transients
Part 4: Potential Transformer Failures
David D. Shipp, PE Thomas J. Dionise, PE
Fellow, IEEE Senior Member, IEEE
Eaton Electrical Group Eaton Electrical Group
130 Commonwealth Dr. 130 Commonwealth Dr.
Warrendale, PA 15086 Warrendale, PA 15086

Case Studies

• Switching Transients – Opening

• Switching Transients – Closing
• Ferro- Resonance – Closing
• Ferro-Resonance – Opening (20 HZ
Saturation)
• Internal Resonance

2
2

CASE 12
Midwest Data Center

• 12.47 kV System / 120 MW Load

• Bkr Pairs with Unloaded wye-wye PTs for Auto
Transfer Sensing at Load End of Cables
• Multiple Open and Closed Operations were
Performed Preceding the Failure.
• 1st failure – Smoke But fuses did not Blow –
Cleared Manually.

3
3

1
CASE 12
Midwest Data Center

• 2nd Failure – Identical Switching Events

• Open Transitioned Back to Source “A”
• A few Minutes Later A Load “Pop” Was heard.
• More Smoke + B Phase Fuse Blew
• Measurements Were Taken – Snuck Up on
Problem without PT Loading – Risked Failure

4
4

CASE 12
Midwest Data Center

5
5

CASE 12
Closing From Generator Source – 300 kW Load

6
6

2
CASE 12
Opening From Generator Source – 300 kW Load

7
7

CASE 12
Primary Measurements - 300 kW Load Opening

8
8

CASE 12
Primary Measurements – Snubbers + 300 KW Load

9
9

3
CASE 12 - Closing
Primary Measurements - 300 kW Load

10
10

CASE 12 – Closing - Snubber

Primary Measurements - 300 kW Load

11
11

PT Failure – Example 13

• PT 1500VA, 14400/120V, open delta, in gen-tie swgr

• System 13.8 kV wye-solidly grounded
• Upstream VCB switching
• Cable lengths of 250 to 3600 feet
• Failures
• #1 – catastrophic failure of PT in gen-tie swgr (2700ft cable)
• #2 – a month later, catastrophic failure of PT in gen-tie swgr
• No switching at time of failure/100s prior to failure
• Possible failure modes
• Ring frequency of transient overvoltage on closing
• Ferroresonance (PT saturation) on opening

12
12

4
 CASE 13
Switching Transient 1st Failure

13
13

PT transient on VCB closing (EMTP simulated)

14
14

PT - test without snubber (measured)

Oscillation continues
beyond ¼ cycle

Transient followed by
high frequency ring

• Transient overshoot to -17.5kVpeak

• High frequency oscillation follows
• Oscillation persists for more than ¼ cycle on phase-b
15
15

5
 PT - test without snubber (zoom view)
Recorded Volts/Amps/Hz Zoomed Detail: 04/27/2011 19:21:07.855.770.000 - 04/27/2011 19:21:07.857.283.100

12000

10000

8000

6000

4000

2000

0
Volts

-2000

-4000

-6000
High frequency
-8000
effects – 50kHz
-10000

-12000 11.27kVpeak
-14000
nominal

-16000

Magnitude -18000
effects – 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700
Time 04/27/2011 19:21:07.854000 + us
2800 2900 3000 3100 3200

17.5kV
• Magnitude is not the only problem
• High frequency as well as high magnitude
• Overtime both will damage PT 16
16

PT - test with snubber

Oscillation well damped

Transient just above

normal crest

• Overshoot to 12.25kVpeak
• Lower frequency oscillation
• Oscillation very well damped 17
17

PT – test with snubber (zoom view)

Recorded Volts/Amps/Hz Zoomed Detail: 04/27/2011 20:09:12.582.800.000 - 04/27/2011 20:09:12.584.990.800
12000

10000

8000 Low frequency

2000Hz
6000

4000

2000
Volts

-2000

-4000

-6000

-8000

-10000

-12000
11.27kVpeak
Low Magnitude nominal
2800 3000 3200 3400 3600 3800 4000 4200 4400 4600 4800
12.25kV Time 04/27/2011 20:09:12.580000 + us

• Both magnitude and frequency are acceptable

• PT is well protected by the snubber
18
18

6
PT saturation on VCB opening

measured

simulated

19
19

PT saturation on VCB opening (EMTP)

20
20

PT with 100W loading resistors VCB opening

21
21

7
6.0 SNUBBER PERFORMANCE MEASUREMENTS

6-1
 Transformer Failure Due to Circuit Breaker
Induced Switching Transients
Part 5: Snubber Performance Measurements
David D. Shipp, PE Thomas J. Dionise, PE
Fellow, IEEE Senior Member, IEEE
Eaton Electrical Group Eaton Electrical Group
130 Commonwealth Dr. 130 Commonwealth Dr.
Warrendale, PA 15086 Warrendale, PA 15086

Snubber Performance Measurements

• Data Center
• Purpose of Measurements
• Test procedures
• Medium voltage measurements
• Selecting a meter
• Using voltage dividers
• Safety concerns
• Data Center challenges

2
2

Purpose of Performance Measurements

• Ensure the proper operation of the snubbers

• Measure the transient overvoltage waveforms at the transformer
primary
• produced during switching of the primary vacuum circuit breakers
• verify they do not exhibit high frequency transients (magnitude, rate-
of-rise and frequency).
• Specify the meters and voltage dividers needed for the transient
measurement
• guide electrical contractor with their installation
• electrical contractor to operate all equipment following the test
procedure
• After completion of the site monitoring during the snubber tests,
analyze the measurement data and document the test results in
an engineering report.

3
3

1
Data Center

Highlights
• Tier III
• LEEDS Certified
• 12.5MVA Capacity
• 13.2KV Ring Bus

4
4

Data Center – Measurement Locations

Vacuum
Breaker

Short
Cable

Cast-Coil
Transformer

5
5

Data Center – Electrical Highlights

• 2 x 24.9 kV lines from Factory Shoals and Buzzard

Roast
• 2 x 12.5 MVA transformers
• 13.2 kV “ring-bus” configuration
• MSA and MSB and generator bus GPS
• Vacuum circuit breakers – 600A mains & ties, 200A feeders
• 3 x 2250 KW generators
• 6 x 3750KVA cast coil transformers, 90kV BIL
• 3 x DE Subs CSA, CSB and CSC
• 3 x SE Subs MDSA, MDSB and MLB
• Feeder cable lengths vary from 109 to 249 Feet

6
6

2
 RC Snubber at MVS - Typical

Cast Coil
3750KVA
Transformer

200 A
Disconnect to
RC Snubber transformer
13.8 kV
0.25 microFarad
30 ohm

7
7

RC Snubber at MVS - Typical

Snubber Specs
13.8 kV
0.25 microFarad
30 ohm

8
8

Test Procedure

• Test Procedure required by contractor

• Prepared 2 weeks in advance of testing
• developed the instructions
• Included safety briefing each day
• Site specifics supplied by the contractor
• LO/TO instructions
• Breaker operations
• All meter connections made de-energized & LO/TO
• No one in transformer room during tests
• Signature of “Responsible Engineer” before test began
• detailed test procedure form
9
9

3
Measurement Equipment

• Measurement equipment included voltage

dividers and a transient recording device
• Voltage Dividers
• Capacitive and resistive elements
• 10MHz frequency response
• Three-Phase Power Quality Recorder
• transient voltage waveshape sampling
• 200 nsec sample resolution
• 5 Mhz sampling

10
10

Test Measurement Setup (Typical)

11
11

Voltage Divider Connections

• Highly stranded No. 8, 15 kV insulated hookup wire

• Insulated wire was a precaution – could have used
bare conductor
• maintained 8inch minimum separation for 15kV
between phases and ground
• Routed wires with gradual curve – no 90 degree
bends
• Connection at surge arrester at bus to transformer
primary windings

12
12

4
Voltage Divider Connections

13
13

Voltage Divider Connections

• Base of each divider grounded

• Grounded to ground bus in primary disconnect
• Used same highly stranded no. 8, 15kV insulated
hookup wire
• Connection to PQ Meter
• Cables to match input impedance of meter to that
of voltage divider
• Voltage divider was calibrated prior to shipping to
site

14
14

CT Connections

• Primary interest was voltage at transformer

• Secondary interest was snubber current
• Connected CTs at ground side of snubbers

15
15

5
Initial view of events

• RMS trend of voltage

• VCB energized and de-energized transformer

Energize De-energize
(see zoom) (see zoom)

16
16

Zoom View of Energize

• Slight overshoot above normal crest

• Ring well damped

17
17

Zoom View of De-Energize

• No transient on opening
• Well damped

18
18

6
Other Tests – transformer natural frequencies
• Sweep Frequency Response Analysis (SFRA)
Impedance
Phase Angle

Frequency 19
19

Data Center Challenges

• Contractor requested measurements

• Advised contractor
• First hookup/measurement – 1 day
• Remaining 5 transformers – 2nd day
• Prepare two voltage divider carts in advance
• Contractor agreed to schedule but…
• This was a construction site
• Project manager had other objectives
• Snubber testing viewed as an “inconvenience”
• Voltage divider carts were not prepared
• Be prepared to negotiate
20
20

Summary

• Prepare test procedure – 2 weeks in advance

• Request onelines and site photos to plan hookups
• Ship all equipment in advance
• For safety, two engineers perform measurements
• Follow test procedure
• Make all connections de-energized, LO/TO
• No one in transformer room during test
• Radio contact with contractor performing switching
• Download data after each test
• Analyze results before proceeding to next test
• Begin building test report while on-site

21
21

7
7.0 SUMMARY

7-1
Transformer Failure Due to Circuit Breaker
Induced Switching Transients
Summary
David D. Shipp, PE Thomas J. Dionise, PE
Fellow, IEEE Senior Member, IEEE
Eaton Electrical Group Eaton Electrical Group
130 Commonwealth Dr. 130 Commonwealth Dr.
Warrendale, PA 15086 Warrendale, PA 15086

The Problem – In a Nutshell

• Switching transients associated with circuit breakers

observed for many years
• Breaking opening/closing interacts with the circuit
elements producing a transient
• The severity of the transient is magnified by breaker
characteristics
• Current chopping on opening
• Pre-strike or re-ignition on closing
• In limited instances, the transient overvoltage exceeds
transformer BIL resulting in failure
• RC snubber in combination with surge arrester
mitigates the transient

2
2

What to look for…

“Rules of Thumb” to screen applications:

• Generally, short distance between circuit breaker and
transformer
• about 200 feet or less
• Dry-type or cast coil transformer
• oil filled not immune and low BIL
• Inductive load being switched
• transformer, motor, etc. (light load or no load)
• Circuit breaker switching characteristics:
• chop (vacuum or SF6) or restrike (vacuum)

3
3

1
Predicting Transients – EMTP Simulations

• For purposes of screening applications for damaging TOVs

• Source, breaker, cable and transformer modeled
• Breaker models for current chop and re-ignition
U RUTIL XUTIL UT C1 LCABLERCABLE C2 T1 LTRAN RTRAN N:1 T2

UN
VUTIL

C/2 C/2 CH CL
RLRG

SYSTEM SOURCE VCB CABLE

TRANSFORMER
AT 13.8 KV BKR 13.8KV
4
4

Designing the Snubber

TX

• 15kV typical snubber & arrester R

• transformer protection SA
C
• non-inductive ceramic resistor
• 25 ohms to 50 ohms
• surge capacitor
• capacitor ratings Surge
Arrester
0.15 μF to 0.35 μF
• 3-phase 13.8kV solidly ground
Resistor
• 1-phase 13.8kV LRG

Surge
Cap

5
5

Performance Measurement Equipment

• Test equipment includes voltage dividers and

a transient recording device
• Voltage Dividers
• Capacitive and resistive elements
• 10MHz frequency response
• Three-Phase Power Quality Recorder
• transient voltage waveshape sampling
• 200 nsec sample resolution
• 5 Mhz sampling

6
6

2
8.0 REFERENCES

Switching Transients

1. Shipp, Dionise, Lorch and MacFarlane, “Transformer Failure Due to Circuit

Breaker Induced Switching Transients”, IEEE Transactions on Industry
Applications, April/May 2011.
2. Shipp, Dionise, Lorch and MacFarlane, “Vacuum Circuit Breaker Transients
During Switching of an LMF Transformer”, IEEE Industry Applications Society
Annual Meeting 2010, October 2010.
3. ANSI/IEEE, A Guide to Describe the Occurrence and Mitigation of Switching
Transients Induced By Transformer And Switching Device Interaction, C57.142-
Draft.
4. D. Shipp, R. Hoerauf, "Characteristics and Applications of Various Arc
Interrupting Methods," IEEE Transactions Industry Applications, vol 27, pp 849-
861, Sep/Oct 1991.
5. ANSI/IEEE, Standard for AC High-Voltage Generator Circuit Breakers on a
Symmetrical Current Basis, C37.013-1997.
6. ANSI/IEEE, Application Guide for Transient Recovery Voltage for AC High-
Voltage Circuit Breakers, C37.011-2005.
7. D. Durocher, “Considerations in Unit Substation Design to Optimize Reliability
and Electrical Workplace Safety”, ESW2010-3, 2010 IEEE IAS Electrical Safety
Workshop, Memphis.
Ferroresonance

1. IEEE Standard Dictionary of Electrical and Electronics Terms, ANSI/IEEE Std

100-1984.
2. Hopkinson, R.H., “Ferroresonant Overvoltages Due to Open Conductors,”
General Electric, 1967, pp. 3 - 6.
3. Westinghouse Distribution Transformer Guide, Westinghouse Electric Corp.,
Distribution Transformer Division, Athens, GA, June 1979, revised April 1986,
Chapter 4 Ferroresonance, pp. 36 - 40.
4. IEEE Guide for Application of Transformers, ANSI/IEEE C57.105-1978, Chapter
7 Ferroresonance, pp. 22 – 28.
5. Distribution Technical Guide, Ontario Hydro, Ontario, Canada, May 1999, original
issue May 1978, pp. 72.1-1 – 72.1-10.
6. Greenwood, A., “ Electrical Transients in Power Systems”, Wiley & Sons, 1971,
pp. 91-93.
7. Kojovic, L., Bonner, A., “Ferroresonance - Culprit and Scapegoat”, Cooper Power
Systems, The Line, December 1998.

8-1
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 47, NO. 2, MARCH/APRIL 2011 707

Transformer Failure Due to Circuit-Breaker-Induced

Switching Transients
David D. Shipp, Fellow, IEEE, Thomas J. Dionise, Senior Member, IEEE,
Visuth Lorch, and Bill G. MacFarlane, Member, IEEE

Abstract—Switching transients associated with circuit breakers

have been observed for many years. Recently, this phenomenon
has been attributed to a significant number of transformer failures
involving primary circuit-breaker switching. These transformer
failures had common contributing factors such as the following:
1) primary vacuum or SF-6 breaker; 2) short cable or bus con-
nection to transformer; and 3) application involving dry-type or
cast-coil transformers and some liquid-filled ones. This paper will
review these recent transformer failures due to primary circuit-
breaker switching transients to show the severity of damage
caused by the voltage surge and discuss the common contributing
factors. Next, switching transient simulations in the electromag-
netic transients program will give case studies which illustrate
how breaker characteristics of current chopping and restrike
combine with critical circuit characteristics to cause transformer
failure. Design and installation considerations will be addressed,
particularly the challenges of retrofitting a snubber to an exist-
ing facility with limited space. Finally, several techniques and Fig. 1. Simplified electrical distribution system for data center.
equipment that have proven to successfully mitigate the breaker
switching transients will be presented, including surge arresters, which can fail distribution equipment such as transformers.
surge capacitors, snubbers, and these in combination. Transformer failures due to circuit-breaker-induced switching
Index Terms—Electromagnetic Transients Program (EMTP) transients are a major concern, which is receiving attention in a
simulations, RC snubbers, SF-6 breakers, surge arresters, switch- draft standard [1] and the focus of this paper.
ing transients, vacuum breakers.

I. I NTRODUCTION A. Forensic Evidence for a Unique Case

Consider the case study of a new data center with a 26-kV
T ODAY, medium-voltage metal-clad and metal-enclosed
switchgears that use vacuum circuit breakers are applied
over a broad range of circuits. These are one of many types
double-ended loop-through feed to six dry-type transformers
each rated at 3000 kVA AA/3390 kVA FA and 26/0.48 kV
of equipment in the total distribution system. Whenever a and delta–wye solidly grounded, as shown in Fig. 1. The trans-
switching device is opened or closed, certain interactions of former primary winding is of 150-kV basic impulse insulation
the power system elements with the switching device can cause level (BIL). A vacuum breaker was used to switch the trans-
high-frequency voltage transients in the system. The voltage- former. The 4/0 cable between the breaker and the transformer
transient severity is exacerbated when the circuit breaker op- was 33 kV and 133% ethylene propylene rubber. Primary
erates abnormally, i.e., current chopping upon opening and arresters were installed. The transformers were fully tested,
prestrike or reignition voltage escalation upon closing. Such including turns ratio, insulation resistance, etc. Functional tests
complex phenomena in combination with unique circuit char- were completed, including uninterruptible power supply (UPS)
acteristics can produce voltage transients involving energies full load, UPS transient, data center room validation, etc. In
the final phase of commissioning, a “pull-the-plug” test was
implemented with the following results.
1) De-Energization Failure #1. Four electricians “simultane-
Manuscript received June 4, 2010; accepted July 9, 2010. Date of publication ously” opened four 26-kV vacuum breakers to simulate
December 23, 2010; date of current version March 18, 2011. Paper 2010-
PPIC-188, presented at the 2010 IEEE Pulp and Paper Industry Conference, a general utility outage. All systems successfully trans-
San Antonio, TX, June 21–23, and approved for publication in the IEEE ferred to standby generation but a “loud pop” was heard
T RANSACTIONS ON I NDUSTRY A PPLICATIONS by the Pulp and Paper Indus- in Substation Room B and the relay for the vacuum circuit
try Committee of the IEEE Industry Applications Society.
The authors are with Eaton Electrical Group, Warrendale, PA 15086 USA breaker feeding transformer TB3 signaled a trip.
(e-mail: [email protected]; [email protected]; visuthlorch@ 2) Energization Failure #2. Minutes later, two electri-
eaton.com; [email protected]). cians “simultaneously” closed two 26-kV vacuum break-
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. ers to substation Room A. Transformer TA3 failed
Digital Object Identifier 10.1109/TIA.2010.2101996 catastrophically.
0093-9994/$26.00 © 2010 IEEE
708 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 47, NO. 2, MARCH/APRIL 2011

TABLE I
H ISTORY OF T RANSFORMER FAILURES R ELATED TO
P RIMARY VACUUM B REAKER S WITCHING

Fig. 2. Transformer failure #2 during energization.

due to primary switching of vacuum breakers. Table I details a

history of transformers related to primary switching of vacuum
breakers occurring within the past three years.
In Case 1, in a hydro dam, the transformer was “value
engineered” with a 13.8-kV primary-winding BIL of 50 kV. The
BIL should have been 95 kV for the 13.8-kV class.
The 1955 switchgear was replaced with modern vacuum
breakers with only 20 ft of cable to the transformer. The user
chose to energize the transformer before conducting a switching
transient analysis and failed the transformer primary winding.
Fig. 3. Transformer failure #1 during de-energization. The post mortem analysis revealed that no surge protection was
applied.
Failure #2 is shown in Fig. 2. Examination of the primary In Case 2, in a hospital, the vacuum breaker was
windings revealed that the coil-to-coil tap burnt off and the close-coupled through 27 ft of cable to a 2500-kVA dry-type
winding terminal showed an upward twist. The burn marks transformer with a 95-kV-BIL primary winding. The vacuum
from the initial Flash indicated the transient concentrated on breakers were supplied with no surge protection because the
the first turns of the windings. Typically, closing the vacuum particular vacuum breaker installed had a very low value of cur-
breaker to energize the transformer is the worst condition. rent chop. During vacuum breaker switching of the transformer,
Failure #1 is shown in Fig. 3. Examination of the primary the transformer failed. The transformer was rewound, and surge
windings revealed Flash and burn marks on the B-phase wind- protection/snubbers were installed.
ing at the bottom and middle. Those at the top indicate a coil- In Case 3, in a railroad substation, vacuum breakers
to-coil failure, not a winding-to-winding failure, and indicate a applied at 26.4 kV were used to switch a liquid-filled rectifier
transient voltage with high dv/dt. Those in the middle were a transformer with 150-kV-BIL primary winding. The switching
result of the cable (used to make the delta connection) swinging transient overvoltage (TOV) failed the middle of the primary
free. Supports were only lacking for this jumper (oversight winding. Forensic analysis determined a rectifier with dc link
during manufacturing) which could not withstand the forces of capacitors, and the transformer inductance formed an internal
the transient. This transformer passed the BIL test at 150-kV resonance that was excited by the switching. Such an LC series
BIL but ultimately failed at 162-kV BIL. resonance typically fails the middle of the transformer primary
All six transformers and cables were identical, but only winding.
two failed during the vacuum-circuit-breaker switching. The In Case 4, in a data center, vacuum breakers applied at
significant difference was that the two failed units had 40 ft 26.4 kV were used to switch six dry-type transformers with
of feeder cable while the others had 80 or 100 ft of feeder 150-kV-BIL primary windings under light load. Two transform-
cable. This short 40-ft cable, high-efficiency transformer, and ers failed, one on breaker closing and the other on opening.
vacuum circuit breaker proved to be the right combination to The failed transformers were connected by 40 ft of cable to the
produce a damaging voltage transient on both energization and vacuum breaker, while the other transformers had either 80 or
de-energization. 100 ft of cable. Arresters were in place at the time of failure,
but there were no snubbers.
In Case 5, in an oil field, a dry-type transformer for a variable
B. History of Failures and Forensic Review
speed drive (VSD) had multiple windings to achieve a 36-pulse
The previous example is not an isolated case. Instead, it effective “harmonic-free” VSD. A vacuum breaker at 33 kV
is representative of a growing number of transformer failures was separated from the transformer by only 7 ft of cable.
SHIPP et al.: TRANSFORMER FAILURE DUE TO CIRCUIT-BREAKER-INDUCED SWITCHING TRANSIENTS 709

TABLE II
C URRENT C HOP V ERSUS C ONTACT M ATERIAL

Arresters were applied on the primary winding. However, upon

closing the breaker, the transformer failed.
Finally, in Case 6, in an oil drilling ship, vacuum breakers
designed to International Electrotechnical Commission (IEC)
standards were applied at 11 kV and connected by 30 ft of Fig. 4. Voltage escalation due to successive reignitions.
cable to a dry-type cast-coil propulsion transformer rated at
7500 kVA. The transformer was also designed to IEC stan- 3–5 A, offering excellent interruption performance and a
dards, and the primary winding had a BIL of 75 kV. The IEC moderate weld strength. Table II shows the average and
transformer BIL is much lower than the American National maximum levels of current chop for copper–chromium,
Standards Institute (ANSI) BIL for the same voltage class copper–bismuth, and other contact materials. It should
winding. The transformer failed upon opening the breaker. be noted that both vacuum and SF-6 interrupters current
chop. Current chop is not unique to vacuum breakers.
C. Common Parameters 4) Reignition. Current chop, even though very small, coupled
with the system capacitance and transformer inductance
The severity of the voltage surge, i.e., high magnitude and
can impose a high-frequency transient recovery voltage
high frequency, and the damage caused by the voltage surge
(TRV) on the contacts. If this high-frequency TRV ex-
are determined by the circuit characteristics. The following are
ceeds the rated TRV of the breaker, reignition occurs.
some “rules-of-thumb” to screen applications for potentially
Repetitive reignitions can occur when the contacts part
damaging switching transient voltages.
just before a current zero and the breaker interrupts
1) generally, short distance between circuit breaker and at high-frequency zeros, as shown in Fig. 4. On each
transformer (about 200 ft or less); successive reignition, the voltage escalates. The voltage
2) dry-type transformer (oil filled and cast coil not immune) may build up and break down several times before inter-
and low BIL; rupting. Although current-chop escalation with modern
3) inductive load being switched (transformer, motor, etc.); VIs is rare, a variation of this concept applies on closing
4) circuit-breaker switching characteristics: chop (vacuum called prestrike.
or SF-6) or restrike (vacuum). 5) Switching inductive circuits. The transformer is a highly
inductive load with an iron core. The effect of switching
D. Underlying Concepts this inductive load and core must be considered. The
current cannot change instantaneously in an inductor.
3) Current Chop. When a vacuum breaker opens, an arc burns
Energy cannot be created or destroyed; only the form
in the metal vapor from the contacts, which requires a
of energy is changed. The energy in the inductor is
high temperature at the arc roots [2]. Heat is supplied by
described by
the current flow, and as the current approaches zero, the
metal vapor production decreases. When the metal vapor √
1/2LI 2 = 1/2CV 2 or V = I L/C. (1)
can no longer support the arc, the arc suddenly ceases or
“chops out.” This “chop out” of the arc called “current
chop” stores energy in the system. If the breaker opens From the energy equation, it can be seen that, for short
at a normal current zero at 180◦ , then there is no stored cables, C is√very small, which results in a very high surge
energy in the system. If the breaker opens, chopping impedance L/C. Energizing a cable produces a travel-
current at 170◦ , then energy is stored in the system. ing wave which reflects when it meets the discontinuity in
Current chop in vacuum circuit breakers is a ma- surge impedance between the cable and the transformer. The
terial problem. Older vacuum interrupters (VIs) used surge impedance of a cable may be under 50 Ω, while the surge
copper–bismuth. Modern VIs use copper–chromium. impedance of the transformer is 300–3000 Ω. In theory, the
Most copper–chromium VIs have a low current chop of reflection can be as high as 2 per unit.
710 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 47, NO. 2, MARCH/APRIL 2011

Fig. 5. Important circuit elements for EMTP modeling.

Vacuum circuit breakers are prone to current chopping and

voltage reignition while SF-6 circuit breakers are more prone
to just current chopping. Air circuit breakers are not prone
to either of any significant magnitude. Manufacturers design
vacuum-circuit-breaker contacts to minimize the severity and
occurrence of abnormal switching leading to severe voltage
surges (the lowest current-chop characteristics are 3–5 A).
Regardless of the circuit-breaker manufacturer, voltage surges
do occur.

E. Characteristics of the Voltage Transient and

Transformer Limits
The voltage transient that develops following the vacuum-
Fig. 6. Simplified electrical system for ship propulsion drive.
circuit-breaker switching is influenced by three factors: stored
energy, dc offset, and the oscillatory ring wave. The voltage
circuit elements modeled in electromagnetic transients program
component is due to the stored energy. The dc offset is de-
(EMTP) consisting of the source, breaker, cable, and trans-
termined by the X/R ratio of the cable and the transformer.
former, as shown in Fig. 5. The cable is represented by a
The oscillatory ring wave is a result of the capacitance and
Pi model consisting of the series impedance and half of the
inductance of the cable and the transformer. The magnitude of
cable charging at each end. In some cases, multiple Pi models
the voltage transient is compared with the transformer BIL. If
are used to represent the cable. The vacuum or SF-6 breaker
the magnitude is excessive, then the transformer winding will
is represented by a switch with different models for opening
likely fail line to ground. If the voltage transient has excessive
(current chop), restrike (excessive magnitude of TRV), reig-
rate of rise (dv/dt), then the transformer winding will likely
nition (excessive frequency of TRV), and closing (prestrike).
fail turn to turn (natural frequency of ring wave). For the
The three-phase transformer model consists of the leakage im-
transformer to survive the transient, the insulation must be able
pedance, magnetizing branch, and winding capacitances from
to withstand both the magnitude and the dv/dt. Dry-type trans-
high to ground and low to ground. For oil-filled transformers,
formers are particularly susceptible to vacuum or SF-6 breaker
the oil acts like a dielectric so the high-to-low capacitance
switching transients. However, oil- or liquid-filled transformers
is modeled. In cases requiring more detail, the transformer
are not immune. The oil has capacitance and acts like a surge
saturation and hysteresis effects are modeled.
capacitor to slow the rate of rise of the voltage transient.
The choice of the integration time step will depend upon
The trend in modern-day power systems is to install trans-
the anticipated frequency of the voltage transient. If it is too
formers with high-efficiency design. As a result, these high-
large, the time steps will “miss” the frequency effects. If it is
efficiency transformers have a very small resistance, which
too small, then this will lead to excessive simulation times. The
offers little or no damping to the voltage transient. Also, the
Nyquist criteria call for a minimum sample rate of twice the
repetitive effect, i.e., small indentations in the insulation, can
anticipated frequency. In switching transients, the anticipated
occur with each successive peak of the voltage transient.
frequency is 3–25 kHz. When the circuit breaker opens, the
transformer primary winding is ungrounded. Also, the ring
II. P REDICTING P ERFORMANCE W ITH S IMULATIONS wave is a function of the natural frequency of the circuit
A. Modeling the Circuit √
fnatural = 1/(2π LC). (2)
When a statistical approach is taken for switching transients,
complex modeling requires a frequency-dependent transformer The iron core of the transformer dominates the inductance
model and an arc model of the circuit breaker. For purposes of the circuit. The capacitance is very small for the dry-type
of screening applications for potentially damaging switching transformer and short cable. Consequently, the circuit’s natural
transients, a simpler approach is suggested with the important frequency is 3–25 kHz with relatively short cables.
SHIPP et al.: TRANSFORMER FAILURE DUE TO CIRCUIT-BREAKER-INDUCED SWITCHING TRANSIENTS 711

Fig. 8. TRV leading to reignition during energization of drive transformer

with and without snubber.
Fig. 7. Matching the simulation to field measurements.
TABLE III
B. Mitigating the Switching Transient C URRENT C HOP AND R EIGNITION C ASES FOR D RIVE P ROPULSION
T RANSFORMER S WITCHING
Various surge protection schemes exist to protect the
transformer primary winding from vacuum-breaker-switching-
induced transients. A surge arrester provides basic overvolt-
age protection (magnitude only). The arrester limits the peak
voltage of the transient voltage waveform. The surge arrester
does not limit the rate of rise of the TOV. A surge capacitor in
combination with the surge arrester slows down the rate of rise
of the TOV in addition to limiting the peak voltage but does
nothing for the reflection or dc offset. The number of arrester
operations is greatly reduced because of the slower rate of
rise. There is a possibility of virtual current chopping. Finally,
adding a resistor to the surge capacitor and surge arrester
provides damping, reduces the dc offset of the TOV waveform,
and minimizes the potential for virtual current chopping. The
resistor and surge capacitor are considered an RC snubber.
Selecting the values of resistance and capacitance are best
determined by a switching transient analysis study, simulating propulsion and identical drives for reverse propulsion, eight
the circuit effects with and without the snubber. 1185-kVA dry-type transformers, and eight 630-A vacuum
circuit breakers. The critical parameters are the vacuum circuit
breaker, 50 ft of cable, and dry-type transformer of 30-kV
C. Matching the Model to Measurements
BIL. Fig. 7 shows that the EMTP simulation results match the
The results obtained from simulation of switching transients transients captured in the field with a high-speed power-quality
in EMTP are only as good as the choice of model and data used. meter (closing). The simulation shows 4.96 kVpeak , which is
When available, field measurements taken during the switching less than 30-kV BIL; however, the oscillation frequency of
transients enable verification of the EMTP model. The EMTP 20.2 kHz exceeds an acceptable limit of dv/dt. Having verified
model can be adjusted as needed to match the actual field- the model, a series of current-chop cases and reignition cases
measured conditions. To illustrate this approach, consider the were run. Fig. 8 shows the TRV leading to reignition and the
ship propulsion electrical system in Fig. 6. TRV with a snubber installed. Reignition occurs because the
The system consists of 3 × 2865-kW generators, a 4160-V TRV peak, time to crest, and rate of rise of recovery voltage
three-phase bus, two 1865-kW drives/motors for forward exceed IEEE ANSI C37.06 limits.
712 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 47, NO. 2, MARCH/APRIL 2011

Fig. 9. Simplified electrical distribution system for Tier III data center.

The snubber reduces the TRV below the IEEE/ANSI limits

for general-purpose vacuum breakers [3] and for generator
breakers [4]. Table III summarizes the reignition cases and
the current-chop cases. In all cases, the snubber is effective in
reducing the transient voltage.

D. Borderline Case
It is important to note that not all applications involving
primary switching of transformers using vacuum breakers re-
quire snubbers. The large majority of applications do not re-
quire snubbers. Switching transient studies are conducted to
determine when snubbers are needed. In this paper, the cases
were selected to show different situations requiring snubbers.
For the system shown in Fig. 9, the results were border-
line; therefore, a snubber was still applied for reliability pur-
poses. The Fig. 9 system is a Tier III data center with two
24.9-kV incoming lines, two 12.5-MVA 25/13.2-kV transform-
ers, a 13.2-kV ring bus, two 2250-KW generators, and six
3750-kV cast-coil transformers.
Data centers fall into the highest risk categories because of
their high load density, close proximities of circuit components,
highly inductive transformers (high-efficiency designs), and
frequent switching. The critical parameters for the Fig. 9 system
are vacuum circuit breakers, 90-kV-BIL transformers, and cable Fig. 10. TOV initiated by current chop during de-energization of cast-coil
lengths ranging from 109 to 249 ft. For the cable of 109 ft, the transformer with and without snubber.
results of opening the vacuum breaker with current chopping of
both within acceptable limits. Finally, field measurements were
8 A are shown in Fig. 10. The TOV is as high as 123 kVpeak
taken after the snubber was installed to ensure that the snubbers
on phase A which exceeds the transformer BIL of 95 kV.
performed as designed. The field test setup for the snubber
The TOV exhibits a significant dc offset because there is very
performance measurements is discussed in Section IV. The
little resistance in the highly inductive circuit. The oscillation
field measurements showed that the snubber limited the TOV
frequency of 969 Hz is slightly less than the acceptable limit.
within acceptable limits.
A snubber is required to reduce the peak below 95-kV BIL.
The results of adding a snubber are shown in Fig. 10. Note
E. Case of Switching a Highly Inductive Circuit
the significant reduction in the dc offset. The resistor in the
snubber provides the reduction in dc offset as well as damping. Now, consider the vacuum breaker switching of a highly
The peak is reduced to 28.6 kV and an oscillation of 215 Hz, inductive circuit, such as the starting current of a large grinder
SHIPP et al.: TRANSFORMER FAILURE DUE TO CIRCUIT-BREAKER-INDUCED SWITCHING TRANSIENTS 713

Fig. 12. TOV during de-energization of LMF transformer with and without
snubber protection.

The examples show that the problem is not unique to one

industry, application, vendor’s breaker, or transformer design.
Fig. 11. Simplified electrical distribution system for LMF.
For the pulp and paper industry, there are many situations
where circuit-breaker-induced switching transients are likely
motor or an electric arc furnace. The vacuum-circuit-breaker to damage transformers. The following examples are some of
switching of an electric arc furnace and ladle melt furnace the more common scenarios encountered in the pulp and paper
transformers raises concern because of their high inductive cur- industry.
rents. High-frequency transients and overvoltages result when 1) Vacuum breaker retrofit for primary load break switch in
the vacuum breaker exhibits virtual current chop and multiple a unit substation. In the pulp and paper industry, there
reignitions. As an example, the arc furnace circuit of Fig. 11 are numerous unit substation installations with primary
consists of a 50-MVA power transformer, 2000-A SF-6 breaker, load break fused switch and no secondary main breaker.
56-MVA autoregulating transformer, 1200-A vacuum breaker, This arrangement results in arc Flash issues on the low-
and 50-MVA furnace transformer. The switching of the SF-6 voltage secondary. Limited space on the low-voltage
and vacuum breaker was studied. The vacuum breaker, because side prevents installation of a secondary main breaker
of the 28-ft bus to the furnace transformer, was the worst to mitigate the arc Flash issues. Retrofitting a vacuum
case. The results opening the vacuum breaker with and without circuit breaker in the primary of the unit substation, in
snubbers are show in Fig. 12. The TOV of 386 kVpeak exceeds place of the primary load break switch, and sensing on
the transformer BIL of 200 kV, and the oscillation of 1217 Hz the secondary is a solution that provides both primary
exceeds the acceptable limit. Application of the snubber results and secondary fault protection [5]. Unit substations may
in a TOV of 56.4 kVpeak that is below the transformer BIL, have oil-filled or dry-type transformers. The secondaries
and the oscillation of 200 Hz is below the acceptable limit. The may be solidly grounded or resistance grounded. With the
results for cases involving current chop and reignition are given vacuum breaker closely coupled to the transformer, surge
in Table IV. arresters and snubbers are most likely needed.
2) Vacuum breaker and rectifier (or isolation) transformer
installation. Rectifier transformers are installed to serve
F. Concerns for the Pulp and Paper Industry
dc drives such as those needed for feed water pumps to
The previous examples illustrate that circuit-breaker-induced the boilers. Also, isolation transformers are installed to
switching transients can fail transformers for specific com- serve a large VSD or groups of smaller drives. Primary
binations of circuit parameters and breaker characteristics. voltages may be 13.8 or 2.4 kV, and secondary voltages
714 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 47, NO. 2, MARCH/APRIL 2011

may be 600 or 480 V. In both situations, vacuum breakers

are installed in the primary and closely coupled to the
transformer through a short run of bus or cable. Often,
these transformers are inside and of dry-type design.
3) New unit substation with primary vacuum breaker.
Recently, a paper mill installed a new metal-enclosed
vacuum switchgear and a new 13.8/2.4-kV 7500-kVA
transformer for a bag house for the generator boilers
to meet Environmental Protection Agency requirements. Fig. 13. Typical snubber and arrester arrangement for transformer protection.
The vacuum breaker was connected to the transformer
through 5 ft of bus. While doing the coordination and arc
Flash studies, the switching transient issue was identified.
The equipment was installed and was awaiting startup
and commissioning when the studies raised the concern.
Before energizing the transformer, snubbers were quickly
sized, obtained, designed, and installed.
The screening criteria previously mentioned identify the
aforementioned examples for potential damaging switching
transient voltages due to vacuum breaker switching. The
vacuum breaker, short distance to transformer, and dry-type
transformer (or aged oil-filled transformer) are key variables
to consider. With such short distance between breaker and
transformer, most of these installations will require snubbers.
One might conclude that standard snubbers could be applied.
However, a switching transient study is still recommended to
determine the unique characteristics of the circuit and custom
design the snubber for the application. Given the limited space
in each of these examples, it is unlikely that off-the-shelf Fig. 14. Example of standard snubber package for transformer.
standard snubbers would fit. A substantial part of the design
effort includes determining how to best fit the snubbers into the 5) Will equipment modification void equipment warranties?
new or existing unit substation or transformer enclosure. 6) Are there standard equipment packages that can accom-
plish the switching transient protection?
7) Is the application an indoor or outdoor application?
III. D ESIGNING THE S NUBBER 8) Is there available space to mount the protective equipment
The preceding analysis has shown that, in some cases, in an external enclosure?
switching transients can produce overvoltages that can result 9) How can it be verified if the switching transient protection
in equipment insulation failure. If the results of the switching components are working effectively?
transient study indicate a risk of overvoltage greater than the 10) What alarms are necessary if the protective equipment
BIL of the equipment and/or if the dv/dt limits are exceeded, components fail?
a surge arrester and snubber should be applied. The switch-
ing transient study may also indicate that multiple locations
B. Custom Engineering Design
require surge arresters and snubbers to protect the generator,
transformer, or large motor. Additionally, the study specifies the Fig. 13 shows the typical snubber arrangement for trans-
necessary protective components and determines how close the former protection. A noninductive ceramic resistor and a
protection must be placed to provide effective protection. surge capacitor are the basic components of a snubber design.
Resistance values typically range from 25 to 50 Ω. Standard
A. Design Requirements capacitor ratings that range from 0.15 to 0.35 μF are the basis
of the design.
At this point, custom engineering design determines how to Fig. 14 shows a standard 15-kV surge protection package.
best provide the protection needed for the equipment. The fol- The arresters are mounted on the top of the enclosure. A three-
lowing questions must be answered to ensure that the snubber phase surge capacitor is mounted on the bottom. Insulators and
design meets all criteria and specifications. bus are located in the center. The cables can enter from top or
1) Is the switching transient protection cost effective? bottom. A ground bus is located on the center right. If space
2) What is the value of the equipment being protected? heaters are required for outdoor locations, they are located on
3) What is the cost of lost production if the equipment fails the lower left.
from switching transients? Fig. 15 shows one phase of a custom snubber circuit. The
4) Can the protection be installed within existing equipment custom design was required because there was not enough room
enclosures? in the transformer for the snubber components. The enclosure
SHIPP et al.: TRANSFORMER FAILURE DUE TO CIRCUIT-BREAKER-INDUCED SWITCHING TRANSIENTS 715

TABLE IV
C URRENT C HOP AND R EIGNITION C ASES FOR LMF T RANSFORMER S WITCHING

Fig. 15. 15-kV snubber mounted above the transformer.

had to be mounted above the transformer. The cable connec-

tions from the transformer were field installed and land on the Fig. 16. 15-kV snubber mounted in switchgear with glow tubes and current
copper bus. A 15-kV nonshielded jumper cable was used to sensors.
make the connection. Each phase passed through an insulation continuity. The purpose of the blue current sensors at the bottom
bushing to the transformer below. Bus work was required to of Fig. 16(a) is to monitor the continuity of the resistor and fuse
provide a solid support for the fragile resistors. Normally, only (optional) and alarm on loss of continuity. A close-up of the
one resistor would be provided, but for this application, to current sensor is shown in Fig. 17(b). Some industries mandate
achieve the delivery schedule, parallel resistors were designed fused protection. If there should be a broken resistor or a blown
to obtain the correct ohmic value (the correct single resistor fuse, an alarm signal can be sent to the plant distributed control
value had long delivery). system or supervisory control and data acquisition system to
Fig. 16(a) and (b) shows a snubber assembly mounted in alert the operating personnel that these snubber components
medium voltage switchgear. The photo on the left shows the have failed. Fig. 16(b) shows a continuation of the same snub-
single-phase surge capacitors mounted vertically. The black ber assembly. Three fuses are attached to the tops of the resistor.
cylinders are ceramic resistors. A variety of options are avail- The fuses will isolate any fault that may occur in the snubber
able to detect if the snubbers are functional. They range from assembly and prevent loss of the breaker circuit.
nothing (oversized but treated like a lightning arrester) to very
sophisticated loss of circuit detection. Glow tube indicators are
C. Special Design Considerations
shown at the top of Fig. 16(a), and a close-up is shown in
Fig. 17(a). These glow tubes are visible through a window in The nature of high-frequency switching transients requires
the switchgear door and provide a visual indication of snubber special design considerations. The snubber designer should
716 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 47, NO. 2, MARCH/APRIL 2011

Fig. 17. 15-kV snubber visual indication (glow tubes) and continuity verifi-
cation (current sensors).
Fig. 18. Snubber with (left) three-phase and (right) single-phase surge capac-
consider the location of the switching transient source when de- itors for 13.8-kV paper mill application.
veloping the custom design layout of the protective equipment.
Abrupt changes in the electrical path should be avoided. A low
inductive reactance ground path should be designed, using non-
inductive ceramic resistors and flat tin braided copper ground
conductors. The minimum clearances of live parts must meet or
exceed the phase-to-phase and phase-to-ground clearances of
NEC Table 490.24. The enclosure should be designed to meet
the requirements of IEEE Standard C37.20.2 1999. When the
enclosure is mounted greater than 10 ft from the equipment to
be protected, NEC tap rules may apply to the cable size required
and additional circuit protective devices may be required.

D. Custom Designs for the Pulp and Paper Industry

As mentioned previously, given the limited space in each
of the examples related to the pulp and paper industry, it is
unlikely that off-the-shelf standard snubbers would fit. Instead,
a substantial part of the design effort includes determining how
to best fit the snubbers into the new or existing unit substation Fig. 19. Snubber for metal-enclosed 13.8-kV vacuum breaker.
or transformer enclosure. Following are three examples of the
custom design effort needed for snubber installation.
1) 13.8-kV Solidly Grounded System in Paper Mill. A vac- for a bag house. Single-phase surge capacitors were used.
uum breaker was retrofitted into the enclosure for the Single resistors of the right ohmic value were available.
primary load break switch of the unit substation with Adequate clearance did not require the use of glastic
a dry-type transformer. The space for the snubber was as shown in Fig. 19.
extremely limited, as shown in Fig. 18 (left). Because the
system was solidly grounded, the voltage on the surge
IV. M EASUREMENTS TO V ERIFY
capacitor was limited to 8 kV line to ground; therefore,
S NUBBER P ERFORMANCE
it was possible to use a three-phase surge capacitor. The
tight clearance required the use of glastic to insulate the Following the installation of the snubbers, power quality
components at line potential from ground. measurements may be taken to ensure the proper operation of
2) 13.8-kV Resistance Grounded System in Paper Mill. An- the snubbers. A high-speed scope or power quality disturbance
other example of retrofitting a vacuum breaker for a analyzer should be used to measure the TOV waveforms at the
primary load break switch with a 30-year-old oil-filled transformer primary produced during switching of the primary
transformer. Because this snubber was needed immedi- vacuum circuit breaker. The measurements are used to verify
ately, the only available resistors had to be paralleled to that the waveforms do not exhibit excessive high-frequency
obtain the desired resistance, as shown in Fig. 18 (right). transients (magnitude, rate of rise, and frequency).
Again, glastic was used for insulation and to support the The test measurement setup generally consists of voltage
resistors. dividers and a transient recording device. The voltage dividers
3) 13.8-kV Low Resistance Grounded System. Snubbers should be made of capacitive and resistive components with
were provided for a new metal-enclosed vacuum a bandwidth of 10 MHz. The scope or power quality meter
switchgear and a new 13.8/2.4-kV 7500-kVA transformer should be capable of transient voltage wave shape sampling,
SHIPP et al.: TRANSFORMER FAILURE DUE TO CIRCUIT-BREAKER-INDUCED SWITCHING TRANSIENTS 717

snubber to an existing facility with limited space. Finally, the

performance of the snubbers is verified with field measurements
at the medium-voltage primary winding of the transformer.

R EFERENCES
[1] ANSI/IEEE, A Guide to Describe the Occurrence and Mitigation of Switch-
ing Transients Induced by Transformer and Switching Device Interaction.
C57.142-Draft.
[2] D. Shipp and R. Hoerauf, “Characteristics and applications of various arc
interrupting methods,” IEEE Trans. Ind. Appl., vol. 27, no. 5, pp. 849–861,
Sep./Oct. 1991.
[3] Standard for AC High-Voltage Generator Circuit Breakers on a Symmetri-
cal Current Basis, C37.013-1997, 1997.
[4] Application Guide for Transient Recovery Voltage for AC High-Voltage
Circuit Breakers, C37.011-2005, 2005.
[5] D. Durocher, “Considerations in unit substation design to optimize reliabil-
Fig. 20. Snubber performance measurement setup for 15-kV transformer ity and electrical workplace safety,” presented at the IEEE IAS Electrical
using high bandwidth voltage dividers. Safety Workshop, Memphis, TN, 2010, Paper ESW2010-3.

David D. Shipp (S’72–M’72–SM’92–F’02) re-

ceived the B.S.E.E. degree from Oregon State Uni-
versity, Corvallis, in 1972.
He is a Principal Engineer with the Electrical
Services and Systems Division, Eaton Corporation,
Warrendale, PA. He is a distinguished scholar in
power system analysis and has worked in a wide
variety of industries. He has spent many years per-
forming the engineering work associated with his
present-day responsibilities, which include a wide
Fig. 21. Voltage divider connections at 18-kV surge arrester on transformer range of services covering consulting, design, power
primary bushing. quality, arc flash, and power systems analysis topics. Over the last few years,
he has pioneered the design and application of arc-flash solutions, modifying
power systems to greatly reduce incident energy exposure. He has written over
8000-Vpeak full scale, and 200-ns sample resolution (5-MHz 80 technical papers on power systems analysis topics. More than 12 technical
sampling). papers have been published in IEEE Industry Applications Society (IAS)
An outage and lockout/tagout are needed to install the volt- national publications and two in EC&M. He spent ten years as a professional
instructor, teaching full time. He occasionally serves as a legal expert witness.
age dividers, make connections to each of the three phases at Mr. Shipp is currently the Chair for the IEEE Industrial and Commercial
transformer primary bushing, and make secondary connections Power Systems-sponsored Working Group on generator grounding. He has
to the transient recorder. Fig. 20 shows the test measurement received an IEEE IAS Prize Paper Award for one of his papers and conference
prize paper awards for six others. He is very active in IEEE at the national level
setup for a typical cast-coil transformer. Voltage divider con- and helps write the IEEE Color Book series standards.
nections to the transformer primary were made, as shown in
Fig. 21.
Additionally, for tests requiring load at a transformer, a
portable load bank may be connected to the transformer sec-
ondary. A resistive load bank configurable to different load
levels (300 or 100 kW) prevents destructive testing.
Thomas J. Dionise (S’79–M’82–SM’87) received
the B.S.E.E. degree from The Pennsylvania State
University, University Park, in 1978, and the
V. C ONCLUSION M.S.E.E. degree with the Power Option from
Carnegie Mellon University, Pittsburgh, PA, in 1984.
This paper has reviewed recent transformer failures due to He is currently a Senior Power Systems Engi-
primary circuit-breaker switching transients to show the sever- neer with the Power System Engineering Depart-
ity of damage caused by the voltage surge and discuss common ment, Eaton Corporation, Warrendale, PA. He has
over 27 years of power system experience involving
contributing factors. Next, switching transient simulations in analytical studies and power quality investigations
EMTP were presented to illustrate how breaker characteristics of industrial and commercial power systems. In the
of current chopping and restrike combine with critical circuit metal industry, he has specialized in power quality investigations, harmonic
analysis and harmonic filter design for electric arc furnaces, rectifiers, and
characteristics to cause transformer failure in unique situations. variable-frequency drive applications.
In these limited instances, mitigation of the transients is accom- Mr. Dionise is the Chair of the Metal Industry Committee and a member
plished with snubbers custom designed to match the specific of the Generator Grounding Working Group. He has served in local IEEE
positions and had an active role in the committee that planned the Industry
circuit characteristics. Design and installation considerations Applications Society 2002 Annual Meeting in Pittsburgh, PA. He is a Licensed
were addressed, particularly the challenges of retrofitting a Professional Engineer in Pennsylvania.
718 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 47, NO. 2, MARCH/APRIL 2011

Visuth Lorch received the B.S.E.E. degree from Bill G. MacFarlane (S’70–M’72) received the
Chulalongkorn University, Bangkok, Thailand, in B.S.E.E. degree from The Pennsylvania State Uni-
1973, and the M.S. degree in electric power engi- versity, University Park, in 1972.
neering from Oregon State University, Corvallis, in He began his career in 1972 with Dravo Corpo-
1976, where he was also a Ph.D. candidate and was ration, Pittsburgh, PA, as a Power System Design
inducted as a member of the Phi Kappa Phi Honor Engineer supporting this engineering/construction
Society. During his Ph.D. studies, he developed the company with expertise in the design and construc-
Short Circuit, Load Flow, and Two-Machine Tran- tion of material handling systems, chemical plants,
sient Stability programs. The Load Flow program pulp and paper mills, steel facilities, ore benefaction,
has been used in the undergraduate power system mine design, and computer automation systems. He
analysis class. He also prepared a Ph.D. thesis on the provided full electrical design services in heavy in-
Load Flow program using the third-order Taylor’s series iterative method. dustrial applications, primarily in the steelmaking and coal-preparation-related
In 1981, he joined Westinghouse Electric Corporation, Pittsburgh, PA. He sectors. His responsibilities included client and vendor liaison and supervision
was responsible for conduction power system studies, including short circuit, of engineers, designers, and drafters. From 1978 to 2003, he was a principal
protective device coordination, load flow, motor starting, harmonic analysis, Process Controls Systems Engineer with Bayer Corporation, Pittsburgh. He
switching transient, and transient stability studies. He also developed the Pro- provided design and specification of instrumentation and control systems for
tective Device Evaluation program on the main frame Control Data Corporation chemical processes. He configured programmable logic controller and distrib-
supercomputer. In 1984, he joined the Bangkok Oil Refinery, Thailand, where uted control systems. He was a Corporate Technical Consultant for power
his primary responsibility was to design the plant electrical distribution system distribution systems. He had been the principal Electrical Engineer on many
as well as the protection scheme for the steam turbine cogeneration facility. He major projects with Bayer Corporation. He joined the Electrical Services and
was also responsible for designing the plant automation, including the digital Systems Division, Eaton Corporation, Warrendale, PA, in 2004. At Eaton,
control system for the plant control room. In 1986, he rejoined Westinghouse he has been involved in a 230-kV substation design, ground grid design for
Electric Corporation. He performed power system studies and developed the substations, as well as in short circuit, protective device coordination, load flow,
Short Circuit and Protective Device Evaluation programs for the personal and arc-flash hazard analysis studies. He has also done power factor correction
computer. In 1998, he joined the Electrical Services and Systems Division, and harmonics studies to resolve power quality issues.
Eaton Corporation, Warrendale, PA, where he is currently a Senior Power
Systems Engineer in the Power Systems Engineering Department. He performs
a variety of power system studies, including switching transient studies using
the electromagnetic transients program for vacuum breaker/snubber circuit
applications. He continues to develop Excel spreadsheets for quick calculation
for short circuit, harmonic analysis, soft starting of motors, capacitor switching
transients, dc fault calculation, etc.
 Vacuum Circuit Breaker Transients
During Switching of an LMF Transformer

David D. Shipp, PE Thomas J. Dionise, PE Visuth Lorch William MacFarlane, PE

Eaton Electrical Group Eaton Electrical Group Eaton Electrical Group Eaton Electrical Group
Power Systems Power Systems Power Systems Power Systems
Engineering Engineering Engineering Engineering
Warrendale, PA Warrendale, PA Warrendale, PA Warrendale, PA
Fellow, IEEE Senior Member, IEEE Senior Member, IEEE

Abstract— Switching transients associated with circuit breakers Frequent switching operations have been enabled by the
have been observed for many years. With the wide-spread development of the vacuum switch. The vacuum switch has
application of vacuum breakers for transformer switching, been designed for hundreds of operations in a day, for long life
recently this phenomenon has been attributed to a significant and low maintenance. With the advantages of the vacuum
number of transformer failures. Vacuum circuit breaker switch, also come the disadvantages of switching transient
switching of electric arc furnace and ladle melt furnace overvoltages. Depending on the characteristics of the vacuum
transformers raises concern because of their inductive currents. switch and the power system parameters, these switching
High frequency transients and overvoltages result when the transient overvoltages can be of significant magnitude and
vacuum breaker exhibits virtual current chop and multiple re-
frequency to cause transformer failure. High frequency
ignitions. This paper will present a detailed case study of
vacuum breaker switching of a new ladle melt furnace
transients and overvoltages result when the vacuum breaker
transformer involving current chopping and re-strike simulations exhibits virtual current chop and multiple re-ignitions.
using the electromagnetic transients program. A technique that According to statistics compiled by one insurance company [1],
involves a combination of surge arresters and snubbers will be the application of vacuum switches has resulted in numerous
applied to the ladle melt furnace to show the switching transients failures of arc furnace transformers. These failures rates have
can be successfully mitigated. Additionally, some practical been reduced by the application of surge arresters, surge
aspects of the physical design and installation of the snubber will capacitors and damping resistors [2]. The transients produced
be discussed. by the vacuum circuit breaker switching of an LMF
transformer and their mitigation are the focus of this paper.
Keywords- Switching Transients, vacuum breakers, SF-6
breakers, LMF transformer, EMTP simulations, surge arresters, A. The LMF Circuit of Interest
RC snubbers.
Consider the new LMF circuit of Fig. 1 that consists of a 50
MVA, 135/26.4 kV power transformer, a 2000 A SF-6
I. INTRODUCTION breaker, a 56 MVA, 27/10 kV auto-regulating transformer, a
Electric Arc Furnaces (EAF) are used widely in the steel 1200 A vacuum breaker and a 50 MVA, 25/0.53 kV furnace
industry in the production of carbon steel and specialty steels. transformer. The SF-6 circuit breaker is separated by 53 feet
The Ladle Melt Furnace (LMF) maintains the temperature of from the auto-regulating transformer. The vacuum circuit
liquid steel after tapping the EAF and facilitates changes in the breaker is separated by 28 feet from the LMF transformer. The
alloy composition through additives. In both cases, the furnace normal configuration (1) consists of the new LMF and the
transformer is a critical component of the furnace circuit that is existing 4EAF operating in parallel. One alternate
exposed to severe duty. The demands of the melt cycle may configuration (2) of the LMF circuit consists of 4OCB and
result in extensive damage to the furnace transformer due to 4EAF out-of-service with 4LTC in standby-service. A second
electrical failures in the transformer. With advances in alternate configuration (3) of the LMF circuit consists of LMF
technology and metallurgy, the operation of arc furnaces today LTC out-of-service with 4LTC switched on-line to source the
is significantly different. Heats of 4 to 5 hours with periods of LMF. Each of these three possible configurations of the LMF
moderate loading have been reduced to 3 to 4 hours with circuit was considered. Of the three, the normal configuration
consistently high loading. Accompanying the shorter heats of results in the shortest bus length between the vacuum breaker
sustained loading are many more switching operations. and the LMF transformer. The normal configuration also
Combined, these factors impose thermal and electrical stresses results in the shortest bus length between the SF-6 breaker and
on the transformer. the LMF transformer.
 UTILITY
4713MVA 3PH SC
II. SWITCHING TRANSIENTS SIMULATIONS
9.26 X/R
Switching transients simulations were conducted in the
138KV electromagnetic transients program (EMTP) to investigate the
1600A possible failure of the new LMF transformer due to transient
50/66/83MVA overvoltages during the circuit switching of the new vacuum
135.3/26.4KV and SF-6 circuit breakers. The LMF circuit model developed
7.5%Z
13OHM
NC
27KV
NC
in EMTP consisted of the source, breaker, cable and
4OCB SF-6 BREAKER
transformer. The cable was represented by a Pi model
2 2000A 2000A consisting of the series impedance and half of the cable
ALUMINUM charging at each end. In some cases, multiple Pi models are
IPS BUS
53FEET
used to represent the cable. The vacuum or SF-6 breaker was
NC NO represented by a switch with different models for opening
NO NC
XFER BUS
(current chop), re-strike (excessive magnitude of TRV), re-
4AUTO LTC 3 1
AUTO LTC ignition (excessive frequency of TRV) and closing (pre-strike).
56MVA 56MVA
27-10KV 2 27-10KV The three phase transformer model consisted of the leakage
NC NO
3.3%Z 3.3%Z impedance, magnetizing branch, winding capacitances from
XFER BUS NC high-to-ground and low-to-ground. For oil filled transformers,
1700A 3
VACUUM BREAKER
1200A
the oil acts like a dielectric so the high-to-low capacitance was
HEAVY DUTY modeled. Two worst-case switching scenarios involving the
COPPER PIPE
28FEET
2000 A SF-6 breaker and the 1200 A vacuum breaker were
EAF XFMR LMF XFMR
simulated: 1) current chop by the breaker on de-energization of
2 50/56MVA 50/56MVA the LMF transformer and 2) re-ignition following opening of
25/.53KV 25/.53KV
2.5%Z 2.5%Z the breaker during energization of the LMF transformer. Also,
4EAF LMF the surge arrester was not modeled to show worst case.
38MW 20MW
(EXISTS) (NEW)

Figure 1. Simplified Electrical Distribution System for New LMF

A. Modelling Current Chop for Vacuum and SF-6 Breakers
When a vacuum breaker opens, an arc burns in the metal
B. Critical Characteristics of the Furnace Ciruit vapor from the contacts which requires a high temperature at
The severity of the switching transient voltage; i.e. high the arc roots [3]. Heat is supplied by the current flow and as
magnitude and high frequency, and the damage caused by the the current approaches zero, the metal vapor production
transient overvoltage (TOV) are determined by critical decreases. When the metal vapor can no longer support the
characteristics of the LMF power supply circuit: arc, the arc suddenly ceases or “chops” out”. This “chop out”
of the arc called “current chop” stores energy in the system. If
• Short distance between circuit breaker and transformer the breaker opens at a normal current zero at 180 degrees, then
• BIL of the transformer there is no stored energy in the system. If the breaker opens
chopping current at 170 degrees, then energy is stored in the
• Inductive load being switched (transformer) system. For modern breakers, current chop can range from 3 –
• Circuit breaker switching characteristics: chop (vacuum or 21 A depending on the contact material and design. Both
SF-6) or re-strike or re-ignition (vacuum) vacuum and SF-6 interrupters current chop. Current chop is
In the case of the furnace circuit, the vacuum or SF-6 not unique to vacuum breakers. Fig. 2 illustrates the LMF
breaker induced switching transients can be amplified by the transformer load current at time of the vacuum circuit breaker
short bus or cable length between the breaker and transformer. opening is 10 Amperes. The Phase-B pole opens first at 2.7891
This amplification is due to the vacuum or SF-6 breaker msec, followed by Phase-A at 6.1875 msec and Phase-C at
chopped current and the system stray capacitance, especially 6.4377 msec. The vacuum circuit breaker was modeled with 6
that of the short bus or cable. In modeling the system for such A chopped current. The SF-6 breaker was modeled similarly.
switching transient analysis, it is important to accurately
represent the vacuum or SF-6 breaker chopped current, stray
capacitance of the short bus or cable and inductance of the
transformer being switched.
The study approach was to evaluate the normal
configuration (shortest bus lengths) shown in Fig. 1 which
produces the worst case TOV during vacuum and SF-6 breaker
switching and size the RC snubber for this worst case. The
performance of the RC snubber was proven for this worst case
of the normal configuration. The RC snubber designed for the
worst case will therefore reduce the less severe TOVs produced
during vacuum and SF-6 breaker switching for the two
alternate configurations. For breaker opening cases, the
transient recovery voltage (TRV) of the breaker was evaluated. Figure 2. Current Chop for the Vacuum Breaker
B. De-Energize the LMF Transformer at Light Load by C. De-Energize the Auto-Regulating Transformer at Both No
Opening the 1200A Vacuum Breaker Load and Light Load by Opening the 2000A SF-6 Breaker
In Case 1, the 1200 A vacuum breaker feeding the 56 In Case 3, the 2000 A SF-6 breaker feeding the 56 MVA
MVA LMF transformer was opened to interrupt light load. auto-regulating transformer was opened to interrupt no load.
The vacuum breaker was modeled as previously described At the time, the 1200 A vacuum breaker feeding the LMF
with 6 A chopped current. This value provides a small safety transformer was open. The SF-6 breaker was modeled with 6
margin for the vacuum breaker with an actual value of current A chopped current similar to that of the vacuum breaker. As
chopping of 3 to 5 A for a vacuum breaker of this design. The with the vacuum breaker, this provides a safety margin for the
results opening the vacuum breaker with and without snubbers manufacturer’s stated actual value of current chopping of 3 to 5
are shown in Fig. 3. The TOV of 386 kVpeak exceeds the A. In Case 4, the switching conditions are the same as for Case
transformer BIL of 200 kV and the oscillation of 1217 Hz 3, except the vacuum breaker is closed and the RC snubber
exceeds the acceptable limit. This TOV is unacceptable and circuit is applied at the primary side of the 56 MVA auto-
indicates the need for an RC snubber in addition to a surge regulating transformer. Fig. 4 compares the results for Cases 3
arrester. and 4 and shows the TOV magnitude at the primary side of the
LMF transformer was negligible in both cases. A snubber is
An RC snubber was designed to protect the LMF not required.
transformer as explained in Sec. III. Sec. III provides the The results of the switching transient study of the LMF
specifications for the RC snubber. In Case 2, the RC snubber operation during current chop conditions are summarized in
was modeled with the same switching conditions of Case 1. In Table I. For each case the magnitude and frequency of the
Case 2, application of the snubber results in a TOV of 56.4 TOV are given. Acceptable and unacceptable levels of
kVpeak which is well below the transformer BIL and the transient overvoltage are noted.
oscillation of 200 Hz is below the acceptable limit as shown in
Fig. 3. This TOV is acceptable and shows the RC snubber
effectively controls the TOV. In Fig. 3, notice the resistor in Case 3 - Voltage Waveform Without Snubber
the snubber reduces the DC offset of the transient voltage (no load)
waveform. The resistor also provides damping of the transient
voltage waveform. The vacuum breaker, because of the short
distance of 28 feet of bus to the furnace transformer, produced
the worst case TOV.

Case 1 - Voltage Waveform Without Snubber

Case 4 - Voltage Waveform With Snubber

(light load)

Case 2 - Voltage Waveform With Snubber

Figure 4. TOV During De-Energization of LMF Tranformer With and

Without Snubber Protection

TABLE I. SUMMARY OF CURRENT CHOP CASES FOR LMF

TRANSFORMER SWITCHING
Current
1
Vacuum SF6 Chop TOV Freq Transf
Case Breaker Breaker (A) RC (kV) (Hz) BIL Note
1 Open IC 6 N 386.5 1,217 200 U
2 Open IC 6 Y 56.4 200 200 A
Figure 3. TOV During De-Energization of LMF Tranformer by the Vacuum 3 IO Open 6 N N/A N/A 200 A
4 IC Open 6 Y 54.7 197 200 A
Breaker With and Without Snubber Protection Notes: U = unnacceptable. A = acceptable. 1 = current chop on breaker opening. IO = initially open. IC = initially closed.
D. Modelling Re-ignition
Case 5 – Voltage Without Snubber
Current chop, even though very small, coupled with the (one phase shown)
system capacitance and transformer inductance can impose a
high frequency transient recovery voltage (TRV) on the
contacts. If this high frequency TRV exceeds the rated TRV of
the breaker, re-ignition occurs. Repetitive re-ignitions can
occur when the contacts part just before a current zero and the
breaker interrupts at high frequency zeros [4]. On each
successive re-ignition, the voltage escalates. The voltage may
build up and break down several times before interrupting.
Figure 6. Voltage Escalation Due to Sucessive Re-ignitions
E. Interrupt Inrush Current to LMF Transformer Followed
by Re-Ignition of the Vacuum Breaker TABLE II. STD C37.06 EVALUATION OF TRV FOR VACUUM BREAKER
In Case 5, after the LMF transformer was energized for FOR CASES 5 & 6

about 3 cycles, the 1200 A vacuum breaker tripped open as

CASE 5 - CASE 5 CASE 6 with IEEE ANSI
shown in Fig. 5. On this initial opening, the TRV has an E2 of First Opening Reiginition Snubber C37.06 Limit
62.95 kVpeak, T2 of 19 μsec and RRRV of 3.3133 kV/μsec. T2 E2 (kV) 62.952 217.670 49.901 71
and RRRV exceed the limits of [5] and [6] of 63 μsec and T2 (μsec) 19 16 130 63
1.1270 kV/μsec as shown in Table II. As a result, re-ignition RRRV (kV/μsec) 3.3133 13.6044 0.3839 1.1270
occurs and the breaker opens at the next current zero. This
second opening of the vacuum breaker produces a TRV with F. Interrupt Inrush Current to Auto-Regulating Transformer
E2 of 217.67 kVpeak, T2 of 16 μsec and RRRV of 16.6044 by the SF-6 Breaker
kV/μsec, all exceed the limits of [5] and [6] of 71 kV, 63 μsec SF-6 breakers do not experience re-ignition. For
and 1.1270 kV/μsec respectively as shown in Table II. The illustration purposes only, the conditions leading to re-ignition
voltage escalation due to the successive re-ignitions is shown in of the vacuum breaker are duplicated for the SF-6 breaker. In
Fig. 6. The initial TRV and subsequent TRVs due to re- Case 7, after the auto-regulating transformer was energized for
ignition are unacceptable and indicate the need for an RC about 3 cycles, the 2000 A SF-6 breaker tripped open. At this
snubber. The conditions of Case 6 are the same as Case 5, time, the 1200 A vacuum breaker to the LMF transformer was
except an RC snubber is applied at the primary side of the 56 open. The conditions of Case 8 were the same as Case 7,
MVA LMF transformer. In Case 6, the vacuum breaker except with the application of the RC snubber circuit at the
interrupting the inductive current of the LMF furnace load side of the 2000 A SF-6 circuit breaker. Fig. 7 compares
transformer produces a TRV with E2 of 49.90 kVpeak, T2 of the results for Cases 7 and 8. For both Cases 7 and 8, the TRV
130 μsec and RRRV of 0.3839 kV/μsec that are well below the across the SF-6 breaker was within the limits of [5] and [6] as
limits of [5] and [6], and re-ignition is avoided. shown in Table III. The TRV was not sufficient to cause re-
ignition. The snubber is not required for SF-6 switching.
Case 5 - TRV Without Snubber
(one phase shown) Case 7 - TRV Without Snubber

Case 6 - TRV With Snubber

Case 8 - TRV With Snubber

Figure 5. TRV Across Vacuum Breaker During Interuption of LMF

Transfomer Inrush Current With and Without Snubber Protection Figure 7. TRV Across SF-6 Breaker During Interuption of Auto-Regulating
Transfomer Inrush Current With and Without Snubber Protection
 However, the snubber provides an additional benefit during TABLE IV. SUMMARY OF RE-IGNITION SWITCHING CONDITIONS AND
SWITCHING TRANSIENT RESULTS
closing of the SF-6 breaker to energize the auto-regulating
transformer. In Case 8, with the application of the snubber Current
1
IEEE
Chop ANSI
circuit, the period of the transient voltage following closing of Vacuum SF6 Transient Recovery C37.06
the SF-6 breaker to energize the auto-regulating transformer Case Breaker Breaker (A) RC Voltage
E2 (kV) 217.7
Limit
71
Note

was reduced from 1,100 μsec to 230 μsec as shown below as T2 (μsec) 16 63
RRRV
shown in Fig. 8. This advantage alone does not justify the 5
Open,
Reignition IC 6 N (kV/μsec) 13.6 1.127 U
installation of an RC snubber at the primary of the auto- E2 32.7 71
Open, No T2 99.4 63
regulating transformer. A surge arrester at the primary of the 6 Reignition IC 6 Y RRRV 0.3293 1.127 A
auto-regulating transformer is adequate. Open, No
E2
T2
29.4
75
136
106
7 IO Reignition 6 N RRRV 0.3923 1.127 A
This results of the switching transient study of the LMF E2 29.4 136
operation during re-ignition conditions are summarized in 8 IO
Open, No
Reignition 6 Y
T2
RRRV
75
0.3923
106
1.127 A
Table IV. The re-ignition was simulated for either the 1200 A Notes: U = unnacceptable. A = acceptable. 1 = current chop on breaker opening. IO = Initially open. IC = initially closed.

vacuum breaker or 2000 A SF-5 breaker. The condition of the

other circuit components for each case is described. For each III. MITIGATING THE SWITCHING TRANSIENT
case the magnitude of the transient recovery voltage (E2), the Various surge protection schemes exist to protect the
time to crest of the recovery voltage (T2) and the rate-of-rise of transformer primary winding from vacuum breaker switching
the recovery voltage (RRRV) are given. Acceptable and induced transients. A surge arrester provides basic overvoltage
unacceptable levels of transient overvoltage are noted. protection (magnitude only). The arrester limits the peak
voltage of the transient voltage waveform. The surge arrester
TABLE III. STD C37.06 EVALUATION OF TRV FOR SF-6 BREAKER FOR
does not limit the rate-of-rise of the transient overvoltage. A
CASES 7 & 8 surge capacitor in combination with the surge arrester slows
down the rate-of-rise of the transient overvoltage in addition to
CASE 7
Reiginition
CASE 8 with
Snubber
IEEE ANSI
C37.06 Limit
limiting the peak voltage but does nothing for the reflection or
E2 (kV) 29.4 29.4 136 DC offset. The number of arrester operations is greatly
T 2 (μsec) 75 75 106 reduced because of the slower rate-of-rise. There is a
.3923 .3923 1.1270
RRRV (kV/μsec)
possibility of virtual current chopping. Finally, adding a
resistor to the surge capacitor and surge arrester provides
Case 7 – Voltage waveform Without Snubber damping, reduces the DC offset of the transient overvoltage
waveform and minimizes the potential for virtual current
chopping. The resistor and surge capacitor are considered an
RC snubber. Selecting the values of resistance and capacitance
are best determined by a switching transient analysis study,
simulating the circuit effects with and without the snubber [7].

A. RC Snubber Ratings
The specifications for the RC snubber circuit are given in
Fig. 9. The resistor average power rating at 40 °C is 1000 W
and the peak energy rating is 17,500 joules. This RC snubber
circuit, which has the same ratings as the one presently
installed at the 3EAF, provides adequate mitigation of the
voltage transients produced by either the vacuum breaker or
Case 8 – Voltage waveform With Snubber SF-6 breaker and simplifies the inventory of spare parts, i.e.
only one type of resistor and capacitor must be stocked.

LMF TX
50/56
MVA
R
100OHM SA
36kV Rating
C 29kV MCOV
0.15uF

Figure 8. Transient Period of During Energizaton of Auto-Regulating

Figure 9. Snubber Specifications and Surge Arrester Arrangement for the
Transfomer Inrush Current With and Without Snubber Protection
LMF Transformer Protection
B. Locating the Snubber to Maximize Effectiveness
To maximize effectiveness, apply the RC snubber circuit to
the primary-side of the 56 MVA LMF furnace transformer.
Cases 1 and 2 have shown the RC snubber circuit reduces the
magnitude of the transient overvoltage during switching of the
vacuum breaker to be within the BIL of the LMF furnace
transformer and reduces the oscillating frequency of the
transient overvoltage to be less than 1,000 Hz. Cases 5 and 6
have also shown the RC snubber reduces the TRV of the
breaker, when interrupting the inductive current of the LMF
transformer, to be well below limits of [5] and [6].
Applying the RC snubber at the load-side of the SF-6
breaker is optional. In Cases 3 and 4 the transient overvoltage
was negligible. In Cases 7 and 8 the transient recovery voltage
is not sufficient to cause re-ignition. The only reason for
applying the RC snubber circuit at the load-side of the SF-6
breaker is to reduce the transient period during energization of
the LMF transformer. This advantage alone does not justify
the installation of an RC snubber at the primary of the auto-
regulating transformer. A surge arrester at the primary of the
auto-regulating transformer is adequate.

C. Custom Designing the Snubber

Given the limited space in the transformer vault, it was
unlikely off-the-shelf standard snubbers would fit. Instead, a
substantial part of the design effort determined how best to fit
Figure 11. Plan View of LMF Transformer Vault Showing RC Snubber
the snubbers into the new transformer vault. Fig. 10 shows one
phase of the custom RC snubber circuit . The custom design The RC snubber is mounted 16 feet above the floor. The
was required because of the 36 kV rating and the need to locate surge capacitor is mounted horizontally on an insulated stand-
the snubber in close proximity to the LMF transformer off bolted to the transformer vault wall. An insulator string
terminals in a highly congested transformer vault. was required to provide a solid support for the surge capacitor
to counter the torque-arm of the 39.5 inch, 150 lb. component .
The 39.5 inch, 20 lb. resistor was also mounted horizontally
and the base was bolted to the transformer wall but did not
require any additional support.
The nature of high frequency switching transients require
special design considerations. The snubber designer should
consider the location of the switching transient source when
developing the custom design layout of the protective
equipment. Abrupt changes in the electrical path should be
avoided. A low inductive reactance ground path should be
designed, using non-inductive ceramic resistors and flat tin
braided copper ground conductors. The minimum clearances of
live parts must meet or exceed the phase to phase and phase to
ground clearances of NEC Table 490.24.
The cable connections from the snubber capacitor land on
the 25 kV, 1.5 inch copper bus supplying the LMF transformer
as shown in Fig. 11. 35 kV non-shielded jumper cable was
used to make the connection between snubber and bus. The
elevation of the bus is 18 feet and that of the surge capacitor is
16 feet with the difference allowing for a gradual curve of the
jumper cable. Because of the high frequency transient
overvoltages, gradual arcs are preferred over 90 degrees bends.
Such an arrangement allows the snubbers to remain in place
during the removal of the transformer, should maintenance or
service be required. Also, the height of the snubbers provides
adequate floor space for maintenance personnel.
Figure 10. 36 kV RC Snubber for LMF Transformer Protection
 IV. CONCLUSIONS
This paper presented the findings of a detailed case study of
the vacuum breaker and SF-6 switching of an LMF furnace
transformer. Through EMTP switching transient simulations,
it was shown high frequency transient overvoltages result when
the vacuum breaker exhibits virtual current chop and multiple
re-ignitions. The simulations showed the SF-6 breaker
switching transients were negligible primarily due to the longer
distance of bus between the breaker and transformer. It was
shown a technique that involves a combination of surge
arresters and RC snubbers applied to the LMF transformer
primary effectively mitigates the switching transients due to the
vacuum breaker switching. Additionally, the practical aspects
of the physical design and installation were discussed including
the nature of high frequency switching transients which require
the avoidance of abrupt changes in the electrical path, and the
use of a low inductive reactance ground path.

REFERENCES

[1] Factory Mutual Insurance Company, “Arc Furnance Transformer

Installations”, Property Loss Prevention Data Sheets, pgs. 1-13, January
2002.
[2] ANSI/IEEE, A Guide to Describe the Occurrence and Mitigation of
Switching Transients Induced By Transformer And Switching Device
Interaction, C57.142-Draft.
[3] D. Shipp, R. Hoerauf, "Characteristics and Applications of Various Arc
Interrupting Methods", IEEE Transactions Industry Applications, vol 27,
pp 849-861, Sep/Oct 1991.
[4] A. Morre, T. Blalock, “Extensive Field Measurements Support New
Approach to Protection of Arc Furnace Transformers Against Switching
Transients”, IEEE Transactions on Power Apparatus and Systems, vol.
PAS-94, no. 2, March/April 1975, pgs. 473-481.
[5] ANSI/IEEE, Standard for AC High-Voltage Circuit Breakers Rated on a
Symmetrical Current Basis, C37.06-1997.
[6] ANSI/IEEE, Application Guide for Transient Recovery Voltage for AC
High-Voltage Circuit Breakers, C37.011-2005.
[7] D. Shipp, T. Dionise, V. Lorch, B. MacFarlane, “Transformer Failure
Due to Circuit Breaker Induced Switching Transients”, 2010 IEEE Pulp
and Paper Industry Conference Record, San Antonio.