Fault Current Management Guidebook - Updated
1012419
Fault Current Management Guidebook - Updated
1012419
Technical Update, November 2006
EPRI Project Manager
R. Adapa
ELECTRIC POWER RESEARCH INSTITUTE
3420 Hillview Avenue, Palo Alto, California 94304-1338 ▪ PO Box 10412, Palo Alto, California 94303-0813 ▪ USA
800.313.3774 ▪ 650.855.2121 ▪ askepri@epri.com ▪ www.epri.com
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF
WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI).
NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY
PERSON ACTING ON BEHALF OF ANY OF THEM:
(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH
RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM
DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED
RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS DOCUMENT IS
SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR
(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING
ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED
OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS
DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN
THIS DOCUMENT.
ORGANIZATION THAT PREPARED THIS DOCUMENT
EPRI Solutions, Inc.
This is an EPRI Technical Update report. A Technical Update report is intended as an informal report of
continuing research, a meeting, or a topical study. It is not a final EPRI technical report.
ORDERING INFORMATION
Requests for copies of this report should be directed to EPRI Orders and Conferences, 1355 Willow Way, Suite 278,
Concord, CA 94520. Toll-free number: 800.313.3774, press 2, or internally x5379; voice: 925.609.9169; fax:
925.609.1310.
Electric Power Research Institute and EPRI are registered service marks of the Electric Power Research Institute, Inc.
Copyright © 2005 Electric Power Research Institute, Inc. All rights reserved.
CITATIONS
This document was prepared by
EPRI Solutions, Inc.
942 Corridor Park Blvd.
Knoxville, TN 37932
Principal Investigators
H. Sharma
N. Abi-Samra
M. McGranaghan
This document describes research sponsored by the Electric Power Research Institute (EPRI).
This publication is a corporate document that should be cited in the literature in the following
manner:
Fault Current Management Guidebook - Updated. EPRI, Palo Alto, CA: 2006. 1012419.
iii
PRODUCT DESCRIPTION
This document is an update of Fault Current Management Guidebook (EPRI report 1010680) on
fault current effects and management in transmission and distribution systems. This guide is a
snapshot of available references, information, and literature on the effects of high fault current on
a number of power system components as well as available and emerging fault current limiters.
Results & Findings
Due to increased load demands and reduced incentives to build new transmission, energy
companies are increasing power flows on existing transmission assets, which will increase fault
current levels throughout the power system. Also, new generation sources to be added at the
transmission and distribution network will increase power flows and, consequently, fault current
levels. Under increased power flow conditions on existing assets, managing fault currents is
crucial for avoiding damage to equipment as well as increasing system reliability. Despite the
importance of proper fault current management, no comprehensive guide on this subject matter is
available to the industry. This guidebook fills the gap by documenting state-of-the-art techniques
for managing fault currents in transmission and distribution assets. Descriptions include
conventional methods such as neutral grounding resistors, current-limiting reactors, increased
transformer impedances, and splitting of bus-bars. Emerging fault-current-limiting technologies
such as superconducting and power electronic devices will be included as well.
Challenges & Objective(s)
Application of cost-effective fault-current-limiting technologies requires close cooperation
between equipment engineers and manufactures. In addition, cost/benefit analysis may not
always be possible, given that a number of these devices are not in wide use today and are not
typically available from many manufactures.
Applications, Values & Use
Understanding the mechanical forces that are developed in major equipment, such as
transformers and other substation equipment, will help utilities minimize equipment failures due
to mechanical forces at high fault current levels.
Users will better understand the effect of high fault current on existing protection systems and
metering systems, current interrupting devices, ground grids, and transmission lines. A highlevel economic analysis also is included so that energy companies can make informed decisions
when choosing options for limiting fault currents.
EPRI Perspective
The Fault Current Management Guidebook lists methods to limit fault currents in increased
power flow scenarios and, thus, avoid equipment failures and save millions of dollars. By
implementing one or more of the options in the guidebook, companies can increase power flows
without damaging equipment due to high fault currents. In addition, system outage costs could be
reduced.
As increased power flows are experienced in utility systems, management of fault currents has
become an important issue. Changing load flows, new generation locations, and other changes
v
have presented transmission asset owners with new challenges. One potential solution, uprating
circuit breakers, often is not cost justifiable or physically possible. This guidebook helps utilities
manage this emerging issue by providing information on how fault currents are increasing and
methods to limit them.
Approach
This Technical Update is part of Project Set 38D, Management of Fault Currents. Although the
effect fault current has on the equipment discussed in this Update will not change, the project
team may, in future versions, describe new and improved fault-limiting devices, make new case
studies available, and discuss new price points. Also, the effect of fault currents on other
equipment not covered in this edition may be added.
Keywords
Fault current
Mechanical forces
Short circuit
Failure mechanisms
Transformers
Core
Shell
Fuse
Breaker
Current limiting
Interrupting capability
SF6
Conductor
Tower
Insulator
Step potential
Touch potential
Protection systems
Metering systems
Current interrupting devices
Ground grids
Transmission lines
Economic analysis
vi
ABSTRACT
Due to increased load demands and reduced incentives to build new transmission, energy
companies are increasing power flows on existing transmission assets, which will increase fault
current levels throughout the power system. Also, new generation sources to be added at the
transmission and distribution network will increase power flows and, consequently, fault current
levels. Under increased power flow conditions on existing assets, managing fault currents is
crucial for avoiding damage to equipment as well as for increasing system reliability. Despite the
importance of proper fault current management, no comprehensive guide on this subject matter
has been available to the industry. This guidebook fills the gap by studying impacts of increased
fault current levels on key power system components, including substation bus structures,
protective devices, groundings, transmission lines, and power transformers. State-of-the-art
techniques for managing fault currents in transmission and distribution assets are documented
and include neutral grounding resistors, current limiting reactors, increased transformer
impedances, and splitting of bus-bars. Emerging fault-current-limiting technologies such as
superconducting and power electronic devices have been included as well.
This guidebook is not intended to be a complete document, but rather a snapshot in time based
on the available references, information, and literature. It is a technology update—a work-inprogress—that will be updated periodically. Studies will need to be conducted to investigate
effects of high fault currents wherever the current state of information is insufficient. Such
studies may include field measurement at various energy company sites. To the extent possible,
functional and economic comparisons of each available fault-managing method has been made
and reported in this guide.
vii
CONTENTS
SYMBOLS ................................................................................................................................ S-1
1 INTRODUCTION ....................................................................................................................1-1
Background ..........................................................................................................................1-1
Reasons for Increased Fault Currents .................................................................................1-1
Guidebook Structure ............................................................................................................1-1
2 MECHANICAL FORCES AND THERMAL EFFECTS IN SUBSTATION EQUIPMENT DUE
TO HIGH FAULT CURRENTS ..................................................................................................2-1
Introduction ..........................................................................................................................2-1
Effects of High Fault Current on Substation Conductors .....................................................2-1
Rigid Bus Bars- IEEE Standard .....................................................................................2-1
Rigid Bus Bars – IEC Standard......................................................................................2-5
Flexible Conductor Buses – Static Method ....................................................................2-7
Force Safety Devices ...................................................................................................2-14
Substation Cable and Conductor Systems ........................................................................2-15
Cable Thermal Limits ...................................................................................................2-15
Cable Mechanical Limits ..............................................................................................2-15
Distribution Line Conductor Motion ..............................................................................2-16
Effects of High Fault Currents on Substation Insulators, Supports and Structures............2-17
Rigid Bus Bars .............................................................................................................2-17
Flexible Conductor Buses – Static Method ..................................................................2-18
Flexible Conductor Buses – Dynamic Method .............................................................2-19
Effects of High Fault Currents on Gas Insulated Substations (GIS) ..................................2-20
Experimental Results .........................................................................................................2-22
Summary and Recommendations......................................................................................2-24
Appendix 2.1 Maximum possible values for dynamic factors (IEC Standard 865-1) ........2-24
Appendix 2.2 Factors for Different Bus-bar Arrangement (IEC Standard 865-1)..............2-26
Appendix 2.3 Factor q for Rigid Conductor (IEC Standard 865-1).....................................2-27
Appendix 2.4 Mechanical Effects on a 110kV arrangement with Slack Conductors..........2-28
Appendix 2.5 Mechanical Effects on Strained Conductors ................................................2-29
Appendix 2.6 Mechanical Effects on a 10kV arrangement with Single Rigid Conductors .2-30
Appendix 2.7 Mechanical Effects on a 10kV arrangement with Multiple Rigid Conductors2-31
Appendix 2.8 Mechanical Effects on a High Voltage Arrangement with Rigid Conductors 2-32
References.........................................................................................................................2-33
3 EFFECTS OF HIGH FAULT CURRENTS ON CURRENT INTERRUPTING DEVICES ........3-1
Air Circuit Breakers ..............................................................................................................3-1
Vacuum Circuit Breakers .....................................................................................................3-2
SF6 Circuit Breakers ............................................................................................................3-3
ix
Loss of Interruption Medium.................................................................................................3-4
Interrupting Ratings of Switching Devices............................................................................3-6
Circuit Breakers..............................................................................................................3-6
Fuses .............................................................................................................................3-8
Next-generation Solid-state Breakers ..................................................................................3-9
Generation 1: 15KV Class Distribution Switchgear Development................................3-10
Generation 2: 35/138KV Class Distribution/Transmission Switchgear Development ..3-10
Identifying the Breakers with Excessive Fault Currents .....................................................3-11
Case Studies......................................................................................................................3-13
Diablo Canyon..............................................................................................................3-13
Dresden and Quad Cities.............................................................................................3-14
Summary and Recommendations......................................................................................3-15
References.........................................................................................................................3-15
4 EFFECT OF HIGH FAULT CURRENTS ON PROTECTION AND METERING.....................4-1
Current Transformer Saturation ...........................................................................................4-1
Saturation of Low Ratio CTs ..........................................................................................4-3
Effect of High Fault Currents on Coordination .....................................................................4-7
Protective Relay Ratings and Settings.................................................................................4-9
Effects of Fault Currents on Protective Relays ..................................................................4-10
Examples .....................................................................................................................4-11
Methods for Upgrading Protection Systems.......................................................................4-11
Update Short Circuit Study...........................................................................................4-11
Update Protective Device Coordination Study .............................................................4-11
Modeling Techniques for Protection Studies......................................................................4-12
Summary and Recommendations......................................................................................4-14
APPENDIX 4.1 CT Saturation Evaluation Spreadsheet....................................................4-15
APPENDIX 4.2 EMTP Model for CT Saturation Evaluation ..............................................4-16
References.........................................................................................................................4-20
5 EFFECT OF HIGH FAULT CURRENTS ON GROUNDING GRIDS ......................................5-1
Introduction ..........................................................................................................................5-1
Summary of Ground Grid Design Procedures .....................................................................5-3
Site Survey.....................................................................................................................5-3
Conductor Sizing............................................................................................................5-3
Step and Touch Voltages...............................................................................................5-5
Ground Grid Layout........................................................................................................5-5
Ground Resistance Calculation......................................................................................5-5
Calculation of Maximum Grid Current ............................................................................5-6
Calculation of Ground Potential Rise (GPR) ..................................................................5-6
Mesh Voltage .................................................................................................................5-6
Step Voltage...................................................................................................................5-7
x
Detailed Design..............................................................................................................5-7
Example Design from IEEE Standard 80 .......................................................................5-7
Effects of High Fault Currents in Ground Grids....................................................................5-8
Failure Mechanisms .......................................................................................................5-8
Reduction in Electrical Safety: Increased Step and Touch Potentials............................5-8
Damage or Failure of Grounding Equipment .................................................................5-9
Case Studies......................................................................................................................5-12
Survey of Substation Grounding System Assessment and Refurbishment Practices..5-12
Safety Assessments of Transit Supply Substations and 161/69-kV Substations.........5-13
Grounding Systems for Electric Traction......................................................................5-14
Ground Current Measurement During a Fault..............................................................5-14
Design of Ground Grid for a Transition Station System ...............................................5-15
Large Industrial Plants .................................................................................................5-15
Deep-Ground-Well Method ..........................................................................................5-15
Design Using Two-Layer Soil Model ............................................................................5-16
Gas Insulated Substation Grounding Grid ...................................................................5-17
Summary and Recommendations......................................................................................5-17
References.........................................................................................................................5-17
6 EFFECT OF HIGH FAULT CURRENTS ON TRANSMISSION LINES..................................6-1
Effect of High Fault Current on Non-Ceramic Insulators......................................................6-1
Conductor Motion Due to Fault Currents .............................................................................6-3
Calculation of Fault Current Motion for Horizontally Spaced Conductors ......................6-4
Calculation of Fault Current Motion for Vertically Spaced Conductors ..........................6-6
Calculation of Mechanical Loading on Phase-to-Phase Spacers...................................6-8
Effect of Bundle Pinch on Conductors and Spacers ............................................................6-8
References.........................................................................................................................6-10
7 SHORT CIRCUIT FORCES IN TRANSFORMERS ................................................................7-1
Typical Values of Mechanical Forces in Transformers ........................................................7-1
Short Circuit Currents in a Transformer ...............................................................................7-2
Effect of the System Impedance on Short Circuit Currents............................................7-2
Short Circuit Tests of Power Transformers ....................................................................7-3
Through fault Currents in a Transformer..............................................................................7-3
Liquid-Immersed Transformers ......................................................................................7-3
Dry-Type Transformers ..................................................................................................7-5
Protection Considerations ..............................................................................................7-5
Mechanical Forces in Transformers.....................................................................................7-6
Critical Forces in Shell Form Transformers....................................................................7-8
Shell-Form Failure Mechanisms ....................................................................................7-8
Critical Forces in Core Form Transformers ..................................................................7-10
Failure Modes in Core Type Transformers...................................................................7-11
xi
Impact of Fault Currents on Transformer Life ....................................................................7-14
Summary and Recommendations......................................................................................7-16
References.........................................................................................................................7-16
8 FAULT CURRENT LIMITING METHODS ..............................................................................8-1
Conventional Methods .........................................................................................................8-1
Emerging Technologies .......................................................................................................8-7
FCL Definitions...............................................................................................................8-7
Requirements for Fault Current Limiters ..............................................................................8-8
Solid-State Current Limiter (SSCL) ......................................................................................8-8
"All Solid-State" Based Designs ...................................................................................8-10
Hybrid Designs.............................................................................................................8-12
EPRI SCR Based FCL .......................................................................................................8-13
Design Description .......................................................................................................8-13
Functional Description of the SSCL Operation ............................................................8-15
Progress Report ...........................................................................................................8-17
Superconducting Fault Current Limiters.............................................................................8-18
Superconducting Current Limiter (SCCL) Operation....................................................8-19
HTS Fault Current Limiter Developments ..........................................................................8-21
Shielded Core ..............................................................................................................8-21
Resistive Type..............................................................................................................8-22
Fault Current Controller (FCC) .....................................................................................8-22
CURL10 .......................................................................................................................8-24
ABB’s Resistive SCFCL ...............................................................................................8-25
SIEMENS Resistive SCFCL.........................................................................................8-26
CESI Project in Italy .....................................................................................................8-26
Super-ACE Project in Japan ........................................................................................8-27
Matrix Fault Current Limiter................................................................................................8-27
Concept........................................................................................................................8-27
Electrical Configuration of an MFCL ............................................................................8-28
Proof-of-Concept MFCL Design ...................................................................................8-29
Matrix ...........................................................................................................................8-30
Development of MFCL Elements .................................................................................8-33
Proof-of-Concept Test Results .....................................................................................8-35
Project Update – DOE Annual peer review, 2006 ........................................................8-40
Other Technologies............................................................................................................8-41
Controlled LC Resonance Circuits ...............................................................................8-41
Liquid Metal FCL ..........................................................................................................8-42
Series Compensator ....................................................................................................8-42
Reactor and Compensating capacitor based Approach...............................................8-43
Comparison of FCL technologies.......................................................................................8-44
Losses..........................................................................................................................8-44
xii
Size ..............................................................................................................................8-45
Recovery ......................................................................................................................8-45
FCL characteristic in the network .................................................................................8-45
Case Studies......................................................................................................................8-46
Air-core CLR in Brazil...................................................................................................8-46
IEE Project in China .....................................................................................................8-46
SCE Distribution Circuit of future .................................................................................8-47
Korean Electric Power Grid..........................................................................................8-47
Summary and Recommendations......................................................................................8-48
APPENDIX 8.1 EMTP Model for FCL Evaluation .............................................................8-49
Parameter Description of FCL Module.........................................................................8-49
FCL Module disabled ...................................................................................................8-50
Current Limiting reactor in Service...............................................................................8-51
Superconducting FCL in Service..................................................................................8-52
References.........................................................................................................................8-53
9 TECHNICAL AND ECONOMIC ANALYSIS OF FAULT CURRENT MANAGEMENT
SOLUTIONS ..............................................................................................................................9-1
Benefits of Fault Current Limiting.........................................................................................9-1
Economic Analysis for Conventional Solutions ....................................................................9-2
Cost Data .......................................................................................................................9-5
FCL Applications ..................................................................................................................9-7
Generator Connection....................................................................................................9-8
Coupling of Networks .....................................................................................................9-9
Coupling of Busbars.......................................................................................................9-9
Transformer feeder ......................................................................................................9-10
Coupling of Local Generation.......................................................................................9-10
Economic Analysis of Individual Components ...................................................................9-11
Power transformers ......................................................................................................9-11
Circuit Breakers............................................................................................................9-12
Economic Analysis for FCLs ..............................................................................................9-12
Solid State FCLs ..........................................................................................................9-12
Superconducting FCLs.................................................................................................9-15
System Integration Issues..................................................................................................9-18
Protection Coordination................................................................................................9-19
Testing of FCLs............................................................................................................9-21
Summary and Recommendations......................................................................................9-22
References.........................................................................................................................9-22
GLOSSARY.............................................................................................................................. G-1
xiii
SYMBOLS
ce
Energy charge
cP
Reference
($/kVA)
____________________________________
cost
of
SFCL
Symbol
Quantity
CC
Cost of capacitor in solidstate FCL application ($)
a
Center line distance between
main conductor mid-points
(m)
CD
Expansion factor
CF
Form factor
am
Effective distance between
main conductors (m)
CF
Yearly cost due to faults ($)
as
Clearance between mid-point
of adjacent sub-conductors
(m)
CFCL
Total cost of operating solidstate FCL ($)
*
C FCL
Normalized life cycle cost of
SFCL
CI
Investment Cost ($)
A
Conductor cross-section area
(sq. mm)
A
Total area of ground grid (sq.
m)
CI
Cost of inductor in solid-state
FCL application ($)
Akcmil
Conductor cross-section area
(kcmil)
CM
Maintenance costs ($/year)
As
Sub-conductor cross-section
area (sq. m)
COM
Operation and maintenance
costs ($/year)
bc
Equivalent conductor sag at
mid-span (m)
Cp
Purchase price of SFCL ($)
CP0
Cost of no-load losses ($)
Maximum
horizontal
displacement of flexible
conductor (m)
CPk
Cost of load losses ($)
CS
Surface layer-derating factor
B
Magnetic Field (Tesla)
CT
BC
Cost Benefit of using SFCL
Cost of control circuit in
solid-state FCL application
($)
B*C
Normalized Cost Benefit of
using SFCL
d
Diameter
of
conductor (m)
BS
Strategic Benefit of using
SFCL
d0
Initial conductor spacing of a
transmission line.
Normalized
Strategiceconomic benefit of using
SFCL
dS
Diameter of a flexible
conductor (m)
dt
Conductor spacing of a
transmission line at time t.
bh
*
BSE
cd
Demand charge
S-1
the
grid
D
Conductor center-to-center
spacing of a rigid bus (cm)
FG
D
Spacing between parallel
conductors in ground grid
(m)
Total bus unit weight,
including ice loading and
connectors (N/m)
Fm
Decrement factor that takes
into account the system
impedance
Force between rigid main
conductors during a short
circuit (N)
Fm2
Maximum distance between
any two points on the grid
(m)
Force between rigid main
conductors during a line-line
short circuit (N)
Fm3
Force on the central rigid
main conductor during a
balanced 3-phase short circuit
(N)
Df
Dm
E
Modulus of elasticity (kPa or
N/m2)
EGPR
Ground potential rise voltage
(v)
Fpi
Pinch force (N)
FSC
Fault current force (N/m)
Em
Mesh voltage (v)
Fst
ESTEP
Step Voltage (v)
Static tensile force in flexible
main conductor (N)
ETOUCH
Touch Voltage (v)
Ft
Short circuit tensile force (N)
f
System Frequency (Hz)
FT
Total vector force on the bus
(N/m)
fb
Natural frequency of the bus
(Hz)
Fv
fc
Capacitor loss factor
kW/Mvar
Short circuit current force
between the sub-conductors
in a bundle (N)
g
gravitational constant (m/s2)
G
International
Annealed
Copper Standard (IACS)
conductivity (%)
in
fe
Cost of energy loss per kWh
fu
LC utilization factor
fv
Present value factor
h
Depth of the ground grid (m)
fM
Maintenance factor
ha
FA
Maximum yield stress (kPa2)
Ratio of active power of subnetwork at the time of the
peak power of the whole
network to the peak active
power of the sub-network
ha
Height of conductor
transformer (inches)
Favg Average force on one conductor of a
transmission line
FDEV
Actuation force for the Force
Safety Device
Ff
Drop force (N)
S-2
in
hr
Ratio of maximum power to
rated power
Kre
Cost of removal of
existing circuit breaker
hs
Depth of the surface material
(m)
KS
H
Magnetic
(A/m)
Constant based on number of
spans and end types, Table
2-3
KS
Spacing
voltage
KVC
Capacitor variable cost per
Mvar in solid-state FCL
KVCB
Circuit breaker variable cost
per MVASC rating
KVL
Inductor variable cost per
MVAr in solid-state FCL
HC
field
strength
Critical magnetic field for
superconductivity
factor
for
the
step
i
Inflation rate
IC
Critical
temperature
superconductivity
IG
Maximum grid current (kA)
ISC
Symmetrical
current (A)
fault
KVT
Thyristor variable cost per
MVAr in solid-state FCL
j
Parameter determining the
bundle configuration
KVTS
Incremental cost of installing
a new transmission system
J
Moment of inertia of the
cross-section (cm4)
l
k
Number of FCL
Maximum
center
line
distance between supports
(m)
K
Pinning factor
lc
Kf
Mounting structure flexibility
factor
Cord length of the main
conductor of a flexible bus
(m)
li
Kf
Material-fusing constant
Length of one insulator chain
(m)
KFC
Fixed capacitor cost in solidstate FCL
lS
Distance
between
adjacent spacers (m)
KFCB
Fixed circuit breaker cost
∆L
Change in bus bar length (m)
KFL
Fixed inductor cost in solidstate FCL
LC
Length of beam (inches)
KFT
Fixed thyristor cost in solidstate FCL
LC
Total length of conductor in
the horizontal grid (m)
Inner conductor factor
Li
Initial bus bar length (m)
Kii
Kin
Installation cost of the new
circuit breaker
LM
Effective buried length of
ground grid (m)
Km
Geometrical factor
LP
Length of the perimeter of
the grid (m)
rms
for
S-3
two
LR
Individual ground rod length
(m)
PWS
Present worth of using an
SFCL
LS
Maximum allowable length
of the bus (cm)
PCE
LS
Effective length of buried
conductor
Present cost of energy losses
(inductor, capacitor and
thyristor) in solid-state FCL
application
Lt
Inductance
of
transmission system
r
Ratio of electromechanical
force on a conductor during
short circuit to gravity
LT
Total length of buried
conductor in ground grid (m)
Rg
Resistance of the grounding
grid (ohms)
Lx
Maximum length of the grid
in the x direction (m)
Rp0.2
Stress corresponding to yield
2
point (N/m )
Ly
Maximum length of the grid
in the y direction (m)
S
Section modulus (m3)
m
Mass per unit length of main
conductor (kg/m)
S
Spring constant of support
(N/m)
m
load factor
S
Span Length
m
Average life of SFCL (years)
SCB
m’S
Mass
of
(Kg/m)
savings due to extension of
life of circuit breaker by
installation of FCL
Sf
Current
split
factor
determined from the detailed
substation
short-circuit
calculation
Sr
Rated power of SFCL (kVA)
STP
savings due to increase in
transmitted power due to
series capacitor in solid-state
FCL installation
the
sub-conductor
n
Number of sub-conductors
n
Effective number of parallel
conductors in a grid
n
Service life in years
N
Average life of conventional
methods of limiting short
circuit currents (years)
N
Number
of
turns
transformer winding
in
t
Thickness of conductor
transformer (inches)
N
Stiffness norm of installation
with flexible conductors
(1/N)
tc
Fusing time of conductor (s)
p
Interest rate
tf
Fault duration (s)
P0
No load losses
ts
Duration of the shock current
(s)
S-4
in
T
Period
of
conductor
oscillation with no current
flow (s)
Ta
Time constant (s)
TC
Critical temperature for a
superconductor
Tf
Ti
Maximum swing out angle
(rad)
m
Γ
Constant based on type of
short circuit and conductor
location (See Table 2-1):
Stress factor for flexible main
conductor
Final conductor temperature
(°C)
Initial conductor temperature
(°C)
th
Thermal expansion
ela
Elastic expansion
st ,
pi
Strain factor of the bundle
conductor
Ti
Installation temperature (°C)
Tk1
Duration of the first short
circuit current flow (s)
Tsc
Period
of
conductor
oscillation
during
short
circuit (s)
h
Hoop
compression
transformer (psi)
v1,v2, v3, v4, ve
Factors for calculating Fpi
b
Bending stress in transformer
(psi)
w
Applied Load
y0
Initial conductor sag of a
transmission line
yavg
Average conductor sag of a
transmission line
yt
Conductor sag at time t of a
transmission line
α
Coefficient
of
expansion (1/°C)
δ
Direction of the resultant
force for a flexible conductor
(rad)
ρ
Earth resistivity (Ω-m)
ρs
Resistivity of the surface
material (Ω-m)
k
factor for calculating Fpi in
case of non-clashing subconductors
m
s
thermal
Swing out angle at the end of
the short circuit (rad)
S-5
in
Bending stress caused by
forces between main rigid
2
conductors (N/m )
Bending stress caused by
forces between subconductors (N/m2)
electricity demand, utilities have been
upgrading their systems for higher
power transfer capability resulting in
higher fault currents.
1
INTRODUCTION
Meshed Networks- Present networks are
getting more interconnected for the
purpose of enhancement of reliability
and flexibility in the power transmission.
A more closely coupled system not only
exhibits reduced source impedance
values from parallel paths but also an
increased number of sources possibly
contributing to a fault.
Background
With the growth of the electricity demand,
utilities have been upgrading their systems
continuously for higher power transfer
capability and, consequently, for higher fault
current handling capability. A more closely
coupled system not only exhibits reduced
source impedance values from parallel paths
but also an increased number of sources
possibly contributing to a fault. Under
increased power flow conditions on existing
assets, managing fault currents is crucial in
order to avoid malfunctioning and damage
of equipment as well as to increase system
reliability.
New Generation – The addition of new
conventional and distributed generation
(thermal solar power and photovoltaic
systems, wind generators, fuel cells,
microturbines, combustion turbines etc.)
to existing generation is constantly
increasing. The addition of distributed
generators results in increased fault
currents throughout the distribution
system.
This guidebook is intended to be a
comprehensive resource on the subject of
managing high fault current levels. The
adverse impacts of the increased current
levels on individual equipments in the
system are addressed in detail. The
conventional as well as emerging solutions
to the resultant problems are investigated
and the recent developments in the field
have been reported and updated. The
theoretical explanations and mathematical
equations in the text have been
complemented by several applets that have
been developed and can be provided with
this guidebook for the benefit of the readers.
Guidebook Structure
This introductory chapter provides a broad
overview of this guidebook on the subject of
fault current management. The structure of
the rest of the guidebook is as follows.
Chapter 2 deals with the mechanical and
thermal impacts of high fault current levels
on substation bus-work and conductors.
There are separate sections for rigid and
flexible bus bar configurations. Both, IEEE
and CIGRE methods for calculating the
thermal and mechanical forces are explained
in detail. The spreadsheets have been
developed that model the mathematical
equations that have been mentioned in this
chapter. These spreadsheets may be used to
evaluate the impact of high fault current on
example systems. The results have been
validated using the test systems and the
same have been included in Appendix to the
chapter for reference. The impact of the high
Reasons for Increased Fault Currents
There are several reasons for ever increasing
fault current levels in transmission and
distribution systems.
Increased Power Transfer Capability – In
order to meet the ever increasing
1-1
fault currents on substation insulators is also
dealt with. The chapter also includes the
results of an experimental study that
investigated the mechanical impact of shortcircuit on outdoor HV substations.
Chapter 5 deals with the impacts of fault
currents on grounding grids. It addresses the
adverse impacts of the flow of the excessive
fault current into the grounding system in
the substation. The various measures that
may be used to strengthen the grounding
grid are introduced. The procedure that is
used to design the ground grid for expected
fault current levels is also explained using
an example IEEE system. The theoretical
content in the chapter has been
complemented by several case studies that
deal with the grounding issues in increased
fault current situations.
There are growing instances in utility
distribution and transmission systems
wherein the fault current levels are
exceeding the interrupting capability of
existing substation circuit breakers. This
increase in fault current level can cause
significant challenges for utilities as it either
requires the replacement of large number of
substation breakers or the development of
some means to limit the fault current. The
impact of high fault current levels on circuit
breakers is addressed in Chapter 3. The
breakers are classified based on the
interrupting medium used. The interrupting
ratings of the switching devices based on
their ANSI ratings have been included. In
case it is not feasible to replace the breakers,
the procedure that may be used to identify
the problem topologies in a substation that
could be avoided is presented. Finally, some
case studies that deal with the replacement
of breakers in response to increased fault
current levels have been included.
The impacts of high fault current levels on
transmission lines are addressed in Chapter
6. The impact of the high fault current on the
insulators is addressed. The chapter talks
about conductor motion during high fault
current that can be an issue for compact
transmission lines. The mathematical
equations have been provided for conductor
motion for horizontally spaced as well as
vertically
spaced
conductors.
The
mechanical forces experienced by the
spacers during a fault are also investigated.
The mechanical impacts of short-circuit
forces inside the power transformers are
evaluated in Chapter 7. Through fault
capability curves are explained for different
categories of liquid immersed and dry-type
transformers. The expressions for the
mechanical forces due to the short circuit
currents for shell form and core form
transformers are presented and various
failure modes are discussed. Finally, the
combined effects of thermal aging and
mechanical forces on the transformer life are
discussed.
Chapter 4 addresses the impacts of high
fault current on protection and metering. CT
saturation at high fault current levels is the
primary issue and the phenomenon is
investigated in detail. The impacts of the
resultant distortion of CT secondary current
on individual and coordinated protection
schemes have been evaluated. The
spreadsheets have been developed that may
be used to predict CT ratio errors arising out
of AC and DC saturation that may occur
during fault conditions based on CT
characteristics and system parameters.
Example cases have been developed on
EMTP platform that can be used to study CT
saturation in example systems in timedomain. The modeling techniques for the
accurate simulation of CT saturation
phenomenon have also been explained.
Chapter 8 deals with the various methods
that may be used for limiting fault currents
to acceptable levels in transmission and
distribution systems. The chapter includes
description of conventional methods such as
neutral grounding resistors, current limiting
reactors, increased transformers impedances,
1-2
and splitting of grids and bus-bars.
Furthermore, emerging fault current limiting
technologies such as superconducting and
power electronic devices have been included
as well. For the sake of completeness other
novel technologies have also been included.
The status of the various research projects
related to fault current limiters has been
included in this chapter. Also, the case
studies that describe potential and actual
applications of fault limiting technologies
have been included from all over the world.
The technical and economic analysis of
various fault current limiting applications is
addressed in Chapter 9. The comparison is
made of the cost of conventional solutions
of fault limiting such as upgrading breakers,
transformers and adding line reactors. Two
simplified example distribution systems
have been used for the purpose of economic
analysis. Finally, the methodologies that
could be used to estimate the costs and
savings associated with solid-state and
superconducting fault current limiters are
explained.
For the reader’s benefit, the list of symbols
used and the glossary of key terms in the
guidebook have also been provided.
1-3
Effects of High Fault Current on
Substation Conductors
2
Rigid Bus Bars- IEEE Standard
MECHANICAL FORCES
AND THERMAL
EFFECTS IN
SUBSTATION
EQUIPMENT DUE TO
HIGH FAULT
CURRENTS
Short Circuit Mechanical Forces – Circular
Cross Section
The force imparted to the bus structure by
fault current is dependent on conductor
spacing, magnitude of fault current, type of
short circuit, and degree of short-circuit
asymmetry. The equation for the force
between parallel, infinitely long conductors
in a flat configuration due to an
unsymmetrical short circuit current as per
IEEE Standard 605 (IEEE 1998.) is:
All substation equipment can suffer
damage from mechanical forces
caused by short circuits. Examples
include bending of buswork, sudden
expansion of transformer coils, and
breaking of insulators and bushings.
This chapter mainly reviews the
literature for the effects of fault
Introduction
current on the substation bus work
and conductors, leaving the effect on
other substation equipment to latter
chapters.
FSC = K f
(
CΓ D f 2 I SC
)
2
D
Eq. 2-1
Where:
C
= 0.2 × 10-4
FSC = fault current force in N/m.
ISC = symmetrical rms fault current in A.
D = conductor center-to-center spacing
in cm.
Γ = constant based on type of short
Mechanical forces and thermal effects
produced by high fault currents can damage
or destroy substation equipment. In addition,
even faults with rather moderate magnitude
may cause long-term effects such as
accelerated aging of dielectric insulation due
to repetitive mechanical stresses. Solving the
problem of increased fault current means
repeating portions of the original design
process. Because substation design has
become an automated procedure (Anders et
al. 1992.) the uprating should be in the
nature of a design review.
circuit and conductor location (See
Table 2-1).
Df = decrement factor that takes into
account the system impedance.
T
Df = 1+ a
tf
−2 t f
⎛
⎜
Ta
⎜1 − e
⎜
⎝
⎞
⎟
⎟
⎟
⎠
Eq. 2-2
Where:
tf
= fault duration in s.
Ta = time constant X/2πfR in s.
2-1
FG
= total bus unit weight,
including ice loading and connectors
in N/m.
Kf = mounting structure flexibility factor.
This is 1.0 unless mounting structures
are higher than 3 m tall (Figure 2-1).
= 1 if the bus conductors are
KV
vertical, otherwise 0.
Table 2-1
Constant Γ
KH
= 1 if the bus conductors are
horizontal, otherwise 0.
For practical buses, the results will be
conservative due to the end effects. The
short circuit forces on a conductor are added
to the other forces to produce a total force,
which must be less than the minimum yield
stress of the conductor material, Figure 2-1
(IEEE 1998. Section 11) The magnitude of
the total force is:
FT =
( FW + K H F SC ) + ( FG + K V F SC )
2
Figure 2-1
Constant Kf (IEEE. 1998. Figure 4)
Table 2-2
Allowable Stress for Common Conductor
Materials (IEEE. 1998. Table 4)
2
Material
Eq. 2-3
The angle of the force below the horizontal
is:
⎛ FG + KV FSC
⎝ FW + K H FSC
θ = tan −1 ⎜⎜
⎞
⎟⎟
⎠
Eq. 2-4
Where:
FT = total vector force on the bus in
N/m.
FW = wind force in N/m.
2-2
Minimum Yield
Stress kPa2
Al alloy 6063-T6 or
6101-T6
172,375
Al alloy 6061-T6
241,325
Al alloy 6061-T6
103,452
Cu No. 110 hard
drawn
275,800
The maximum When increased forces
allowable
are anticipated due to
length between increased short circuit
spans is limited levels, supporting them
by
the with additional
maximum of insulators can protect
vertical
substation bus bars
deflection or
the span length
for fiber stress. The vertical deflection is
primarily an aesthetic concern, which is not
affected by short circuit forces. The
maximum allowable length based upon fiber
stress is calculated from the maximum
allowable
stress
in
Table 2-2:
LS = 3.16
K S FA S
FT
Eq. 2-5
Where:
LS = maximum length of the bus in cm.
Figure 2-2
Dwight Curves for Proximity Factor
KS = constant based on number of spans
and end types (Table 2-3).
Table 2-3
Conductor Span Constant Ks
FA = maximum yield stress in kPa2.
S
3
= section modulus in cm .
2-3
Number of
Spans
Fixed
ends
Pinned
ends
KS
1
0
2
8
1
2
0
12
1
1
1
8
2
N/A
N/A
8
3
N/A
N/A
10
4+
N/A
N/A
28
Short Circuit Mechanical Forces –
Rectangular Cross-Section
E = modulus of elasticity (Table 2-4)
in kPa.
In the case of substations with rectangular
cross-section bus bars, a proximity factor K
is used (CDA. 2001.):
J = moment of inertia of the cross4
section in cm .
FSC = K
2
µ0 8 Γ (D f I SC )
2π
D
m = mass per unit length in kg/m.
If the resonant frequency calculation leads to
the suspicion of a possible resonance
problem, a dynamic (Bergeron et al. 1999)
or static (Bergeron and Trahan. 1999) finite
element analysis should be performed.
Eq. 2-6
Where, K is taken from Figure 2-2. K is
equal to 1.0 for a round conductor, and is
almost 1.0 for a square conductor.
Conductor shape is most significant for thin,
strip conductors.
In the case of hollow bus bars, internal
weights or stiffeners may be added to
dampen vibration modes.
Dynamic Effects of Short Circuits
Table 2-4
Modulus of Elasticity for Common Conductor
Materials (IEEE. 1998. Table 2)
When excited The stimulus of a short
by
a circuit will provide a
displacement
twice power frequency
force, a rigid periodic force which may
conductor will be amplified if the
vibrate at its natural frequency of the
natural
bus bar is greater than
frequency,
or equal to the power
subject
to frequency.
damping
forces. (IEEE
1998. Section 7). If additional supports are
being added to stiffen the bus due to
increased fault currents, the dynamic effects
should be checked as well. The bus bar
natural frequency is:
fb =
π K2
2
20 L
EJ
m
Material
Modulus of
Elasticity, kPa
Al alloy 6061-T6
Al alloy 6063-T6
6.895 x 107
Al alloy 6101-T6
Cu
11.03x 107
Short Circuit Thermal Effects
Heating of bus bars can cause annealing,
thermal expansion or damage to attached
equipment. The limit for thermal expansion
adopted during the design of the substation
is generally used. Annealing can occur at
temperatures of 100 °C or more.
Eq. 2-7
Where:
The amount of current required to heat a
conductor from the ambient to a given final
temperature during the duration of a fault
can be calculated (IEEE 1998. Section 5.2.):
fb = natural frequency of the bus in Hz.
K = pinning factor:
- 1.00 if both ends are pinned.
- 1.22 if one end is fixed and one end
is pinned.
- 1.51 if both ends are fixed.
2-4
∆L = change in bus bar length in m.
T f − 20 + (K G )
1
I = C × 10 A
log10
tf
Ti − 20 + (K G )
6
α = coefficient of thermal expansion in
1/°C.
Eq. 2-8
If both ends of the bus bar are fixed, then a
force results, which may damage attached
equipment:
Where:
I =
maximum
allowable
symmetrical fault current in A.
rms
(
FTE = 0.1AEα Ti − T f
A = conductor cross-section area, sq.
mm.
A = cross-sectional area in cm2.
Rigid Bus Bars – IEC Standard
= fault duration in s.
Calculation of Electromagnetic Force
Tf = final conductor temperature in °C.
During a three-phase short-circuit in a threephase system, the maximum force is
experienced by the central conductor and
may be computed as per the equation below.
Ti = initial conductor temperature in
°C.
K = constant: 15150 for aluminum,
24500 for copper.
Fm 3 =
G = International Annealed Copper
Standard (IACS) conductivity in percent.
Where:
When the ends
Where thermal expansion
of a bus bar are
is anticipated due to
not
fixed,
increased fault currents,
thermal
expansion fittings can be
expansion will added to long bus
result
from structures.
heating. This
may cause damage to attached equipment,
such as switches, insulators, and other
devices. The amount of expansion (IEEE
1998. Section 13) is:
(
∆L α T f − Ti
=
Li
1 + α Ti
)
Eq. 2-10
Where:
C = constant: 92.9 for aluminum
conductors, 142 for copper conductors.
tf
)
µ0 3 2 l
i p3
2π 2
am
Eq. 2-11
ip3 = peak value of short-circuit current
in the case of balanced 3-phase fault
in A.
l = maximum center-line distance
between supports in m.
am = effective distance between main
conductors in m.
The effective distance am may be computed
as:
- a m = a , for circular conductors
Eq. 2-9
- am =
Where:
Li = initial bus bar length in m.
a
, for rectangular conductors
k12
Where, k12 shall be taken from Figure 2-3.
Ti = installation temperature in °C.
2-5
1
1
1
1
=
+
+ ... +
a s a12 a13
a1n
conductors
-
k
1 k12 k13
=
+
+ ... + 1n
a s a12 a13
a1n
conductors
-
Where, k12……k1n shall
Figure 2-3.
for
circular
for rectangular
be
taken
from
Calculation of Stresses
For the rigid conductors the, axial forces
may be disregarded. In that case, the
bending stress equation for the main
conductors is given below.
Figure 2-3
Factor k12 for computing effective distance
σ m = Vσ Vr β
Similarly, the maximum force between
short-circuited conductors during a line-line
short circuit in a 3-phase system or in a 2line 1-phase system is given below.
Fm 2 =
µ0 2 l
i p2
2π
am
S
V ,Vr are the dynamic factors and may be
obtained from Table in Appendix 2.1. Value
for may be taken from Table in Appendix
2.2. The bending stress equation for the subconductors is given below.
σs = VσsVrs
Then, the maximum force between coplanar
sub-conductors is given below.
2
⎞ ls
⎟⎟
⎠ as
= Section modulus of the main
conductor.
Eq. 2-12
ip2 = peak value of short-circuit current
in the case of line-line fault in A.
⎛ ip
⎜⎜
⎝n
Eq. 2-14
Where:
Where:
µ
Fs = 0
2π
Fm l
8S
Fs l s
16 S s
Eq. 2-15
where
Eq. 2-13
Ss = Section modulus of the subconductors.
Where:
ip = ip2 or ip3
Permitted Conductor Stress
ls = maximum center-line distance
between adjacent connecting pieces in m
The stress caused by short-circuit forces is
permissible for a single conductor if the
following condition is satisfied.
as = effective distance between subconductors in m
σ m < qR p 0.2
The effective distance as may be computed
as:
Where:
2-6
Eq. 2-16
Unwanted forces on support structures
q = factor taken from Table in
Appendix 2.3.
Increased thermal
conductors
Rp0.2= Stress corresponding to yield point
2
in N/m .
Eq. 2-17
σs <= R p 0.2
Eq. 2-18
the
Possibility of damage due to increased
drop force
Pinch effect damage to conductors due to
clashing, and to spacers due to
compression, and to suspension
insulators and supports due to impulse
tension.
Forces on Supports
The dynamic force on the support of the
rigid conductor is calculated as per the
equation below.
Fd = VF Vr αFm
on
Possibility of arcing due to decreased
minimum
clearance
between
conductors during swing
In the case of a conductor comprising of
sub-conductors, the following conditions
should be met.
σ m + σ s < qR p 0.2
stress
Flexible conductor substation buses are
discussed in detail in IEC Standard 865 (IEC
1993 and 1994) with further explanations in
(CIGRE. 1996). The design standard is
intended for horizontal buses up to 60 m
long in a temperature range of –20 to +60 ºC
and maximum sag of 8%. Automatic
reclosing does not increase the effect of
short circuits on flexible conductors. The
simplified calculation procedure used in the
standards has been verified by tests and
detailed finite element method (FEM)
simulations. (Stein, Miri and Meyer. 2000;
Miri and Stein. 2003; Herrmann, Stein, and
Keißling. 1989).
Eq. 2-19
Where, the maximum value of VFVr may be
taken from the table in Appendix 2.1. Also,
value of is dependent on the type and
number of supports and may be taken from
the table in Appendix 2.2.
Flexible Conductor Buses – Static Method
Flexible conductor buses may be
constructed as strain buses, suspended from
insulator strings (Figure 2-4). This type of
construction is usually used for substation
main buses at high voltages (> 100 kV). The
slack bus construction (Figure 2-5), with
post insulators, is normally used for
connections between equipment within a
substation. When high current carrying
capability is needed, conductors are often
bundled (Figure 2-6), separated from 8 to 60
cm with spacers at regular intervals of 2 to
30 meters. (CIGRE 1996. p. 12)
Conductor Motion During a Fault
The electromagnetic force per unit length on
flexible conductors for a phase-to-phase and
three-phase fault is approximately given as:
FSC =
The effects of high fault currents on flexible
conductor buses are:
2
µ 0 0.75I SC
lc
a
l
2π
Eq. 2-20
Where:
ISC = short circuit current.
Increased tension on the conductors
lc = cord length of the main conductor.
Increased tension on insulators
2-7
a = center line distance between main
conductor mid-points.
l
l = maximum center line distance
between supports.
lc
The forces
will
also The forces will cause the
cause
an conductors to separate
(swing out), gravity will
elastic
then bring them together
expansion
of
the (drop force), and they will
oscillate
with
a
conductor
characteristic period.
material,
while the
high currents will cause thermal expansion.
If the bus is constructed from bundled
conductors instead of single conductors,
then the short circuit will force them
together through the pinch effect, which
produces tension on the conductor.
bc
Figure 2-5
Slack Bus From Post Insulators
Spacers
as
ds
bc
li
lc
ls
n sub-conductors
li
l
Figure 2-6
Details of Flexible Conductor Bundle With
Spacers
IEC standard (IEC 1993. Section 2.3)
defines the ratio of electromagnetic force
from the short circuit to the weight of the
conductor:
r=
Figure 2-4
Strain Bus From Suspension Insulators
FSC
n m ′s g
Eq. 2-21
Where:
n
= number of sub-conductors.
m’S = mass of the sub-conductors in
kg/m.
2-8
g
= gravitational constant in m/s2.
⎧ ⎡
⎛
T
⎪δ ⎢1 − cos⎜⎜ 2π k 1
⎪
⎝ TSC
δ k = ⎨ ⎢⎣
⎪
⎪2δ
⎩
Then the direction of the resultant force is:
δ = arctan r
Eq. 2-22
⎞⎤
T
⎟⎟⎥ for 0 ≤ k 1 ≤ 0.5
TSC
⎠⎥⎦
T
for k 1 > 0.5
TSC
The static conductor sag, without current
flow, is:
Eq. 2-26
Where:
n m ′s g l 2
bc =
8Fst
Eq. 2-23
Tk1 = duration of the fault in s
Using r, we calculate a quantity χ,
Where:
⎧1 − r sin δ k for 0 ≤ δ k ≤ π 2
for δ k > π 2
⎩1 − r
Fst = static force on the conductors in N.
χ=⎨
Eq. 2-27
and the maximum swing-out δm:
for 0.766 ≤ χ ≤ 1
⎧1.25 arccos χ
⎪
δ m = ⎨π 18 + arccos χ for - 0.985 ≤ χ ≤ 0.766
⎪π
for χ < -0.985
⎩
Eq. 2-28
Once the angle of displacement is known,
the objective is to determine the minimum
conductor clearance during the fault. This
will require calculating the conductor tensile
force and the thermal and elastic expansion.
Figure 2-7
Curves for Determining the Factor ψ From
IEC Standard 865, Figure 7
If the conductor oscillates at small swing-out
angles, again, with no current flow, the
period of oscillation is:
T = 2π 0.8
bc
g
Short-circuit tensile force on the conductors
may be calculated as follows:
1. Calculate the stiffness norm N
Eq. 2-24
N=
And with a short circuit:
1
1
+
Sl nE S AS
Eq. 2-29
Where:
TSC
T
=
⎡ δ2⎤
4
1 + r 2 ⎢1 −
⎥
⎣ 16 ⎦
Eq. 2-25
S = spring constant of both supports in
N/m.
ES = actual Young’s modulus in N/m2.
The maximum angle which the conductor
can swing out for a given short circuit
current magnitude and fault duration can
now be calculated. The swing-out angle at
the end of the short-circuit is:
AS = cross section of one sub-conductor
in m2.
Calculate the stress factor ζ
2-9
ζ =
(ngm s′ l )2
between Midpoints of a Slack Bus
Eq. 2-30
24 Fst3 N
The thermal expansion of the conductors is
given by:
Calculate the load parameter ϕ
2
⎡ I ′′ ⎤
ε th = cth ⎢ k 3 ⎥ T
⎣ nAS ⎦
⎧ ⎛
2
⎞
for 0 ≤ δ ≤ π 2
⎪3⎜ 1 + r − 1⎟
ϕ=⎨ ⎝
⎠
⎪⎩3(r sin δ + cos δ − 1) for δ > π 2
Eq. 2-35
Where:
Eq.
⎧T
⎫
T = max ⎨ sc , Tk1 ⎬
⎩ 4
⎭
2-31
Determine the factor ψ from Figure 2-7
(Figure 7 in IEC Standard 865), or
calculate from a real solution of
I k′′3 is the initial three-phase symmetrical
rms short circuit current in A
⎧ϕ 2ψ 3 + ϕ (2 + ζ )ψ 2 + (1 + 2ζ )ψ − ζ (2 + ϕ ) = 0
⎨
⎩0 ≤ ψ ≤ 1
⎧0.27 × 10 −18 m 4 /( A 2 s) for Al, Al alloy and Al/Steel
⎪
⎪conductors with a cross - section ratio of Al/St > 6
c th = ⎨
−18
for Al/St ≤ 6
⎪0.17 × 10
−
18
⎪0.088 × 10 for Cu
⎩
Eq. 2-32
Calculate the conductor bundling factor,
KB
⎧1.0
KB = ⎨
⎩1.1
for n = 1
for n ≥ 2
Eq. 2-37
The elastic expansion of the conductors is
given by:
Eq. 2-33
ε ela = N (Ft − Fst )
Then
Ft = K B Fst (1 + ϕψ )
Eq. 2-36
Eq. 2-38
The thermal and elastic expansions are
combined into an expansion factor CD:
Eq. 2-34
2
3⎡ l ⎤
C D = 1 + ⎢ ⎥ (ε els + ε th )
8 ⎣ bc ⎦
a
Eq. 2-39
bh
amin
The actual shape of the conductor sag can
vary from a catenary at low current to a
triangle at high short circuit currents. A form
factor CF is used:
δm
for r ≤ 0.8
⎧1.05
⎪
C F = ⎨0.97 + 0.1r for 0.8 ≤ r < 1.8 Eq. 2-40
⎪1.15
for r ≥ 1.8
⎩
The maximum horizontal displacement
(Figure 2-8) for a slack bus (l c = l ) is:
Figure 2-8
Horizontal Displacement and Distance
2-10
for δ m ≥ π 2
⎧C F C D bc
Eq. 2-41
bh = ⎨
⎩C F C D bc sin δ m for δ m < π 2
lS = distance between two adjacent
Eq. 2-42
spacers
If there are no more than four subconductors, which clash according to the
above definition, then the force Ft, described
above, may be used. Otherwise, the force
Fpi, as described below, should be
calculated.
For a strain bus (l c = l − 2li ) :
⎧C F C D bc sin δ
bh = ⎨
⎩C F C D bc sin δ m
for δ m ≥ δ
for δ m < δ
Eq. 2-43
The closest approach of the conductors is:
a min = a − 2bh
In order to calculate Fpi the following
quantities are required:
Eq. 2-44
1. The maximum of the single phase to
ground and the three-phase short
circuit current
Where:
a = distance between the centers of the
conductors while at rest in m
I ′′ = max (I k′′1 , I k′′3 )
Once the conductors have reached their
maximum
height,
they
will
fall,
experiencing the drop force:
F f = 1.2 Fst 1 + 8ζ
Short-circuit current force
Fv = (n − 1)
δm
for r > 0.6 and δ m ≥ 70 o
π
v1 =
Pinch Forces on Bundled Conductors
2
⎛ I ′′ ⎞ l s v 2
⎜ ⎟
⎜ n ⎟ a v
⎝ ⎠ s 3
In bundled conductor
configurations (Figure
1-5), short circuit
forces cause the subconductors to come
together rapidly.
(a s − d s )m s′
f
sin
Eq. 2-48
π
µ0
2π
n
2
⎛ I ′′ ⎞ n − 1
⎜ ⎟
⎜ n ⎟ a
s
⎝ ⎠
Eq. 2-49
Where, v2 and v3 are given by Figure 2-9
and Figure 2-10 (Figures 8 and 9 in IEC
Standard 865).
First strain factor for bundle contraction
Sub-conductors are said to clash when:
⎧a s d s ≤ 2.0 and l s ≥ 50a s
⎪
or
⎨
⎪a d ≤ 2.5 and l ≥ 70a
s
s
⎩ s s
µ0
2π
Factor v1
Eq. 2-45
This discussion
applies to n
sub-conductors
arranged in a
circular
configuration,
separated by a
distance aS.
Eq. 2-47
ε st = 1.5
Eq. 2-46
Fst l s2 N
(a s − d s )2
⎛π ⎞
sin 2 ⎜ ⎟
⎝n⎠
Eq. 2-50
Second strain factor for bundle
contraction
Where:
ε pi = 0.375
as = clearance between mid-point of
adjacent sub-conductors
Fv l s3 N
⎛π ⎞
sin 3 ⎜ ⎟
(a s − d s )3 ⎝ n ⎠
Clashing factor
dS = diameter of a flexible conductor
2-11
Eq. 2-51
ε pi
1 + ε st
Eq. 2-52
Overswing factor for non clashing sub
conductors
If j≥1 the sub-conductors clash. Go to step
7.
⎛ ls ⎞
µ ⎛ I ′′ ⎞
1 ⎡9
⎟⎟
ve = + ⎢ n (n − 1) 0 ⎜ ⎟ N v2 ⎜⎜
2 ⎣8
2π ⎝ n ⎠
⎝ as − d s ⎠
j=
2
4⎛ π
Factor v4
as − d s
ds
12
⎤
⎞
sin ⎜ ⎟ ⎧
⎥
⎫
arctan
v
1
n
⎪
⎪
4
⎝ ⎠ 1−
⎥
−
⎬
⎨
η 4 ⎪⎩
v4
⎪⎭ 4 ⎥
⎥
⎦
If j<1 the sub-conductors do not clash. Go
to step 10.
v4 =
4
Eq. 2-57
Eq. 2-53
Overswing factor for clashing of sub
conductors
⎛ ls
µ ⎛ I ′′ ⎞
1 ⎡9
+ n (n − 1) 0 ⎜ ⎟ N v 2 ⎜⎜
2 ⎢⎣ 8
2π ⎝ n ⎠
⎝ as − d s
2
ve =
⎤
⎛π ⎞
sin 4 ⎜ ⎟ ⎧
⎥
⎫
⎝ n ⎠ ⎪1 − arctan v 4 ⎪ − 1 ⎥
⎨
⎬
ξ 3 ⎪⎩
v4
⎪⎭ 4 ⎥
⎥
⎦
⎞
⎟⎟
⎠
4
12
Eq. 2-54
Figure 2-9
Curves for Determining the Factors v1 and
v2 From IEC Standard 865, Figure 8
Where:
ξ is the pinch force factor taken from
Figure 2-11 (Figure 10 of IEC Standard
865).
The tensile force is then
⎛
v ⎞
F pi = Fst ⎜⎜1 + e ξ ⎟⎟
ε
st ⎠
⎝
Eq. 2-55
Factor v4
v4 = η
as − d s
a s − η (a s − d s )
Eq. 2-56
Where:
Figure 2-10
Curves for Determining the Factors v1 and
v2 From IEC Standard 865, Figure 9
η = non-clashing sub conductor factor
taken from Figure 2-12, Figure 2-13 and
Figure 2-14 (Figure 11 of IEC Standard
865).
2-12
Figure 2-11
Curves for Determining the Factor ξ as a
Function of j and ε From IEC Standard 865,
Figure 10
Figure 2-13
Curves for Determining the Factor η From
IEC Standard 865, Figure 11
Figure 2-14
Curves for Determining the Factor η From
IEC Standard 865, Figure 11
Figure 2-12
Curves for Determining the Factor η From
IEC Standard 865, Figure 11
The tensile force is then
⎞
⎛
v
F pi = Fst ⎜⎜1 + e η 2 ⎟⎟
⎠
⎝ ε st
2-13
Eq. 2-58
Spacer compression may be calculated with
the Manuzio formula (Lilien et al. 2000.):
⎛a
Pmax = 1.45 I ′′ Fst log⎜⎜ S
⎝ dS
⎞
⎟⎟
⎠
Eq. 2-59
Where:
Pmax = Compression force on the spacer
in N.
Fst = Initial static tension on the
conductor bundle in N.
as = conductor spacing in mm.
Figure 2-15
Operation of Force Safety Device (Miroshnik.
2003)
ds = conductor diameter in mm.
Tests by Lillien, et al., showed that the
Manuzio formula underestimated the stress
by 50%. Better results (within ±10%) were
obtained using finite element analysis.
When short circuit forces increase, force
safety devices can mitigate pinch force
effects.
Force Safety Devices
A Force Safety Device (FSD) (Miroshnik.
2003) is a deformable mechanical link
which can be placed in series with a flexible
substation bus to limit damage due to shortcircuit forces. It is similar to a fuse, in that it
is non-recoverable, and must be replaced
after a short circuit. The principle of
operation (Figure 2-15) is that of a metallic
cramp, having two weakened cross-sectional
areas, which are calibrated for an actuation
force of Pdev. The FSD is connected
between the support structure and the
suspension insulator, Figure 2-16. When the
total force Ft exceeds Pdev, the FSD will be
deformed, as shown in Figure 2-15, limiting
the force. A graph of force limitation is
plotted in Figure 2-17.
Figure 2-16
Connection of FSD to Flexible Substation
Bus Structure. (Miroshnik. 2003)
Figure 2-17
Limitation of Bus Tension by FSD.
(Miroshnik. 2003)
2-14
Substation Cable and Conductor Systems
2
T2 + Tk
⎛I⎞
⎜ ⎟ t = 0.0297 log10
A
T1 + Tk
⎝ ⎠
There are many types of cables and
conductors used in substations. (IEEE.
1992.) These include:
Eq. 2-60
Where,
I = symmetrical short circuit current in
A
1. High-voltage power cables, defined as
> 1000 V. These may connect to
other substations, to substation
equipment, or to customer loads.
A = conductor cross-section in circular
mils
T1 = initial conductor temperature in
°C, the maximum continuous conductor
temperature for the insulation system is
used, typically 75 or 90 °C for low
voltage cables.
2. Low-voltage power cables, defined as
< 1000 V. These supply auxiliary
power to substation equipment.
3. Control cables, including instrument
transformer secondary cables
T2 = final conductor temperature in °C,
the short-circuit temperature limit of the
insulation system is used, typically 250
°C for low voltage cables.
4. Instrumentation cables, primarily for
SCADA systems.
5. Overhead secondary conductors. In
distribution
substations,
these
medium voltage open wire lines are
the termination points of distribution
feeders.
Tk = heating temperature constant for
the conductor material, 234 °C for Cu
and 228 °C for Al.
Similar limits are available for the sheaths of
medium voltage cables. They should be
used for ground fault currents. In the case of
increased fault current levels, protective
relay settings should be changed, as
necessary, to protect the cables. If this is not
possible, resizing of the cable may be
necessary.
Cable Thermal Limits
Protective
Cables are subject to
relay
thermal damage from
operating
times
and prolonged exposure to
short-circuit currents.
circuit-breaker
clearing times
must be fast enough to prevent prolonged
overheating. (IEEE. 1993c. Section 5.6.2.)
Although the protection requirements of the
National Electrical Code (NEC) (NFPA.
2005.) do not apply in most substations, they
should be considered when evaluating cable
protection systems.
Cable Mechanical Limits
When a short duration fault bends a cable,
the mechanical effect is more significant
than the thermal (Rüger. 1989.). Permanent
deformation may occur to plastic insulated
single-core cables. When cleats confine a
cable, such that short circuit forces create
outward bows with small bending radii, the
cable may be damaged. Friction between the
cleats and the cable may damage the outer
sheath. Softening of the insulation by the
simultaneous heating further increases the
damage caused by bending of the
In addition to the NEC requirements, it is
recommended that cable protection adhere
2
to the I t limits of the cable damage curve
for the insulation type, published by the
cable manufacturers.
2-15
conductors. Proper support of cables can
prevent this type of damage from occurring.
Distribution Line Conductor Motion
When overhead
distribution lines A fault on the
distribution
line
enter
causes
the
overhead
substations, the
conductors to swing
opportunity
exists for the side-to-side closer to
substation to be the substation.
exposed
to
damage from distribution faults. (Ward.
2003.) A fault on the distribution line causes
the overhead conductors to swing side-toside closer to the substation. As a result, the
conductors may move close enough to arc
(0.1 meters) or even touch. Thos causes a
second fault, which may cause increased
stress on the substation transformer and
cause backup protective devices to operate.
Figure 2-18
Critical Clearing Curve for Different Span
Lengths (Ward 2003)
The impact of the phase spacing and
conductor temperature on the critical
clearing times is shown in Figure 2-19 and
Figure 2-20 respectively. It can be seen that
increased phase-phase spacing and lower
conductor temperature result in higher
values for critical clearing time.
Ward prepared a computer program that
calculates critical clearing time curves for
overhead distribution conductors, based
upon conductor motion. The critical clearing
curve has been developed for a typical 34.5
kV distribution line having a phase spacing
of 1.1 m and utilizing 477 kcmil AAC
conductors (Figure 2-18). These curves were
obtained for a given span by applying
different magnitudes of current for a
duration that would cause conductors to
swing into each other close enough to cause
a breakdown in air. The area above the
curve is problem area and needs to be
avoided. It can be seen that longer spans are
likely to be more problematic.
Figure 2-19
Impact of Phase-Spacing on Critical Clearing
Curve (Ward 2003)
2-16
suspension insulators (Burnham et al.
2002.), however, the same problems could
occur in post type insulators. In terms of
short circuit stresses, exceeding the
mechanical loading limits could result in
cracks or splits in the rod or in damaged
seals. Water intrusion leads to brittle
fracture, through the leaching of acids in
combination with tensile stress. (de Tourreil
et al. 2000.)
Rigid Bus Bars
Figure 2-20
Impact of Conductor Temperature on Critical
Clearing Curve (Ward 2003)
High fault current forces on rigid bus bars
are transmitted to supporting insulators,
which will be subject to forces that may
exceed their design limits. The effects on
the insulators could be cracks, fractures or
breakage. These, in turn, will weaken the
bus support structure resulting in greater
damage should a second fault occur before
the damage is repaired. The action of
reclosers is of particular concern here.
Possible solutions to the problem of damage
caused by distribution conductor motion are:
1. Use faster recloser time curves. This
is the preferred solution, if it is
possible.
2. Installing fiberglass spacers at midspan, shortens the effective span.
This is fairly inexpensive.
The short-circuit force on a bus bar is
transmitted to the insulator (IEEE. 1998.
Section 12.) through the bus-support fitting
(Figure 2-21 and Figure 2-22):
3. Adding intermediate poles to shorten
spans. This is expensive.
4. Increase phase spacing. This requires
replacing cross arms, and is
expensive.
FSB = LE FSC
Eq. 2-61
Where:
5. Removing slack in the lines to reduce
conductor motion. This is timeconsuming and expensive.
The first two options, faster reclosing times
and fiberglass spacers at mid-span are the
best alternatives if increased levels of fault
current result in added stress to the
substation due to overhead distribution line
faults.
FSB = bus short circuit force transmitted
to the bus support fitting in N.
LE = effective length of the bus span in
m.
Similarly, the gravitational forces are:
FGB = LE FG
Eq. 2-62
Where:
Effects of High Fault Currents on
Substation Insulators, Supports and
Structures
FG B= gravitational force transmitted to
the bus support fitting in N
FG = weight of the bus in N
Brittle fracture of nonceramic insulators
(NCI) have mostly occurred in polymer
The cantilever force on the insulator is then:
2-17
(
)
H i + H f FWB ⎤
⎡F
FIS = KV K1 ⎢ WI +
⎥
Hi
⎣ 2
⎦
H
H
F
+
⎡F
i
f
GB ⎤
+ K H K 3 ⎢ GII +
⎥
H
2
i
⎣
⎦
⎡ H i + H f FSB ⎤
+ K2 ⎢
⎥
Hi
⎣
⎦
(
(
Bus Support
Fitting
FIS F +F
GB
SB
FGI
)
Insulator
)
Hi/2
Hi/2
Bus
Hf
Eq. 2-63
Figure 2-21
Insulator Configuration for Vertical Bus
Where:
K1 = overload factor for wind forces,
typically 2.5.
Bus
FWB+FSB
Bus Support
Fitting
FIS
K2 = overload factor for fault current
forces this should be 2.5 also, unless
certain resonance criteria are met.
Hf
Insulator
K3 = overload factor for gravitational
forces, typically 2.5.
Hi/2
FWI = wind force on the insulator in N.
FWI
Hi = height of the insulator in cm.
Hf = height of the bus centerline above
the insulator in cm.
Hi/2
If the cantilever force is exceeded as
prospective fault currents increase, two
possible solutions are to increase the number
of insulators, decreasing LE, or replace the
insulators with units having greater
cantilever strength. If insulator spacing is
changed,
the
mechanical
resonant
frequencies will have to be recalculated, and
a dynamic study may need to be performed.
Experimental results (Barrett et al. 2003)
show that the IEEE method is conservative,
and it is unlikely that increased fault
currents will damage an IEEE designed
insulator structure.
Figure 2-22
Insulator Configuration for Horizontal Bus
Flexible Conductor Buses – Static Method
Post Insulators
In accordance with (IEC. 1993.), the
maximum value of Ft, Ff or Fpi shall not
exceed the withstand value of insulators and
their supports. Connectors shall be rated to
withstand the maximum value of 1.5Ft, Ff or
Fpi.
2-18
Chain Insulators
Table 2-5 ). Only conductors were needed to
be modeled for this exercise. It can be seen
that results of dynamic approach are more
accurate as they are closer to the test
measurements.
For this configuration, insulators, their
supports and structures should all be able to
withstand the maximum value of Ft, Ff or Fpi.
Flexible Conductor Buses – Dynamic
Method
The method discussed in the previous
section is based on the assumption that the
static load can be used to represent the
transient impact of the short circuit currents
on bus structures. This means that results of
the approach can be over-conservative and
the same has been supported by field
observations. In order to get a more accurate
assessment of the structures response to the
short circuit currents, a non linear finite
element dynamic incremental analysis
methodology has been presented in (Jesson
et al. 2006). It involves modeling of both
conductors and structures using finite
element analysis software (ADINA version
8.2)
The model was validated by comparing the
measured results for CIGRE test cases (Case
10 and 11 in IEC Standard 865-1) against
the static and dynamic methodology (
2-19
Table 2-6. These are for external faults,
where the GIS is tested in the same manner
as circuit breakers.
Table 2-5
Model Verification for Case 10 (Jesson et al
2006)
Criteria
CIGRE
test
Static
Method
Dynamic
Method
Peak
swing
force
23
14.3
26
Peak drop
out force
34.7
41.5
32
Maximum
swing out
angle
66
48.2
60
The model was extended to include the
supporting structures with foundations, the
latter modeled as elastic blocks onto a rigid
plane. This allowed for the dynamic
simulation of overturning of the structures.
Three different structural arrangements were
investigated for different fault current
magnitudes (31.5 kA and 60 kA) and
durations (150 ms, 350 ms, 450 ms and 1 s).
As per the dynamic analysis, all the three
structures were found to be stable with
adequate safety factor (above 1.5). The same
structures were found to fail the overturning
assessment using the static methodology.
Effects of High Fault Currents on Gas
Insulated Substations (GIS)
Gas Insulated Substations are designed and
tested in accordance to (IEEE. 1993a), and
have short circuit ratings as listed in
2-20
Where:
Table 2-6
GIS Short Circuit Ratings (IEEE C37.1221993)
Short-time current carrying capability (kA,
rms) for a specified time of 1 s or 3 s
20
50
25
63
31.5
80
40
100
t
= time in ms
d
= thickness of the Aluminum in mm
I
= current in kA
A rotating arc can puncture a GIS wall in
two different ways (Boeck. 2003.). If there
is an oblique arc, which rotates, the burst
will be similar to that shown in Figure 2-23.
If an insulating barrier stops the moving arc,
the vertical arc will puncture a hole in a
much shorter time. The internal design of
GIS is intended to keep arcs moving, and to
prevent them from sweeping over the same
locations more than once.
Table 2-7
GIS Phase to Ground Burn-Through Times
(IEEE C37.122-1993)
Phase to ground burn-through times
Current (kA, rms)
Time (s)
< 40
0.2
≥ 40
0.1
The
The internal arcing fault
internal
withstand capability of GIS
arcing
is based upon the thickness
fault
of the metal walls and the
withstand
gas pressure, (IEEE. 1993a),
capability
and is thus not easy to
of GIS is upgrade.
based
upon the
thickness of the metal walls and the gas
pressure, (IEEE. 1993a), and is thus not easy
to upgrade.
Figure 2-23
GIS Enclosure Punctured by a Rotating Arc
(Boeck 2003)
A statistical analysis (Trinh. 1992) shows
that “An increase in the mean fault current
from 20 to 30 kA raises the risk of burnthrough from 0.34 to 0.78, which illustrates
the importance of designing the GIS in
terms of the distribution of the local fault
current.” This is expressed as a probability
formula:
The withstand times are listed in Table 2-7.
IEC standards are similar in regard to both
short circuit ratings and burn-through times.
The time to puncture an aluminum plate is
approximately (Boeck and Krüger. 1992.)
t=C
d2
I
C = (60K500) kA ms mm − 2
R=
Eq. 2-64
∞
∞
0
0
∫ p(ic )∫ P(tc ) p(tc )dtc dic
Where:
2-21
Eq. 2-65
R = risk of burn-through of a GIS
having an envelope thickness d
associated with a fault at a certain
location on the transmission system
fed into the busbar, at the north portal the
busbar was short-circuited.
The lower edge of the crossarm was placed
at a height of 8,22 m above ground for the
100-kV- and 11,22 m for the 400-kVarrangement. The centre-line phase distance
was 2,0 m (100 kV), or 3,0 m respectively
(400 kV). The static tensile force of the span
was adjusted in such a way that the
conductor sag in mid-span was 600 mm
(100 kV) or 800 mm (400 kV). The double
insulator strings used consisted of 2 x 7 (100
kV) or 2 x 24 (400 kV) cap and pin
insulators.
p(ic) = probability density of the local
fault current ic
P(tc) = probability that the burn-through
time will not exceed tc
P(tc) = probability density of the faultclearing times.
When a GIS unit is inspected and
maintained or replaced after a fault, very
specific safety procedures should be
followed
(IEEE.
1993b).
Sulfur
Hexafluoride (SF6) is nontoxic, but
produces numerous toxic byproducts during
arcing and burning.
Experimental Results
The experimental results about the
mechanical impact of short-circuits on the
outdoor HV substations have been presented
(Pitz, V et al. 2004). The research project
was funded by German Federal Ministry of
Economy and Labour and the results were
intended to be used by IEC standardization
committees and CIGRE working groups.
There were two set-ups called”100-kVarrangement” and “400-kV-arrangement”
that differed in the height of cross-arms,
distance between the conductors and static
tensile forces. The experiments were carried
out to see the impact of the magnitude and
duration of short-circuit currents, spacers
and bundling.
Figure 2-24
Test set-up for studying the mechanical
impacts of short-circuits (Pitz, V et al. 2004)
The results showing variation of pinch force
and swing out force as a function of
bundling and the magnitude of short circuit
current are shown in the plots in Figure 2-25
and Figure 2-26 respectively.
A schematic drawing of the test set-up is
shown in Figure 2-24: two lattice-type portal
towers (mid and north) with adjustable
crossarms supported a two-phase flexible
busbar with a centre-line distance between
towers of 40 m (conductor type: ACSR
537/53). At the mid portal the current was
2-22
corresponding calculation. The comparison
is shown for the “100-kV-arrangement” in
Figure 2-27 and for the “400-kVarrangement” in Figure 2-28 respectively. It
is found that for 100-kV arrangement, most
of the calculated values are on the safe side
(above 0% line). For 400-kV tests, most of
the calculated values are safe and the error
of the values under the 0% line is tolerable.
Figure 2-25
Pinch Force Results for 400-kv-arrangement
(Pitz, V et al. 2004)
Figure 2-27
Comparison of Tensile Force between
Calculations and Test Results for 100-kvarrangement (Pitz, V et al. 2004)
Figure 2-26
Swing-out Force Results for 400-kvarrangement (Pitz, V et al. 2004)
The experimental results have been
validated against the analytical calculations
as per the method in IEC Standard 865. In a
first step, the maximum of the tensile forces
Fpi, Ft and Ff is determined which is decisive
for the conductors, clamps, insulators and
their anchoring: the maximum Fm for each
measured case and the maximum Fc for its
Figure 2-28
Comparison of Tensile Force between
Calculations and Test Results for 400-kvarrangement (Pitz, V et al. 2004)
2-23
by adding expansion fittings to the long bus
structures and considering deflection of a
bus conductor, bus-conductor bends,
insulators, or mounting structures for short
buses.
Summary and Recommendations
Increased levels of fault current result in the
increased thermal and mechanical stresses
on the substation buswork and conductors.
The impact can be evaluated using the
equations that have been summarized in this
chapter and have been coded into a
spreadsheet application that has been
developed and provided with this book. The
results of the application have been
validated for few example cases (IEC
Standard 865-2) and have been provided in
the Appendices to this chapter.
In the case of flexible buses, a Force Safety
Device (FSD) that is a deformable
mechanical link can be placed in series with
a substation bus to limit the damage due to
short-circuit forces.
The current standards that deal with the
impact of the fault currents on station bus
structures are based on assumption that
static loading can be used to represent the
transient impact of short circuit forces. It has
been shown through simulations and field
measurements that it results in overconservative results. Dynamic assessment
methodology explained in the chapter is
found to provide more accurate results and
standards need to reflect the same.
If the mechanical forces are found to be
excessive for rigid bus bars, supporting them
with additional insulators can protect them.
In such cases, dynamic effects should be
checked as well. In the case of hollow bus
bars, internal weights or stiffeners may be
added to dampen vibration modes.
Similarly, the increased thermal expansion
due to higher fault currents can be countered
Appendix 2.1 Maximum possible values for dynamic factors (IEC Standard 865-1)
2-24
2-25
Appendix 2.2 Factors for Different Bus-bar Arrangement (IEC Standard 865-1)
2-26
Appendix 2.3 Factor q for Rigid Conductor (IEC Standard 865-1)
2-27
Appendix 2.4 Mechanical Effects on a 110kV arrangement with Slack Conductors
Data
-Three phase initial symmetrical short-circuit current (rms) = 19 kA
- Duration of fault = 0.3 s
- Maximum center line distance between supports = 11.5 m
- Center-line distance between main conductor mid-points = 2.0 m
- Spring constant of the supports = 100000 N/m
- Conductor information
•
Number of subconductors = 1
•
Sub-conductor cross-section = .000242 sq. m
•
Sub-conductor mass per unit length = 0.670 Kg/m
•
Young modulus = 5.5e10 N/sq. m
- Static conductor tensile force at a conductor temperature of -20 deg = 400 N
- Static conductor tensile force at a conductor temperature of +60 deg = 273 N
Results
- Maximum horizontal span displacement = 0.61 m
- Minimum air clearance = 0.78 m
- Design load for insulator and supports = 3035 N
- Design load for connectors = 3985 N
2-28
Appendix 2.5 Mechanical Effects on Strained Conductors
Data
-Three phase initial symmetrical short-circuit current (rms) = 63 kA
- Duration of fault = 0.5 s
- Maximum center line distance between supports = 48 m
- Cord length of the main conductor = 37.4 m
- Center-line distance between main conductor mid-points = 5.0 m
- Spring constant of the supports = 500000 N/m
- Conductor information
•
Number of subconductors = 2
•
Sub-conductor cross-section = .001090 sq. m
•
Sub-conductor mass per unit length = 3.25 Kg/m
•
Sub-conductor diameter = 0.043 m
•
Young modulus = 6e10 N/sq. m
- Static conductor tensile force at a conductor temperature of -20 deg = 23100 N
- Static conductor tensile force at a conductor temperature of +60 deg = 18900 N
Results
- Maximum horizontal span displacement = 1.33 m
- Minimum air clearance = 2.34 m
- Design load for insulator and supports = 79074 N
- Design load for connectors = 79074 N
2-29
Appendix 2.6 Mechanical Effects on a 10kV arrangement with Single Rigid Conductors
Data
-Three phase initial symmetrical short-circuit current (rms) = 16 kA
- Factor for calculation of peak short circuit current – 1.35
- Number of spans = > 3
- Maximum center line distance between supports = 1 m
- Center-line distance between main conductor mid-points = 0.2 m
- Conductor information
•
Rectangular
•
Number of subconductors = 1
•
Dimensions = 0.06 m X 0.01 m
•
Sub-conductor mass per unit length = 1.62 Kg/m
•
Young modulus = 7.0e10 N/sq. m
- Minimum stress corresponding to yield point = 1.2e8 N/m2
- Maximum stress corresponding to yield point = 1.8e8 N/m2
Results
- Bus-bar will withstand the short circuit force
2
- Bending stress = 73 N/mm
- Dynamic bending force for outer support = 631 N
- Dynamic bending force for inner support = 1736 N
2-30
Appendix 2.7 Mechanical Effects on a 10kV arrangement with Multiple Rigid Conductors
Data
-Three phase initial symmetrical short-circuit current (rms) = 16 kA
- Factor for calculation of peak short circuit current – 1.35
- Number of spans = > 3
- Maximum center line distance between supports = 1 m
- Center-line distance between main conductor mid-points = 0.2 m
- Conductor information
•
Rectangular
•
Number of subconductors = 3
•
Distance between spacers = 0.5 m
•
Dimensions = 0.06 m X 0.01 m
•
Sub-conductor mass per unit length = 1.62 Kg/m
•
Young modulus = 7.0e10 N/sq. m
- Minimum stress corresponding to yield point = 1.2e8 N/m2
- Maximum stress corresponding to yield point = 1.8e8 N/m2
Results
- Bus-bar will withstand the short circuit force
- Bending stress ( Total : 40.6 N/mm2, Sub-conductor: 16 N/mm2)
- Dynamic bending force for outer support = 873 N
- Dynamic bending force for inner support = 2400 N
2-31
Appendix 2.8 Mechanical Effects on a High Voltage Arrangement with Rigid Conductors
Data
-Three phase initial symmetrical short-circuit current (rms) = 50 kA
- Factor for calculation of peak short circuit current – 1.81
- Number of spans = 2
- Maximum center line distance between supports = 18 m
- Center-line distance between main conductor mid-points = 5 m
- Conductor information
•
Rectangular Tubular
•
Number of subconductors = 1
•
Outside diameter = 0.16 m
•
Wall thickness = 0.006 m
•
Sub-conductor mass per unit length = 7.84 Kg/m
•
Young modulus = 7.0e10 N/sq. m
- Minimum stress corresponding to yield point = 1.6e8 N/m2
- Maximum stress corresponding to yield point = 2.4e8 N/m2
Results
Without 3-Phase Reclosing
- Bending stress = 156 N/mm
2
- Bus-bar will withstand the short circuit force
- Dynamic bending force for outer support = 4722 N
- Dynamic bending force for inner support = 15742 N
With 3-Phase Reclosing
- Bending stress = 280 N/mm2
- Bus-bar will not withstand the short circuit force
- Dynamic bending force for outer support = 3830 N
- Dynamic bending force for inner support = 12767 N
2-32
CIGRE. 1996. “The mechanical effects of
short-circuit currents in open-air
substations (rigid and flexible busbars),” CIGRE, Paris, CIGRE
brochure no. 105, vol. 1 and 2, Apr.
1996.
References
1. Anders, G.J., G.L. Ford, M. Vainberg,
S. Arnot and M. Germani. 1992.
“Optimization of Tubular Rigid Bus
Design.” IEEE Trans. on Power
Delivery, Vol. 7, No. 3, pp. 11881195.
Copper Development Association (CDA).
2001.
Copper
for
Busbars,
Publication 22, June 1996, Reprinted
2001.
Barrett, J.S., W.A. Chisholm, J. Kuffel,
B.P Ng, A-M. Sahazizian, and C. de
Tourreil.
2003.”Testing
and
Modelling Hollow-Core Composite
Station Post Insulators under ShortCircuit Conditions.”
IEEE PES
General Meeting, 13-17 July 2003.
Vol. 1, pp. 211-218.
Dasbach, A. and G. J. Pietsch. 1990.
“Calculation of Pressure Waves in
Substation Buildings Due to Arcing
Faults.” IEEE Trans. on Power
Delivery, Vol. 5, No. 4, pp. 17601765.
de Tourreil, C., L. Pargamin, G. Thévenet
and S. Prat. 2000. “”Brittle Fracture”
of Composite Insulators: Why and
How they Occur.” IEEE PES Summer
Meeting, 16-20 July, vol. 4, pp. 25692574.
Bergeron, D.A., and R. E. Trahan, Jr.
1999. “A Static Finite Element
Analysis of Substation Busbar
Structures.” IEEE Trans. on Power
Delivery, Vol. 14, No. 3, pp. 890-896.
Bergeron, D.A., R. E. Trahan, Jr., M.D.
Budinich and A. Opsetmoen. 1999.
“Verification of a Dynamic Finite
Element Analysis of Substation
Busbar Structures.” IEEE Trans. on
Power Delivery, Vol. 14, No. 3, pp.
884-889.
de Wendt, G., A.M. Miri, N. Stein, T.
Tietz, A. Kühner, Ph. De Coninck.
1998. “Rigid Conductors With Elbow
Bends
to
Connect
Different
Horizontal Busbar Levels Electrical
Tests and Calculation of Short-Circuit
Stresses.”
6th
International
Conference on Optimization of
Electrical and Electronic Equipment
OPTIM 1998.
Boeck, W. 2003. “Solutions of Essential
Problems of Gas Insulated Systems
for Substations (GIS) and Lines
(GIL).” ICPADM 2003, Nagoya,
Japan.
Herrmann, B., N. Stein, and G. Keißling.
1989. “Short-Circuit Effects in HV
Substations with Strained Conductors
Systematic Full Scale Tests and a
Simple Calculation Method.” IEEE
Trans. on Power Delivery, Vol. 4, No.
2, pp. 1021-1028.
Boeck, W. and K. Krüger. 1992. “Arc
Motion and Burn Through in GIS.”
IEEE Trans. on Power Delivery, Vol.
7, No. 1, pp. 254-261.
Burnham, J.T., et al. 2002. “IEEE Task
Force Report: Brittle Fracture in
Nonceramic Insulators.” IEEE Trans.
on Power Delivery, Vol. 17, No. 3,
pp. 848-856.
IEC. 1993. International Standard 865-1:
1993.
Short-circuit currents—
Calculation of effects. Part 1:
Definitions and calculation methods.
Genève: CEI.
2-33
IEC. 1994. International Standard 865-2:
1994.
Short-circuit currents—
Calculation of effects. Part 2:
Examples of calculation. Genève:
CEI.
Compression for a Triple Conductor
Bundle.” IEEE Trans. on Power
Delivery, Vol. 15, No. 1, pp. 236-241.
Miri, A.M. and N. Stein. 2003.
“Calculated Short-Circuit Behaviour
and Effects of a Duplex Conductor
Bus Variation of the Subconductor
Spacing.” Journal of Electrical
Engineering.
Vol.
3.
2003.
www.jee.ro
IEEE. 1992. Standard 525-1992. IEEE
Guide for the Design and Installation
of Cable Systems in Substations. New
York, NY: IEEE.
IEEE. 1993a. Standard C37.122-1993.
IEEE Standard for Gas-Insulated
Substations. New York, NY: IEEE.
Miroshnik, R. 2003. “Force Safety
Device for Substation With Flexible
Buses.” IEEE Trans. on Power
Delivery, Vol. 18, No. 4, pp. 12361240.
IEEE. 1993b. Standard C37.122.1-1993.
IEEE Guide for Gas-Insulated
Substations. New York, NY: IEEE.
National Fire Protection Association
(NFPA). 2005. Standard NFPA-70.
National Electrical Code. Quincy,
MA: NFPA.
IEEE. 1993c. Standard 141-1993. IEEE
Recommended Practice for Electric
Power Distribution for Industrial
Plants. New York, NY: IEEE.
Pitz, V. et al. 2004.
“Short-circuit
Mechanical Effects on Outdoor HV
substations with Widw Bundling.”
CIGRE, Session 2004.
IEEE. 1998. Standard 605. IEEE Guide
for Design of Substation Rigid-Bus
Structures. New York, NY: IEEE.
IEEE. 2000. Standard 80. IEEE Guide
for Safety in AC Substation
Grounding. New York, NY: IEEE.
Rüger, W. 1989. “Mechanical ShortCircuit Effects of Single-Core
Cables.” IEEE Trans. on Power
Delivery, Vol. 4, No. 1, pp. 68-74.
Jesson, T. et al. 2006. “Assesment of the
Dynamic Performance of flexible
Busbar Systems under Varying Fault
Conditions”. CIGRE, Session 2006.
Stein, N., A.M. Miri and W. Meyer.
2000. “400 kV Substation Stranded
Conductor Buses – Tests and
Calculations
of
Short-Circuit
Constraints and Behaviour—“ 7th
International
Conference
on
Optimization of Electrical and
Electronic Equipment OPTIM 2000.
Brasov (Romania), 11-12 May 2000.
Proceedings pp. 251-258
Kock, B. 1988. “Electromagnetic Forces
in a Plane Six-Bus System.” IEEE
Trans. on Power Delivery, Vol. 3, No.
3, pp. 947-953.
Labridis, D.P. and P.S. Dokopoulos.
1996. “Electromagnetic Forces in
Three-Phase Rigid Busbars with
Rectangular Cross-Sections.” IEEE
Trans. on Power Delivery, Vol. 11,
No. 2, pp. 793-800
Triantafyllidis, D.G., P.S. Dokopoulos
and D.P. Labridis. 2003. “Parametric
Short-Circuit Force Analysis of
Three-Phase
Busbars—A
Fully
Automated
Finite
Element
Approach.”
Lilien, J-L., E. Hansenne, K.O. Papailiou
and J. Kempf. 2000. “Spacer
2-34
Trinh, N. G. 1992. “Risk of BurnThrough – A Quantitative Assessment
of Gas-Insulated Switchgear to
Withstand Internal Arcs.” IEEE
Trans. on Power Delivery, Vol. 7, No.
1, pp. 225-236
Ward, D. J., 2003.“Overhead Distribution
Conductor Motion Due to ShortCircuit Forces.” IEEE Trans. on
Power Delivery, Vol. 18, No. 4, pp.
1534-1538.
2-35
technology based on solid-state switching
for the circuit breakers is introduced.
3
Air Circuit Breakers
EFFECTS OF HIGH
FAULT CURRENTS ON
CURRENT
INTERRUPTING
DEVICES
Every year, the Doble Conference elicits
responses from clients to a technical
questionnaire on various subjects. The
following results on medium-voltage circuit
breaker (CB) failures from 1991–1994
confirm the above assertion:
•
1991 - 133 replies, 54% return, 71
clients reported 165 CB failures
• 1992 - 158 replies, 61% return, 71
clients reported 180 CB failures
• 1993 - 125 replies, 48% return, 44
clients reported 117 CB failures
• 1994 - 109 replies, 42% return, 54
clients reported 109 CB failures
Of particular interest in the Doble findings is
the diversity of the nature of the failures.
These include but are not restricted to:
For a variety of reasons, the available
fault currents on the power systems
exceed the interrupting rating of the
existing circuit breakers.
From the technical point of view, the
most eminent reason for utilities to
maintain or even reduce their fault
current levels is to ensure proper
functioning of circuit interrupting
devices such as circuit breakers and
fuses. However, safety plays a major
role in the need to contain or reduce the
fault current duties on these current
interrupting devices.
•
•
•
•
•
This chapter will summarize some of the
aspects of the current interrupting
devices which are used for medium and
high voltage.
•
•
•
The fault currents in the power system must
be limited to the levels that can be safely
interrupted. Present breaker technology
imposes an upper bound on the fault current
that can be interrupted. In cases that this
capability is exceeded, alternate means such
as avoiding the problem topologies and
incorporating fault limiting technology may
be utilized. This chapter introduces various
circuit breaker technologies that uses
mediums such as air, vacuum and SF6. The
interrupting ratings of the switching devices
are presented. Finally, an emerging
•
•
•
•
3-1
Arc chute melting
Failure to clear faults
Insulation breakdown
Component deterioration
Bushing insulation destruction due to
moisture and electrical stress
Arc chute carbon buildup and
condensation
Interrupter disintegration
Fire destruction caused by failure to
clear
Jamming of the contact blade by arc
chute
delamination
preventing
closure of contacts
Operating rod failure during an
opening operation
Support arm failure
Motor drive and ratchet wheel
failures
•
interrupters, the arc in vacuum circuit
breakers is supported by ionized metal
vapors that are generated by the contacts
themselves. Thus, at contact zero, vapor
condensation and the collapse of ionization
is almost instantaneous.
Operating coil and anti-pump relay
failures
Arc chute
Circuit breakers that are
moisture
25–40 years old can be
contaminati
to
be
on
and considered
approaching
the
end
of
reduced
their
expected
design
life.
circuit
breaker
operating velocities should be a major
concern. Research in 1992 on more than 500
air circuit breakers at major midwestern
utilities showed that approximately 25% of
all circuit breakers exhibited operating
velocities below the manufacturers’
specification, despite the fact that all circuit
breakers had been timed at acceptance and
at approximately five-year intervals
thereafter. The in-service dates of the circuit
breakers ranged from the early 1950s to the
mid-1980s (Genutis, 1992).
Outwardly, the vacuum interrupter appears
to be the ultimate in switching devices that
are independent of the operating system, as
its breaking capability is dependent only on
the material and geometry of the contact
structure and the quality of the vacuum. This
is not strictly true, however, because care
has to be taken to ensure the correct travel
and force characteristics of the operating
mechanism with respect to the interrupter.
The high dielectric strength of the vacuum
permits a small contact gap in the open
position, typically 1/4 to 3/4 inch (0.64 to
1.9 cm). As a result of this fact and the low
resistance of the metal vapor arc, the arc
energy is very low. Following arc
quenching, most of the metal vapor
condenses back onto the contacts and, by
doing so, results in contact restoration; thus,
contact erosion is extremely low.
Vacuum Circuit Breakers
Although the earliest attempts to make
circuit-interrupting devices using contacts in
a vacuum date back to the 1920s, it was not
until the 1960s that the technology reached
the point at which reliable vacuum
interrupters capable of acceptable short
circuit breaking current ratings could be
manufactured economically.
At the moment of contact separation through
which current is flowing, an arc discharge
occurs. With currents approximating 10 kA,
the arc burns diffusely on the whole contact
surface between the contacts until the next
current zero. Contact burn-off is low, as is
the specific thermal load of the contact
surface. For currents above 10 kA, the arc is
severely pinched through the pressure of its
own magnetic field. The concentrated
burning arc, occasioned by the high current
density, causes large amounts of contact
material to vaporize very rapidly. The
contact material, the shape of the contacts,
and the manufacturing methods for the
contacts play a significant role in the
ultimate performance of a vacuum
interrupter.
Figure 3-1 and Figure 3-2 provide an outline
and the internal construction of a modernday vacuum interrupter. Under working
conditions, when contact separation occurs,
arcing plasma that consists of metallic ions
develops. The formation of metallic vapors
sustains the arc. In proximity to the natural
power-frequency current zero, these vapors
condense rapidly onto metal screens. Thus, a
condition is created for ultra-fast recovery of
the dielectric strength.
At contact separation, where the arc is
supported by ionized gas in other
3-2
tank, compressed, and pumped into a
separate small reservoir. With the opening
of the circuit breaker, the compressed gas
was discharged through nozzles around the
contacts by a valve, resulting in the
extinguishing of the arc.
In distribution systems, the first technique
used in SF6 circuit breakers was that of a self
generated principle, referred to as the
“puffer” system. This was a step toward
simplifying the design with fewer moving
parts. Eliminating compressors, highpressure seals, and heating elements
obtained a much higher degree of reliability.
The design did not, however, overcome the
requirement for relatively long strokes and
high operating forces, resulting in the need
for powerful operating mechanisms with
large energy output requirements. A quantity
of SF6 gas is compressed with the
commencement of a circuit breaker opening
action prior to the separation of the arcing
contacts. When the contacts part, an arc is
drawn, and upon reaching current zero, the
flow of gas rapidly cools the remaining
plasma and sweeps away the arc products,
leading to an extremely rapid increase in the
dielectric strength of the gap, realizing a
successful clearance.
Figure 3-1
Schematic of Vacuum Interrupter
An evolving design process developed selfblast interrupters, which use the arc energy
to heat the gas and increase its pressure,
allowing the gas to expand. This means that
the pressure is raised by thermal means
rather than by mechanical energy. It is a
method that makes possible a lower-energy
operating mechanism than the puffer
interrupter. The ensuing arc extinguishing
process occurs in a similar manner to that of
the puffer interrupters.
Figure 3-2
Cut-Away View of Vacuum Interrupter
SF6 Circuit Breakers
Historically, the SF6 circuit breaker found its
application in transmission systems, where it
was shown to be a simpler alternative to the
air blast circuit breakers that were used in
the 1960s. The original circuit breakers were
of the two-pressure type, where the contacts
and arc control devices were immersed in a
metal tank full of SF6, serving the dual
purpose of an insulating and arcextinguishing medium. Essentially, the gas
used for arc extinction was drawn out of the
Figure 3-3 illustrates this thermal expansion
method. From the figure, the travel of the
gas can be observed. When heated by large
arc currents, the gas rapidly expands and
pressurizes the upper vessel chamber.
Pressure differential between the upper and
3-3
interrupting current. Figure 3-5 represents a
plot of arc quenching performance versus
interrupting current. From the curves, the
comparative performance for the different
interrupting techniques as a function of the
interrupting current can be readily observed.
lower vessel chambers creates a flow of gas,
facilitating the cooling of the arc and the
evacuation of the heat discharged by the arc.
Thus, the gas flow effectively extinguishes
the arc.
Figure 3-3
Thermal Expansion Method
Figure 3-5
Arc Quenching Performance vs. Interrupting
Current
The rotary arc quenching method, shown in
Figure 3-4, is based on the interrupting
current developing a rotating magnetic flux,
which extinguishes the arc. Because the arc
is rotated at approximately the speed of
sound and is always exposed to the SF6 gas,
extinction is ensured. The purpose of
rotating the arc is to bring the intense heat of
the arc column into contact with as much of
the gas as possible, thus raising its
temperature and, consequently, its pressure.
With the advent of the latest technologies
now available in SF6, the closing energy has
been reduced to levels comparable to
vacuum. This translates to fewer parts in the
circuit breaker mechanism. In fact, the
difference is so minimal that one European
manufacturer of both SF6 and vacuum circuit
breakers actually utilize the same
mechanism in each case.
Loss of Interruption Medium
It should
Both sealed vacuum and SF6
be noted interrupters are susceptible
that SF6 to the possible loss of their
circuit
interrupting medium. Most
breakers
users
have
shown
a
can
be willingness to accept this
monitored susceptibility.
in service,
while no
practical monitoring for commercial
purposes is currently available for vacuum
circuit breakers.
Figure 3-4
Rotary Arc Quenching Method
Sophisticated methods akin to those used in
a laboratory are available, but economics
make their use prohibitive.
By adopting the rotary arc and the thermal
expansion method, a proportionate arcquenching force is generated that is
equivalent to the magnitude of the
3-4
For load currents in an ungrounded
system, the vacuum circuit breaker
carried the current and interrupted it
satisfactorily. In a 600 A grounded
system, the phase in which the
vacuum
interrupter
was
at
atmospheric pressure continued to
arc for 15 seconds. All arcing was
confined
within
the
vacuum
interrupter, although the envelope
cracked. The SF6 interrupter at
atmospheric pressure switched the
load current in both grounded and
ungrounded circuits. For 10 kA and
at 25 kA the phase that had the
vacuum interrupter at atmospheric
pressure again continued to arc for
30 cycles until the backup circuit
breaker interrupted the circuit. All
arcing was confined inside the
interrupter’s envelope. In these
experiments, the envelope showed
no evidence of rupture. In fact, the
only evidence of thermal stress to the
ceramic
envelope
was
some
cracking. The SF6 puffer interrupter
interrupted the 10 kA current, but at
25 kA the SF6 did not interrupt the
circuit and the continued burning of
the arc caused the puffer to explode
(Storms, 1992).
For the SF6 circuit breaker, reliable pressure
switches can be furnished within the circuit
breaker’s poles, and these switches are
wired into a two-stage alarm system to
register any loss of the gas, facilitating
action prior to a dangerous condition
(Swindler, 1989). It should also be noted
that, even with a total loss of SF6 gas, a 30kV withstand level is maintained, and load
currents can still be safely interrupted on a
one-time basis (Swindler, 1993).
The mechanical connection into the gas
bottle and the pressure switch itself are two
more
potential
leak
paths.
The
manufacturer’s history of gas bottle leakage
should be a factor in choosing the claim of a
higher reliability in gas integrity.
SF6 gas is five
SF6 gas is five times
times heavier than heavier than air.
air, the possibility
of some air replacing the SF6, even after the
pressure has equalized, cannot be ruled out.
However, a large proportion of air is needed
to significantly reduce the breakdown
strength, and designers account for this
remote possibility by designing the open gap
contact system to withstand the service
voltage at atmospheric pressure. This should
be contrasted with a vacuum interrupter
suffering a leak, where the voltage strength
of the gap falls to a minimum level of a few
hundred volts in the glow discharge region
of 0.1–10 torr (13.3–133.2 pascals),
recovering to around 30 kV/cm at
atmospheric pressure. When the vacuum is
leaking and the voltage drops, failure is
possible (Blower, 1986).
It should be noted that contemporary designs
of both rotary arc and puffer SF6 circuit
breakers
employ
an
over-pressure
diaphragm at the bottom of each interrupter.
In the event that the interrupter fails to
interrupt due to loss of gas, the diaphragm
ruptures, directing the pressurized gas
downward.
The consequences of attempted interruption
of fault currents by a vacuum interrupter
(with loss of vacuum) and by an SF6 circuit
breaker (with loss of SF6) have not been
adequately addressed in the available
technical literature. The following statement
appears in a 1992 Doble Conference paper
by a vacuum CB manufacturer:
In early designs of SF6, problems were
experienced with leaking seals. This is no
longer the case with better contemporary
varieties that use EP-type rubber O-rings.
These O-rings have a longer life than the
circuit breaker itself (see Figure 3-6). For all
production runs, the leakage rate of gas is
tested by highly sensitive detecting
3-5
equipment that can detect leakage rate
values up to 10-9cc/sec.
Circuit Breakers
Circuit breakers in the US are rated by the
American National Standards Institute
(ANSI) and the Institute of Electrical and
Electronics Engineers (IEEE). Circuit
breakers are classified by location, indoor or
outdoor, and by voltage rating. The ratings
are summarized in Table 2-1. Indoor circuit
breaker ratings are shown in Table 2-2.
Outdoor distribution circuit breakers are
summarized in Table 2-3. Outdoor high
voltage circuit breakers are summarized in
Table 2-4. The maximum interrupting
ratings for these various types are as
follows:
•
•
Figure 3-6
Relation Between O-Ring Life and
Temperature
•
63 kA for indoor distribution, 15 kV
class
40 kA for outdoor distribution, 38
kV class
80 kA for outdoor transmission, 145
kV class.
Table 3-1
Summary of ANSI Breaker Ratings Showing
Maximum for Each Class
Interrupting Ratings of Switching Devices
Exceeding the interrupting rating will result
in continued arcing, fire, explosions and
failure to interrupt the fault. Danger to any
nearby personnel is likely. Damage to
substation structure, switchgear, buses, and
other equipment is also likely. Backup
equipment must interrupt the fault.
For a variety of reasons, users might
discover that available fault currents on their
power systems exceed the interrupting rating
of their existing circuit breakers. The
increasing sophistication of tools for
determining short circuit (fault) currents, the
addition of transformer capacity necessary
to support plant life extensions, the addition
of motor loads as back-electromotive force
(EMF) fault contributors, and the
reconfiguring of systems all contribute
toward this situation. Safety considerations
make the need to address the situation
obvious.
Table 3-2
Typical Indoor Distribution Circuit Breaker
Ratings (ANSI. 2000a, Table 1)
Rated
Maximum
Voltage,
kV, rms
3-6
Rated
Continuous
Current,
A, rms
Rated
ShortCircuit
Current,
Rated
Closing
and
Latching
kA, rms
Current,
kA, rms
1200, 2000
31.5
82
1200, 2000
40
104
1200, 2000,
3000
50
130
8.25
1200, 2000,
3000
40
104
15
1200, 2000
20
52
1200, 2000
25
65
1200, 2000
31.5
82
1200, 2000,
3000
40
104
1200, 2000,
3000
50
130
1200, 2000,
3000
63
164
1200
16
1200, 2000
4.76
27
38
Table 3-3
Typical Outdoor Circuit Breaker Ratings up
to 72.5 kV (ANSI. 2000a, Table 2)
Rated
Maximum
Voltage,
kV, rms
Rated
Continuous
Current,
A, rms
Rated
ShortCircuit
Current,
kA, rms
Rated
Closing
and
Latching
Current,
kA, rms
15.5
600, 1200
12.5
33
1200, 2000
20
52
1200, 2000
25
65
1200, 2000,
3000
40
104
1200, 2000
12.5
33
1200, 2000
25
65
1200, 2000
16
42
42
1200, 2000
20
52
25
65
1200, 2000
25
65
1200
16
42
1200, 2000
31.5
82
1200, 2000
25
65
40
104
1200, 2000,
3000
31.5
82
1200, 2000,
3000
1200, 2000
20
52
1200, 2000,
3000
40
1200, 2000
31.5
82
1200, 2000,
3000
40
104
1200, 2000
20
52
1200, 2000
31.5
82
1200, 2000,
3000
40
104
25.8
38.0
48.3
104
72.5
3-7
The ability of a circuit breaker to
successfully clear a fault is dependent on the
magnitude of fault current. The breaker
failure probability curve for a breaker which
is fully operational (all elements function
properly) is shown in Figure 3-7. It can be
seen that breaker in such a state has a high
probability of successful operation while
interrupting currents below the threshold
current (I0). Threshold current may represent
the expected fault current level and should
be significantly below the breaker rated
current (IN). If the fault current is beyond
maximum breaker capability (I1), there is no
chance of successful operation. Thus, it can
be seen that higher the fault current above
threshold value, greater is the probability
that it might fail to successfully interrupt the
fault.
Table 3-4
Typical Outdoor Circuit Breaker Ratings 123
kV and Above (ANSI. 2000a, Table 3)
Rated
Maximum
Voltage,
kV, rms
Rated
Continuous
Current,
A, rms
Rated
ShortCircuit
Current,
kA, rms
Rated
Closing
and
Latching
Current,
kA, rms
123
1200, 2000
31.5
82
1600, 2000,
3000
40
104
2000, 3000
63
164
1200, 2000
31.5
82
1600, 2000,
3000
40
104
2000, 3000
63
164
2000, 3000
80
208
1600, 2000
31.5
82
2000, 3000
40
104
2000, 3000
50
130
2000, 3000
63
164
1600, 2000,
3000
31.5
82
2000, 3000
40
104
2000, 3000
50
130
2000, 3000
63
164
2000, 3000
40
104
2000, 3000
50
130
2000, 3000
63
164
2000, 3000
40
104
3000, 4000
50
130
3000, 4000
63
164
2000, 3000
40
104
3000, 4000
50
130
3000, 4000
63
164
145
170
245
362
550
800
Figure 3-7
Breaker failure probability vs Current
Magnitude (PSERC 2006)
Fuses
There are two major types of fuses: Currentlimiting and expulsion type (Figure 2-7).
Fuse ratings are summarized in Table 2-5.
While current-limiting fuses are enclosed in
a sealed cylinder, and are usually contained
in metal enclosed switchgear, limiting the
danger even if their short-circuit rating is
exceeded, expulsion fuse links are mounted
in open tubes on utility poles. These present
several hazards when subjected to excessive
short circuit currents:
3-8
•
Exceeding the interrupting rating will
result in continued arcing, fire,
explosions and failure to interrupt the
fault.
•
Danger to public from falling arcing
products if fuse is on a pole along a
street.
•
Damage to disconnecting switch, poles,
enclosures, and other equipment is
likely.
•
Table 3-5
Preferred Ratings for Fuses (ANSI 2000b)
Backup equipment must interrupt fault.
CurrentCurrent-limiting
fuses
limiting
generally
have
higher
short
fuses are
circuit
ratings
than
available
expulsion fuses of the same
which fit in
size.
cutouts,
replacing
expulsion fuse links.
Upgrading of
expulsion fuses to current-limiting fuses and
in general replacing fuses with units having
higher interrupting ratings is the preferred
solution to increasing available fault current
Rated
Symmetrical
Interrupting
Current,
kA, rms
2.54-2.8
31.5-63
5.08-5.5
31.5-50
8.3
4.0-50
15-17.2
4.0-80
23-27
4.0-50
38
5.0-40
48.3
3.15-31.5
72.5
2.5-25
121
1.25-16
145
1.25-12.5
169
2.5-12.5
Next-generation Solid-state Breakers
There are growing instances in utility
distribution and transmission systems
wherein the fault current levels are
exceeding the interrupting capability of
existing substation circuit breakers. This
increase in fault current level either requires
the replacement of large number of
substation breakers or the development of
some means to limit the fault current. Also,
many mechanical circuit breakers are
operating much more than originally
intended in applications such as capacitor
switching. This continual use of mechanical
breakers requires intensive maintenance to
be performed or periodic replacement of the
whole breaker. Environmental problems are
also on the horizon with the use of both SF6
gas and oil within mechanical breakers,
which may pose long term problems for
many utilities.
levels.
(a)
Rated Maximum
Voltage, kV,
rms
(b)
Figure 3-8
(a) Current Limiting Fuses (b) Pole Mounted
Fuse Cutout (Cooper Power Systems)
3-9
In a recent study [EPRI 2004], EPRI
conducted a comprehensive survey to assess
utility needs and perceptivities with respect
to power electronics based solid state
switchgear that can limit fault currents.
Some of the key findings that are relevant to
the outcome of the economic assessment
presented in this chapter include:
•
•
•
Generation 1: 15KV Class Distribution
Switchgear Development (breaker, a current
limiter, DG isolation switch, capacitor
switch, tie breaker, and transfer switch)
Detailed design analysis and
functionality verification of family of "All
Solid-State" and "Hybrid" Switchgear
Topologies (Voltage Ratings: 15KV, 35KV,
138KV & Current Ratings: 600A, 1200A,
3000A) which can be used as a breaker, a
current limiter, DG isolation switch,
capacitor switch, tie breaker, and transfer
switch (2006)
Phase
Survey results indicate that up to
20% of utilities expecting to replace
5 to 10% of their circuit breakers in
the next 10 years would use a fault
current limiting devices that are
priced at 1 to 5 times a circuit
breaker
I:
Phase II: Bench model development and
testing of 15KV 600A distribution
switchgear using hybrid technology which
can be used as a breaker, a current limiter,
DG isolation switch, capacitor switch, tie
breaker, and transfer switch (2006-2007)
Utilities having a greater expectation
for circuit breaker replacement are
even more likely to use a solid-state
fault current limiter – the percentage
increases to 30% of utilities when the
range of circuit breaker replacement
need expands to 5 to 30%
Phase III: Field
prototype development,
deployment, and testing of 15KV 600A
distribution switchgear using hybrid
technology which can be used as a breaker,
a current limiter, DG isolation switch,
capacitor switch, tie breaker, and transfer
switch (2008-2009)
Cases where breakers with the
required ratings are not available, or
where excessive fault current levels
carry more than only cost of a
breaker upgrade alone, 50% of the
utilities value a solid-state fault
current limiter at 2-5 times the cost
of a breaker
Phase IV: Field testing and debugging of
field prototype (2008-2009)
Phase V: Finalization of design, packaging
and preparation of product family
specification
and
subsequent
commercialization (2010)
The report (EPRI 2005) describes the
findings of the feasibility assessment
research that has been done for developing a
next generation of solid-state breakers. The
hybrid switchgear design has been proposed
that would meet the requirements of rapid
fault clearing, instantaneous fault isolation,
fast current limiting, soft switching
capabilities and rapid load transfer. The
proposed approach is based on innovative,
multi-functional, modular, and hybrid design
of power electronics based switchgear.
Distribution and transmission switchgear
development effort has been broken down
into two distinct product families; namely:
Generation
2:
35/138KV
Class
Distribution/Transmission
Switchgear
Development (breaker, a current limiter, DG
isolation switch, capacitor switch, tie
breaker, and transfer switch)
Phase I: Bench model development and
testing of 35KV 600A distribution
switchgear using either "all solid-state" OR
"hybrid" technology which can be used as a
breaker, a current limiter, DG isolation
switch, capacitor switch, tie breaker, and
transfer switch (2011)
3-10
Phase II: Bench model development and
Identifying the Breakers with Excessive
Fault Currents
Due to the addition It may be feasible
of new generation to avoid some
topologies
that
and changes in
result in excessive
network
fault current duties
configurations, the
on breakers.
fault current levels
in the system may
exceed the interruption capability of some
circuit breakers. One way to deal with the
problem is to replace the affected breakers
(discussed in next section). In case
replacement of breakers is not feasible, it is
necessary to avoid the operating conditions
that result in excessive fault current levels.
testing of 138KV 600A distribution
switchgear using either "all solid-state" OR
"hybrid" technology which can be used as a
breaker, a current limiter, DG isolation
switch, capacitor switch, tie breaker, and
transfer switch (2012)
Phase III: Field prototype development,
deployment, field testing and debugging
(2013)
Phase IV: Finalization of design, packaging
and preparation of product family
specification
and
subsequent
commercialization (2014)
Figure 3-9 shows the circuit configuration of
the 15kV, 600A HV-IGBT based Universal
Hybrid Switch that requires 4 solid-state
switches in series, Sss1 – Sss4.
Sm
V1
vge1
Ts1
IGBT
MOV
Is
Snubber
TVS
Iline
The procedure to check for the possibility of
the fault current exceeding the interruption
capability of any breaker due to change in
system configuration is explained using
IEEE 14 bus system (Figure 3-10) in
(PSERC 2006). The standard test system is
modified by introducing a small generator
(12% Z, 10MVA) at bus 5. It is found that it
resulted in increase in short circuit currents
at each bus with percent increase being the
maximum at Bus 5 from 13.2 pu to 14.2 pu
(7.56%). Consequently, the maximum fault
current seen by breakers at bus 5 is found to
exceed their interruption ratings (14.0 pu).
Sss1
Ts4
IGBT
MOV
V4
vge4
Snubber
TVS
Sensor
Conditioning,
Control, &
Gate Drives
Sss4
Figure 3-9
Circuit Configuration of the 15kV, 600A HVIGBT Based Universal Hybrid Switch.
3-11
Ii5 = Fault current input into bus 5 from
bus i
Figure 3-10
IEEE 14-bus test system with additional
generator at bus 5 (PSERC 2006)
Bi = ith breaker
The next step is to find out which breaker
sees the maximum fault current for different
substation topologies due to difference in
fault current paths. The flowchart that has
been used to identify the breaker seeing the
maximum fault current for a given topology
is shown in Figure 3-11. It involves
computing fault current through each
breaker for a particular fault location by
making use of the conductance matrix for
the given topology. It is important to
consider all the possible fault locations in
the topology.
th
Si = i bus section
Figure 3-12
Example Substation Topology (PSERC 2006)
Figure 3-11
Flowchart for finding the Breaker seeing the
Maximum Fault Current (PSERC 2006)
An example topology (Breaker B9 in “Open”
state and fault in section S7) for substation at
bus 5 assuming the breaker-and-a-half
connection is shown in Figure 3-12 and the
corresponding equivalent circuit is shown in
Figure 3-13. In the figures,
Figure 3-13
Equivalent Circuit for Topology in Figure
3-12 (PSERC 2006)
For this topology, the breaker that sees the
maximum fault current was identified using
the procedure explained in Figure 3-11 . The
same procedure was repeated for some other
topologies and the results are shown in
Table 3-6. The system will be safe if the
Ig = Fault current contribution of
generator
Iload = Equivalent load current at bus 5
3-12
identified
avoided.
problem
topologies
can
be
Figure 3-14
Flowchart for Online Assessment of Fault
Current (PSERC 2006)
Table 3-6
Breakers seeing Maximum Fault Current
(ANSI 2000b) (PSERC 2006)
Case Studies
Sub
Topology
Fault
Location
Breaker
with the
largest
fault
current
Largest
Fault
Current
(pu)
Exceed
Rating ?
All closed
S6
B8
7.7012
No
B3 open
S2
B2
12.1315
No
B6 open
S4
B5
13.2164
No
B9 open
S6
B8
14.1717
Yes
B2,B3
open
S7
B1
14.2
Yes
These case studies describe the replacement
of air-magnetic circuit breakers whose
interrupting ratings have been exceeded with
new SF6 or vacuum interrupters. Both of
these cases are taken from nuclear power
plants.
Diablo Canyon
At Diablo Canyon, (EPRI. 2003) the prime
motivation for upgrading air-magnetic
circuit breakers to SF6 was an increase in
the fault level and the elimination of a
potentially hazardous design issue rather
than the desire to save maintenance and
other costs. Beginning in March 1994, a
dedicated user team set out to evaluate the
various options available. The team had
extensive experience in preparing the type of
quality
assurance
(QA)/dedication
requirements that were to be imposed on the
selected vendor for this project. The project
consisted of 110 units.
In practical operation, online assessment of
fault current will be required to inform the
operator of a condition in which the fault
current would exceed the interruption
capability of any breakers. Flowchart of the
software that can perform such a function is
shown in Figure 3-14.
Maintaining the existing switchgear to
provide a 350-MVA circuit breaker within
the same footprint as the 250-MVA design
necessitated a compact circuit breaker
design. At the time, only one circuit breaker
met this criterion. Others did not represent
completed, tested designs. The fact that this
element was SF6 provided additional
comfort because the user preferred
monitoring the state of the interrupting
medium and desired a surge-free
performance. These criteria in conjunction
with the QA requirements stipulated by
Pacific Gas and Electric were the defining
characteristics in selecting the vendor for
this particular project. The Diablo project is
a useful model for the nuclear industry for
converting circuit breakers due to the
combination of high seismic levels, the
3-13
qualification/dedication complexities, and
the technical challenges that are clearly
defined.
maintenance cost savings over the circuit
breaker’s installed life.
Dresden and Quad Cities
The significant lessons learned from this
project fall into two distinct areas that are
unique to the nuclear environment. The first
area concerns how to administer an adequate
nuclear QA program in the dedication and
building of circuit breaker conversions
without losing sight of the primary mission
of building reliable conversions in a timely
and efficient manner. Many of the issues
associated with the manufacturing process
arose because of the intrusive nature of
implementing the QA program. Careful
consideration of QA needs was balanced
with the requirements of a production
facility that is geared to the non-nuclear
aspects of building conversions.
There were several factors involved in the
decision-making process at the Dresden and
Quad Cities plants, (EPRI. 2003). These
factors extended from the problems
associated with obsolescence and lack of
spare parts; the required addition of
cubicles; the need to upgrade interruption
levels; the objective of minimizing outage
duration
during
modifications;
and
recognition of the state of the 20-year-old
motors, transformers, and cables.
Initially,
wholesale
removal
and
replacement of the switchgear, with the
inclusion of additional cubicles, was
considered. The Original Equipment
Manufacturer (OEM) was favored in order
to maintain continuity of products and
maintenance services. The preferred
technology was air magnetic because of its
established acceptable overvoltage response,
but this medium was no longer available
from the OEM.
The second area concerns the interface of
the new design with the existing cubicles
and the flexibility to undertake essential
changes to cubicle interlock interfaces while
being cognizant of nuclear constraints on the
design of the conversion.
A further example of flexibility relates to a
deviation in testing procedures. There is no
economical method to field test the pressure
switches contained within the poles of an
SF6 circuit breaker. After numerous
meetings and detailed design reviews of the
product and its associated manufacturing
methods, acceptance was sanctioned as the
lesser of two evils—in other words,
accepting the pressure switch factory
reliability test results and the track record of
the product in the field as opposed to the
incorporation of a vacuum element, which
cannot be monitored
The costs for the subject proposal of full
switchgear replacement were judged
excessive given the extensive rigging effort,
necessary extension in outage time, and the
risk of damage to the old power and control
cables (Shah, et al. 1988., Dinkel et all.
1986., Fish. 1994, and Heath et al. 1987.).
This can be readily appreciated in
recognition of the fact that the total number
of existing cubicles was 100, and 12
additional units had to be accounted for.
An alternative approach, based on the
concept of a 250 MVA to 350 MVA
conversion circuit breaker that would fit into
the existing 26-inch-wide cubicle, seemed
appealing as it became apparent that it was
economically attractive and would eliminate
the considerable risks pertaining to full
replacement. At the time, IEEE Standard
To the users, this project is a success. A
replacement circuit breaker was effectively
integrated into the existing plant power
system, significantly increasing fault
protection and ensuring a considerable
3-14
C37.59 was in draft form and could serve
well as a guide for the planned scope of
work.
Voltage Expulsion and CurrentLimiting Type Power Class Fuses and
Fuse Disconnecting Switches, New
York, NY: IEEE.
The final selection of vacuum or SF6 circuit
breaker technology was made based on cost,
after factoring in the associated costs for
surge arresters, spare parts, training, and
vendor services. It was originally assumed
that the existing 250 MVA bus rating was
insufficient to withstand the 350 MVA
momentary ratings. Testing in a high power
laboratory proved otherwise, and the
approximately 20-year-old cubicle with its
20-year-old bus and bracing successfully
withstood the momentary test stresses. This
certification allowed the project to abandon
the time-consuming and expensive changeout of the existing bus bracing.
3. Blower, R. W., 1986. Distribution
Switchgear, William Collins and Co.,
Ltd., London
4. Dinkel, Darell G., Walter G. Watts,
and James R. Langlois, 1986, “13.8
kV Switchgear Uprate,” presented at
the Tappi Engineering Conference.
5. EPRI. 2003.
Considerations for
Conversion or Replacement of
Medium-Voltage
Air-Magnetic
Circuit Breakers Using Vacuum or
SF6 Technology: Revision to TR106761, EPRI, Palo Alto, CA: 2003.
1007912.
Summary and Recommendations
6. EPRI 2004. Fault Current Limiters Utility Needs and Perspectives (EPRI
Report 1008696 & 1008694, 2004)
The ability of a circuit breaker to
successfully clear a fault is dependent on the
magnitude of the fault current. There are
limits on the maximum current that can be
interrupted safely by circuit breakers. If
feasible, the problem topologies that result
in excessive fault duty on substation
breakers should be avoided. Other steps
such as fault mitigation techniques may be
used to reduce the fault duty on the breakers.
In case of prospective fault currents still
being excessive, the circuit breakers will
need to be replaced with one having higher
ratings.
7. EPRI 2005. EPRI family of Multifunctional low cost Solid State
Switchgear: Requirements Definition
Phase, EPRI, Palo Alto, CA:2005.
8. Fish, Michael W., 1994, “When You
Have to Retrofit 15 kV Switchgear,”
presented at the IEEE Pulp and Paper
Technical Conference.
9. Genutis, Don A., 1992. “Problems
with Medium Voltage Air Magnetic
Circuit Breakers,” NETA World,
Spring.
References
10. Heath, W., T .S. Freeman, and F. R.
Cochran, 1987, “Replacement of Air
Magnetic Breakers with Vacuum
Breakers in Medium Voltage
Switchgear Assemblies,” presented at
the Electrical Systems & Equipment
Committee
Meeting
No.
58,
Engineering and Operations Division,
Electric Council of New England.
1. ANSI. 2000a. Standard C37.06-2000
AC High-Voltage Circuit Breakers
Rated on a Symmetrical Current
Basis— Preferred Ratings and
Related Required Capabilities, New
York, NY: IEEE.
2. ANSI. 2000b. Standard C37.46-2000
American National Standard for High
3-15
11. Shah, K. R., E. R. Detjen, and D. H.
Epley, 1988, “Impact of Electric
Utility and Customer Modifications
on Existing Medium Voltage Circuit
Breakers,” presented at the IEEE
Industry Applications Society Annual
Meeting
12. Storms, Alan D., 1992. “Medium
Voltage Circuit Breaker Retrofit
Technology: The Current Control for
Costly Outages,” presented at the
Doble Conference
13. Swindler, David L., 1989. “Medium
Voltage Application of SF6 and
Vacuum Circuit Breakers,” presented
at the Tappi Engineering Conference
14. Swindler,
David
L.,
1993.
“Applications of SF6 and Vacuum
Medium Voltage Circuit Breakers,”
presented at the IEEE PCIC
Conference, Denver, CO
3-16
4
EFFECT OF HIGH
FAULT CURRENTS ON
PROTECTION AND
METERING
In this chapter, the effects of high fault
currents on protection and metering
equipment will be reviewed.
The
capabilities and limitations of existing
short circuit protection devices will be
included in the literature search. High
fault currents are well known to cause
saturation of iron core Current
Transformers (CTs). This can adversely
affect the performance of system
protection devices. High fault currents
can also exceed the range of operation of
current operated protective devices and
produce high voltages in the CT
secondary circuits.
Figure 4-1
Current Transformer Equivalent Circuit
CTs are intended to deliver a secondary
current that is directly proportional to the
primary current with as little distortion as
possible. In most cases, the secondary output
current is usually reduced to less than 5
amperes. Although there are CTs with 1
ampere or 10-ampere secondaries, the most
common rating in the United States is 5
amperes.
CTs are rated for a certain turns-ratio of
operation. For example, a CT with a turnsratio of 500:5 reduces 500 amperes on the
primary to 5 amperes on the secondary. A
properly designed CT circuit yields a
secondary current of 5 amperes or less at
rated primary current. Although the CT
secondary and the relay are not intended for
continuous operation at higher than 5
amperes, they are designed to withstand
greater values of current for short periods.
For example, short-circuit currents may be
20 or more times the normal current in a
power system.
Increase in the fault currents in power
systems has an impact on the behavior of
existing protection schemes. For example,
protective relays may act differently to faults
when the new distributed generation also
starts contributing to the fault currents
(PSERC 2005). In the response to higher
fault currents, protection system needs to be
regularly updated in terms of settings of
relays, fuses and circuit breakers.
Current Transformer Saturation
During normal operation, the CT secondary
winding induces a magnetic flux that
opposes and nearly cancels the primary
induced flux. As a result, the flux density is
very low and the resulting voltage at the
secondary terminals is also very low.
Relays, meters, or other connected devices
Current transformers (CTs) are connected in
series with the circuit whose current is to be
measured as shown in Figure 3-1, from
(EPRI. 2004).
4-1
are constructed with only a few turns of
relatively large wire—this low impedance
effectively functions as a short circuit across
the CT secondary. The secondary voltage of
a CT remains at a low value as long as the
secondary circuit remains closed. An open
circuit on the secondary side of a CT that
still has current flow on the primary side can
result in a dangerously high secondary
voltage. Opening the secondary removes the
opposing secondary flux, thus allowing the
primary flux to generate a very high voltage
at the secondary terminals. Electrical arcing
caused by an open-circuited CT can injure
personnel and damage equipment. When the
primary side is carrying any current, great
care must be taken to ensure that the
secondary circuit remains closed at all times.
When saturated, most of the primary current
maintains the core flux, and the shape of
both the exciting and secondary currents
departs from the normal sine wave. The
secondary voltage and current then collapse
to zero, where they remain, until the next
primary current zero is reached. The process
is repeated each half-cycle and results in a
distorted secondary waveform.
For this reason, a CT must be carefully sized
so that it will perform properly for the
maximum expected fault current. Low-ratio
CTs (for example, 50:5 or 75:5) are
particularly susceptible to saturation during
fault conditions.
The following problems may exist if a CT is
allowed to saturate due to high fault
currents:
The ability of a CT to produce a secondary
current proportional to its primary current is
limited by the highest secondary voltage that
it can produce without saturation. Beyond a
certain level of excitation (actual values are
readily available from the manufacturer), the
CT is said to enter saturation. The
phenomenon is shown for an example CT in
Figure 4-2 (Kojovic 2002). It is shown that
the ratio error increases as the operating
point gets further deep into the non-linear
region.
•
•
•
Figure 4-2
600/5 A, C 100 CT Saturation Characteristics
4-2
False tripping - Differential relays used
for transformer protection may respond
to a through fault condition. Numerical
relays, however, pose far less burden
(0.5 VA) on current transformers.
Numerical relays are capable of
detecting CT saturation and blocking the
relay from tripping, minimizing the
effect of false tripping.
Delayed tripping - A distorted secondary
reproduction of the primary current can
delay relay time response. This delay in
tripping may result in deenergizing a
larger portion of the system due to loss
of relay coordination caused by the CT
saturation.
Failure to trip - Failure to trip may occur
if the CT secondary current is very low
or extremely distorted. Backup relays
must then respond to clear the fault.
Digital relays, however, have the added
advantage that they can detect these
conditions and still deliver a trip because
the relay is not dependent on the power
that must be supplied by the CT to trip.
This is usually slightly higher than the
"knee-point voltage," which is defined in the
standard as the point on the saturation curve
where a tangent drawn to the curve has an
angle of 45° to the horizontal axis for a nongapped core and 30° for a gapped core. See
the illustration of knee-point concept (Figure
3-2).
Saturation of Low Ratio CTs
High
levels of High levels of fault current,
especially when DC offset
fault
is present, cause the
current,
secondary current of a CT
especially
when DC to be significantly distorted
diminished
in
offset
is and
magnitude,
even
with
a
very
present,
cause the small burden.
secondary
current of a CT to be significantly distorted
and diminished in magnitude, even with a
very small burden.
Secondary Voltage (V)
100
This is most significant with low-ratio CTs.
CT saturation may cause overcurrent relays
to misoperate or fail to operate, resulting in
a failure of the protection system.
The
Saturation of a CT from
performance
of protective a high fault current can
prevent a protective
relaying
systems in the relay from operating
presence of properly.
CT saturation
has been discussed in, among others, IEEE
standards (ANSI/IEEE. 1996), textbooks on
relaying (Blackburn. 1998 and Elmore.
1994), and IEEE committee reports (Power
Systems Relaying Committee. 1976. and
Linders et. al. 1995.).
10
1
0.1
0.01
0.1
1
10
Exciting Current (A)
Figure 4-3
The "Knee-Point" of a CT Saturation Curve is
the Point Where a Tangent to the Curve
Forms a 45° Angle With the Horizontal Axis
Section 4.5.2(a) of ANSI/IEEE Std.
C37.110-1996 suggests that the effective CT
ratio error will be significant if the
calculated secondary voltage that the CT
must support exceeds the saturation voltage,
VX. That is,
AC Saturation
AC Saturation is a gradual process, where as
the rms value of the current increases, the
ratio accuracy of the CT decreases. AC
Saturation begins to affect protective relay
performance if the rms excitation voltage
begins to exceed the "saturation voltage,"
where the rate of increase of the excitation
current with respect to excitation voltage
greatly increases.
According to the
definition given in the standard, the
"saturation voltage," VX, is the point of
intersection of lines extended from the
straight portions of the saturation curve.
V X < VS , where
VS = I S × Z S
=
Eq. 4-1
IP
× ZS
N
and IS is the primary current IP divided by
the turns ratio, N, and ZS is the total
secondary burden.
After an example
calculation of this type, where VS approaches
4-3
VX, one author states: “Although this is near
the knee of the saturation curve, the small
excitation current does not significantly
decrease the fault current to the relays.”
(Blackburn. 1998, p. 146)
present, it is no longer possible to think in
terms of conventional sinusoidal response.
Practically, an accurate measurement of
current is only important in the case of time
overcurrent relays.
With instantaneous
relays, it is only important to know that the
available current will be greater than the
setting of the relay. It is not important to
know by how much the available secondary
current will exceed the instantaneous relay
setting or whether the waveform of that
current will be a respectable sinusoid.
Another commonly used criterion is from
section 5.10 of the standard: “A rule of
thumb frequently used in relaying to
minimize the CT saturation effects is to
select a CT with a C voltage rating at least
twice that required for the maximum steadystate symmetrical fault current.”
It is
furthermore stated in (Linders et. al. 1995.)
that: “One basic rule-of-thumb has applied
in the application of CT’s, namely: The knee
point voltage of the CT as defined by the CT
excitation curve should not be less than
twice the voltage required to drive the
maximum secondary symmetrical current
through the combined burden of the relay,
connecting wiring and CT.” While the “C”
rating is often assumed to be approximately
equal to the knee-point voltage, in fact, “the
knee-point voltage may be 50% to 75% of
the standard accuracy class voltage rating of
the CT (e.g., C400).” (ANSI/IEEE. 1996,
section 4.5.2).
Therefore, the pertinent question is: Given
the performance characteristics of the CT
and the associated CT burden, will the CT
be "reasonably linear" for all fault
magnitudes for which the time overcurrent
relay is expected to operate? Based on this
perspective, the following conditions should
be checked:
For feeder relays having both time and
instantaneous elements, will the calculated
secondary voltage be less than the CT
saturation voltage, VX, for the maximum
fault at which the time overcurrent relay is
expected to operate, namely, the current
level at which the instantaneous element is
calibrated to pick up?
It is important to carefully consider the
implications of AC saturation before
applying these criteria. As stated above, the
effect of AC saturation is to cause an error
in the effective ratio of the CT such that the
secondary current available to relays and
other measuring devices will be less than
what would be expected knowing the
primary fault current and the nominal CT
ratio. The fact that the calculated voltage
exceeds the saturation voltage, VX, does not
mean that the CT collapses entirely - just
that the ratio error increases significantly.
For main and tie relays with no
instantaneous elements, will the calculated
secondary voltage be less than the CT
saturation voltage, VX, for the maximum
fault at which the time overcurrent relay is
expected to operate, namely, the expected
(calculated) maximum current through the
circuit breaker associated with the CT in
question?
DC Saturation
There is more to saturation that an increase
in the effective ratio error of the CT. The
presence of significant saturation also causes
the waveform of the secondary current to
depart from the normal sinusoidal pattern.
Hence, when significant saturation is
The standard suggests two criteria regarding
DC saturation. The first (section 4.5.2) is
that the effective CT ratio error will be
significant when the following condition is
met:
4-4
⎛
X R + RB ⎞
⎟
I S × Z S ⎜⎜1 + × S
R
Z S ⎟⎠
⎝
VX <
1 − per unit remanence
will also dissipate with a few seconds of
loading, if 60% of the saturation voltage is
exceeded, but it can be a concern if the
breaker subsequently recloses into a fault
before flux dissipation can occur.
Eq. 4-2
Remanent flux may either aid or oppose the
magnetization imposed by DC transients.
Practically, remanent flux is not a real
concern in industrial applications where
there is no automatic reclosing.
where RS+RB is the resistive component of
burden, X and R are the primary side system
reactance and resistance up to the fault.
DC
saturation
does
not
occur
instantaneously, but rather builds up with
time. A CT is often able to replicate the
offset in primary current accurately for few
cycles before the core begins to enter the
area of saturation. Hence, the second
criterion to be considered in evaluating DC
saturation is the time to saturate (in
fundamental frequency cycles):
Alternatively, one author (Elmore. 1994. p.
80) suggests that if
V X ≥ I S (R S + R B )
X
R
Eq. 4-3
DC saturation will not occur. These two
equations may be simplified as representing
either VS ⎛⎜1 +
⎝
X⎞
⎟
R⎠
or V S
X
R
, where Vs is the
TS = −
CT secondary voltage. These equations may
be seen in some CT application calculations,
and are valid as long as there is not
significant inductance in the CT secondary
circuit or significant remanence in the CT
itself. Dropping the factor of 1 in equation
(3) simply means that X/R is assumed to be
high.
⎛ K −1 ⎞
X
⎟
ln⎜⎜1 − S
ωR ⎝
X R ⎟⎠
Eq. 4-4
The saturation factor KS, is defined as:
KS =
VX
I S (RS + R B )
Eq. 4-5
This formula does not include the effect of
remanence, which will decrease the time-tosaturation. It has also been shown that
greater the degree of DC offset, the sooner
the core will reach the onset of saturation.
For the same degree of offset, the magnitude
of offset current is proportional to the
magnitude of fundamental component.
Therefore, the greater the fault current
magnitude, the lesser is the time-tosaturation.
Remanence is the tendency of the iron core
of a CT to retain magnetic flux based on
prior history. Remanence flux levels of up
to 80% of saturation level have been
observed (NFPA. 2005. par. 4.6.1). The
worst case for remanence comes about when
a DC continuity test is used to verify CT
circuit continuity (ANSI/IEEE. 1996. par.
4.6) (Seveik and DoCarmo. 2000), but this
remanent flux will dissipate if the CT is
demagnetized following the test.
The effect of DC saturation is to interfere
with operation of instantaneous relays.
Given the performance characteristics of the
CT and the associated CT burden, will DC
saturation occur quickly enough, and with
sufficient severity, to interfere with the
operation of these instantaneous relays?
A more common issue comes about when
the circuit breaker interrupts an offset fault
current. Interruption of the DC component
of this current leaves one or more CT's
partially magnetized. This remanent flux
4-5
Based on this perspective, the following
tests should be performed:
susceptible to DC saturation. Because the
formulas
are
based
upon
certain
assumptions, there is an error of -0 +0.5
cycles in the time to saturation. The actual
response of the relay to the saturated current
waveform (Figure 3-3) is subject to many
imponderables. An answer to the question
of whether a particular instantaneous relay
will respond within an acceptable time delay
can only be answered by test.
For feeder relays having both time and
instantaneous elements, will the calculated
secondary DC saturation voltage be less than
the CT saturation voltage, VX, for the
maximum fault at which the instantaneous
relay is expected to operate, namely, the
maximum available fault current and will
the "time to saturate" be shorter than the
time required for the measuring algorithm in
the instantaneous relay to respond to the CT
secondary current prior to the point at which
significant DC saturation appears (typically,
one-half cycle)?
For main and tie relays with no
instantaneous elements, DC saturation is not
an issue
Figure 4-4
Typical Waveforms of CT Primary and
Secondary Current With DC Saturation
(Power Systems Relaying Committee. 1976)
If the effect of remanence flux is taken into
account, all relay-CT combinations may
have time-to-saturation of less than one
cycle, effectively preventing relay operation.
This situation is typical of most applications
of overcurrent relays with current
transformers in medium-voltage switchgear.
It is a generally accepted compromise that
non-operation of overcurrent relays may
occur when the remanence flux is large. For
this reason, back-up overcurrent protection,
and bus differential protection, which are
not susceptible to CT saturation, is utilized
in power systems where DC saturation may
be a problem, such as in generating stations.
The standard (ANSI/IEEE. 1996, section
4.5.2) states: "These requirements generally
result in impractically large CTs and hence
compensating steps must be taken to
minimize saturation effects on the relay
protection plan. Some high speed
instantaneous relays can operate before
saturation has time to occur."
The instantaneous relay may operate
immediately if the time to saturation is long
enough to allow it to operate. If the time to
saturation is too short, the relay may operate
several cycles later when the CT emerges
from saturation. The maximum delay for
instantaneous tripping is determined by the
coordination time intervals (CTIs) defined
for the protective system.
Saturation reduces the magnitude of CT
secondary from its ideal value. The impact
of saturation on the response of digital
overcurrent relay is demonstrated by ATP
simulations in Figure 4-5 (Folkers, R, 1999).
In this example, saturation initially reduces
the relay magnitude response by one half, a
reduction that may affect relay performance
in different ways. For example, a high-set
instantaneous 50 element could pick up for
one cycle and then drop out for one to two
cycles. A time-delayed overcurrent element
could respond up to three cycles late. It is
also shown that CT saturation will cause
In the existing relaying system, the timeovercurrent relay functions serve to backup
the instantaneous relay functions. It is the
instantaneous functions that may be
4-6
distance relay to undereach due to reduced
secondary current flowing through it.
The name of the inverse-time curve is selfexplanatory. The higher the magnitude of
the fault or the load current, the less time is
required for the relay to operate. This
provides quick removal of the faulted
portion from the power system. Under
normal load condition or for a lowmagnitude fault current, the relay allows
enough time for other downstream devices
to clear the fault. Because the
microprocessor relay has to process the
current mathematically to decide the time to
trip, the relay can be equipped with different
types of inverse curves.
Figure 4-5
Overcurrent Relay Response to CT Saturated
Current
The user at the time of the study decides the
curve to be used. The relay simply actuates
that curve equation and processes
accordingly. IEEE has created standard
inverse curves. The manufacturer of the
microprocessor relay must comply with
these curves. Depending on the time to trip
for higher current or for lower current,
different curves are available.
The saturation phenomenon in the CT
applications can be evaluated using the
equations that have been summarized in this
chapter and have been coded into a
spreadsheet application that has been
developed and provided with this book. The
application description is included as an
appendix to the chapter. Similarly, the
transient response of the CT during a fault
condition may be evaluated using an EMTP
test circuit that has also been developed.
This application description may also be
found as an appendix to the chapter.
The following list includes a few of the
many popular curves:
•
•
•
•
•
Effect of High Fault Currents on
Coordination
The coordination of protection devices is
necessary to maintain selectivity, (that is, to
remove the portion of the power system that
is experiencing the fault). The inverse-time
current curve provides coordination at the
same time that it offers the speed and
accuracy needed to clear the correct fault
quickly. By the nature of the design, the
inverse-time curves utilize the magnitude of
the current to decide the tripping time,
unlike the instantaneous relays with definite
time.
Moderately inverse
Inverse
Normally inverse
Very inverse
Extremely inverse
Figure
3-4
depicts these
curves with
the equations
set forth by
IEEE.
4-7
The protective device
coordination in a radial
feeder
or
for
any
distribution feeder is very
important.
Bus
Breaker
51
Relay
Line
Fuse 1
Time
F1
Relay
F2
Branches
Fuse 1
T1
T2
F1
CTI
F2
Current
Figure 4-7
Coordination of a Distribution Feeder
The impact of the increased fault current due
to addition of distributed generation on
protective relaying is shown in (PSERC
2005). The impact of adding DGs on the
fault current levels in a 4-bus test system
(Figure 4-8) is shown in
Table 4-1. It can be seen that increased fault
current is reflected throughout the system.
Figure 4-6
IEEE Extremely Inverse Time Overcurrent
Curve
Example of mis-coordination of a feeder
circuit when fault current increases, see
Figure 3-5:
•
•
•
•
Fuse 1 clears fault F1 before relay
operates.
Coordinating time interval (CTI) T1
is sufficient.
Fault current increases beyond
design value T2 < CTI
Breaker opens at same time as Fuse
1, causing wider outage.
Figure 4-8
A simple 4-bus system with new DGs at bus
3 and 4 (PSERC 2005)
Table 4-1
Impact of DG on fault Currents (PSERC 2005)
4-8
Bus
Fault
current
before
installing
DG (pu)
Fault
current
after
installing
DG (pu)
Change
in Fault
current
(%)
1
1.39
2.36
96.8
2
1.37
3.23
135.7
3
1.39
4.17
200
4
1.31
3.53
169.5
condition is detected. Protective relays work
in concert with sensing and control devices
to accomplish this function. There are
several reasons to use protective relaying:
•
•
Protective Relay Ratings and Settings
•
Protective relays (EPRI.2004) have a
reputation for providing reliable service for
many years. Nonetheless, protective relays
are delicate instruments that are susceptible
to the degradation of components that may
affect performance. Due to their design,
numerical relays, Figure 3-6, have
eliminated the degradation that can be
expected from the mechanical components
of electromechanical relays, Figure 3-7.
Numerical relays also use minimal
electronic components when compared to
electronic relays. The failure of a protective
relay to contain and isolate an electrical
problem can have severe plant-wide
repercussions. When an expected protective
action does not occur, the end result of an
electrical abnormality may be catastrophic
equipment damage and prolonged downtime
instead of localized minor damage. Because
of the severe consequences of a failure,
protective relays should be maintained in a
high state of readiness. Critical applications
should be carefully evaluated for redundant
protection.
•
•
•
To provide alarms when measured
process limits are exceeded, thereby
allowing operators an opportunity to
intervene with corrective actions
To isolate faulted circuits or
equipment from the remainder of the
system so that the system can
continue to function
To limit damage to faulted
equipment
To minimize the possibility of fire or
catastrophic damage to adjacent
equipment
To minimize hazards to personnel
To provide post-fault information to
help analyze the root cause
Under normal power system operation,
protective relays remain idle and serve no
active function. However, when required to
operate because of a faulted or undesirable
condition, it is imperative that the relays
function correctly. Another point of concern
is the undesired operation of a protective
relay during normal plant conditions or
tolerable transients. Inadvertent relay
operation can result in unnecessary system
or plant downtime. A maintenance and
surveillance program will help to ensure that
the protective relays respond properly to
normal and abnormal conditions. This
frequency of testing can be extended to
longer periods than electromechanical
devices. The number of tests can also be
reduced due to the design and construction
of the relays. An effective maintenance
program for protective relays accomplishes
two primary goals. First, it provides a high
degree of confidence that the electrical
power protection system will respond to
abnormal conditions as designed. Periodic
Protective relaying is an integral part of any
electrical power system. The fundamental
objective of system protection is to quickly
isolate a problem so that the unaffected
portions of a system can continue to
function. Protective relays are the decisionmaking device in the protection scheme.
They monitor circuit conditions and initiate
protective action when an undesired
4-9
assurance that protective relays are in an
operable status is particularly important.
Relay problems are generally detected by
internal test routines and during operational
checks via a human machine interface
(HMI). Secondly, an effective maintenance
program preserves the relay’s readiness and
helps to counteract normal and abnormal inservice deterioration that can affect a relay’s
electronic components over time. Even
under normal conditions, electrical, thermal,
and environmental stresses are continually at
work, slowly degrading the relays. This
deterioration is much slower in numerical
relays than in electromechanical relays
because numerical relays are not affected by
mechanical
deterioration.
Routine
maintenance checks help to identify any of
the deterioration in the device. The life of
numerical relays cannot be prolonged by
recalibration,
cleaning,
and
general
maintenance because numerical relays are
either functional or not. Failed devices must
be removed, repaired, and/or replaced.
Failed devices or device components such as
printed circuit boards should be sent to the
supplier for repair.
Figure 4-10
Electromechanical Protective Relay (General
Electric Company)
Effects of Fault Currents on Protective
Relays
Operating voltages range from 110 VAC–
480 VAC and 24 VDC–250 VDC. Currents
are provided in either 1A or 5A (AC) rating.
The maximum design rated voltage is the
highest rms alternating voltage or direct
voltage. The maximum current is the highest
rms alternating current or direct current.
These maximum values are the limits at
which the relay can operate continuously.
The
operating
coils
of
older
electromechanical and solid-state relays
typically determined the relay's rating.
Today, numerical relays are supplied with
multiple range or universal power supplies
and binary inputs/outputs that operate at
specific voltages that are set by using
internal links or jumpers in the relay. The
current and voltage sources can also have
multiple operating ratings. Relays can be
provided with secondary current selections
of 1 A or 5 A. Secondary voltages to the
relay can be configured for a range of
voltages and can typically be connected wye
or delta by a setting parameter rather than a
hardware configuration. Voltage inputs can
be configured for phase-phase or phaseground.
Figure 4-9
Numerical Protective Relay (Schweitzer
Engineering Laboratories)
4-10
Relay
The absence of the current
burden is transformers subject the relay
greatly
to the fault currents. Hence,
reduced
high fault currents are
with the directly seen by such relays,
use
of and hence subjecting their
numerical
internal circuitry to a higher
relays.
risk from high fault currents.
The
power
consumption of numerical relays can be
expected in the 0.04–0.10 VA range. This
means that in many applications, the relay
burden is negligible. In older applications
where the relay burden was much higher,
intermediate current transformers may have
been required. Numerical relays normally no
longer
require
these
intermediate
transformers.
Limits: 15 A continuous. 500 A @ 1 sec.,
linear to 100 A. 1250 A @ 1 cycle.
•
Methods for Upgrading Protection
Systems
Update Short Circuit Study
In order to obtain complete coordination of
the protective equipment applied, it may be
necessary to obtain some or all of the
following information on short-circuit
currents for each node or bus:
•
Maximum and minimum momentary
(first cycle) short-circuit current
• Maximum and minimum interrupting
duty short-circuit current
• Maximum and minimum groundfault current
The momentary currents are used to
determine the maximum and minimum
currents to which instantaneous and directacting trip devices respond. The maximum
interrupting current is the value of the
current at which the circuit protection
coordination interval is established. The
minimum interrupting current is needed to
determine if the protection sensitivity of the
circuit is adequate.
Thermal
Relays contain sensitive
overload
electrical components that
capacities
are
also are designed to operate at
higher. For specific values of voltage
and current.
example,
the
effective thermal overload can reach a rating
of 500 A (1 sec) and a dynamic rating of
1250 A (half cycle).
Examples
Electromechanical Relay (GE Multilin.
1997) GE IAC53 1.5-12 A time overcurrent
taps,
instantaneous
range10-80
A.
Continuous current rating of the time
overcurrent unit: 10-30.5A
Short time
current rating of the time overcurrent unit:
I2t = 67,600.
•
For example, with 500:5 CT, 50 kA
of fault current for 1 second is the
limit.
A short-circuit study and coordination study
should be updated when the available short
circuit of the source to a plant is increased.
Update Protective Device Coordination
Study
For example, with a 500:5 CT, 26
kA of fault current for 1 second will
reach the thermal limit of this relay.
The objective of a coordination study is to
determine the characteristics, ratings, and
settings of overcurrent protective devices to
ensure that the minimum unfaulted load is
interrupted when the protective devices
isolate a fault or an overload anywhere in
Digital (Schweitzer. 2003) SEL351 0.5-16 A
time overcurrent range, instantaneous
overcurrent range 0.25-100 A on 5A CT.
4-11
result in mis-operation of relays. The proper
CT models need to be used in the
electromagnetic transient type programs. In
widely used transient programs (PSCAD,
EMTP), CT models are available that have
been developed using Jiles-Atherton theory
of ferromagnetic hysteresis (Jiles and
Atherton, 1986.) and representation of
magnetization curves using a non-integer
power series (Lucas, 1998). These models
compute the secondary current from the CT
winding in response to the primary current.
The electromagnetic models of current
transformers have been developed based on
the non-linear power curve formulation
(Lucas et. al. 1992). The hysteresis loop
model is based upon a current injection that
reflects the parallel combination of nonlinear inductance and a non-linear resistor
(Figure 4-11).
the system. At the same time, the devices
and the settings selected should provide
satisfactory protection against overloads on
the equipment and should interrupt shortcircuit currents as rapidly as possible.
The coordination study provides data useful
for the selection of instrument transformer
ratios, protective relay characteristics, and
settings and fuse ratings. It also provides
other information pertinent to the provision
of optimum protection and selectivity in
coordination of these devices. When plotting
coordination curves, a certain time interval
must be maintained between curves of the
various protective devices in order to ensure
correct sequential operation of the devices.
This interval is called the coordination time
interval.
When coordinating inverse curves, the time
interval is usually 0.3–0.4 sec. This interval
is measured between relays in series at the
instantaneous setting of the load side of the
feeder relay or maximum short-circuit
current, which can flow through both
devices simultaneously.
A basic understanding of time-current
characteristics is essential in any study.
Initial planning and power system data are
also essential for any coordination study.
Figure 4-11
Basis of hysteresis loop
The magnetizing component in the figure is
represented by the non-linear magnetization
curve as
The initial planning process should include
the following activities:
•
•
•
•
•
Develop a one-line diagram
Determine the load flow
Collect data
Conduct a short-circuit study
Determine time current coordination
curves for all the devices in the
system
H = ∑ K i B αi
Eq. 4-6
Where Ki and i are positive in nature.
Typically, 3 terms of power are found to
give good approximation over a wide range.
The core loss is represented by the nonlinear resistor in Figure 4-11 that includes
eddy current loss and hysteresis loss. The
transformer excitation current is then the
summation of the magnetizing current and
the core loss current. The additional
injection is needed to model the asymmetry
in the flux-current loop due to the presence
Modeling Techniques for Protection
Studies
Protective coordination studies need to be
done to ensure that CT saturation does not
4-12
of any remanence. Thus, these models
accurately reflect the behavior of the
instrument transformers during saturation
conditions as well as the presence of
remanence from previous magnetizations
The authors have successfully demonstrated
the accuracy of these models by validating
them against the published practical tests.
There is a close agreement between the
hysteresis loop of a Silectron 53 steel in a
CT obtained from manufacturer (Figure
4-12) and the one obtained using the model
(Figure 4-13).
Figure 4-13
Hysteresis Loop for Silectron 53 from model
In this model, the magnetizing loops of the
transformer core are self generated rather
than being predetermined. The option of
non-iterative based solution in the models
enables them to be used with the real time
relay test simulators. In offline simulations,
iterations are possible resulting in more
accurate solution.
As mentioned earlier, CT models have also
been developed in which the hysteresis loop
of the core material is developed using JilesAtherton theory of ferromagnetic hysteresis.
As per this theory, magnetization of the
hysteresis model is composed of the
reversible component that is related to the
domain bending during the magnetization
process and the irreversible component that
corresponds
to
the
domain
wall
displacement during the pinning effect.
Figure 4-12
Manufacturer’s Hysteresis Loop for Silectron
53
M = M irr + M rev
Eq. 4-7
The saturation characteristics are modeled
using a modified Langevin function. The
models based on this theory involve solving
a set of first-order differential equations in
which the desired parameters can be
determined by the measurements of the
hysteresis loop. The basics of the theory and
the descriptions of the equations and the
4-13
parameters are included in (Jiles and
Atherton, 1986).
Authors in (Annakage, et al, 2000) mention
that the models discussed so far are adequate
for most situations other than the situations
involving reclosing into a permanent fault
after a long duration (> 1 s) when accurate
modeling of renascence is required. Also,
the authors have determined that the shape
of the hystersis loop in the shoulder area is
very critical in situations where burden is
small (e.g. digital relays) in presence of
residual flux. The Langevin function in
Jiles-Atherton model was found to be
unable to accurately replicate the shoulder
area for such situations. The authors have
introduced an improved anhysteretic
function that has an additional flexibility to
achieve a better shape for the simulated B-H
loop.
Summary and Recommendations
Currents in the secondary windings of CTs
are used for protection and metering
applications. Excessive fault currents can
cause CTs to enter saturation. Consequently,
current in CT secondary is not a true
reflection of actual system current flowing
through CT primary. The resultant error can
cause protection system to fail (mis-operate
or failure to operate). Thus, protection
system needs to be updated and coordinated
based on the existing fault current levels and
system
conditions.
The
protection
performance under high fault current
conditions can be improved by using relays
that put reduced burden on CT secondaries
and/or provide compensation for expected
CT saturation.
4-14
APPENDIX 4.1 CT Saturation Evaluation Spreadsheet
This spreadsheet will allow a user to evaluate saturation phenomenon for a CT application. The
sheet that deals with AC saturation phenomenon is shown in Figure 4-14. User needs to enter all
the parameters in the fields shaded in “Green” and the sheet will compute if the CT ratio error
due to AC saturation is going to be significant or not.
Secondary resistance of CT ( Rs)
Leakage Reactance (Xl)
Burden resistance (Rb)
Secondary to Primary turns-ratio (N2:N1)
Primary fault Current (Ip)
Saturation Voltage (Vx)
Secondary Voltage (Vs)
CT ratio error Significant
0.61
3.46
1.6
240
15000
250
257
Yes
ohms
ohms
ohms
A
V
V
Figure 4-14
AC Saturation sheet
Similarly, the sheet that deals with DC saturation phenomenon is shown in Figure 4-15. Again,
user needs to enter the various parameters in fields shaded in “green” and the sheet will compute
if the CT ratio error due to DC saturation is significant. It will also compute the time to
saturation (in cycles) once the fault occurs.
Secondary resistance of CT ( Rs)
Leakage Reactance (Xl)
Burden resistance (Rb)
Burden Imepedance including leads and secondary devices (Zb)
Secondary Impedance (Zs)
Secondary to Primary turns-ratio (N2:N1)
Primary fault Current (Ip)
Primary system resistance up to fault (R)
Primary system reactance up to fault (X)
Remnance
Secondary Voltage (Vs)
Saturation Voltage (Vx)
CT ratio error Significant
Saturation factor (Ks)
Time to saturation (Ts)
0.61
3.6
0.8
1.74
5.95
240
5600
0.5
10
0
796.8333
250
Yes
7.60
1.3
ohms
ohms
ohms
ohms
ohms
A
ohms
ohms
pu
V
V
cycles
Figure 4-15
DC Saturation sheet
The screenshots that are shown here correspond to a 1200/5, C400 CT (ANSI/IEEE. 1996,
section 4.4).
4-15
APPENDIX 4.2 EMTP Model for CT Saturation Evaluation
A circuit has been developed on EMTP-RV platform to evaluate the CT saturation phenomenon
in transmission systems (See Figure 4-16). EMTP is chosen as it is one of the most widely used
programs for performing transient studies. The circuit represents a 2-bus transmission system
(Folkers 1999).
RL1
+
b
b
SW1
R7
?i
+
+
1E15|1E15|0
1e-6
Fault
+.2
8
R_RELAYB
+
8
R_LEADN
R_LeadB
+.2
R_LEADC
+.2
1
+
CT_B
+
1
8
kv = 230
CT_C
2
2
CT_A
+
R_leadA
+.2
+
R_RELAYC
R_RELAYA
Figure 4-16
EMTP Evaluation Circuit
The test system details are:
•
•
•
•
•
•
•
+ AC
c
c
+
RL2
+
2
kv = 230
SRC2
a
a
AC +
1
SRC1
Positive-sequence line impedance, Z1L = 8.19 + j77.57
Zero-sequence line impedance, Z0L = 36.81 + j245.15
Positive- and zero-sequence source impedances, ZS = 0.1 ZL
Peak source voltage, VP = 189500 V, VSend leads VReceive by 30°
The line is split into two elements, one on each side of the fault location.
CT secondaries are wye-connected and grounded through lead and relay impedances.
CTs are modeled as saturable transformer components as shown in Figure 4-17
4-16
.
RL1
?i
RL2
Tideal_unit
+
i
+
k
1e-6,0
+
+
0.576,0
?i
+
100M
Rmag
0.004166666666666667
Lnonl
m
j
L nonlinear
Data function
Figure 4-17
Saturable Transformer Model for CT
The CT model is developed as per the following steps:
1. Model the CT secondary on Winding 1 of the saturable transformer component.
2. On Winding 2, set resistor in RL2 equal to zero. Set inductor in RL2, which must have a value
greater than zero, equal to 10E-6.
3. Set inductor in RL1 equal to zero, since a C-class CT secondary leakage reactance is very
small.
4. Set resistor in RL1 equal to the CT secondary winding resistance.
5. Set magnetizing resistance, Rmag, to 100M-ohm as it very large.
6. Select seven to ten excitation-current versus voltage points from the CT excitation curve
(1200/5 A CT curve in Figure 4-18), to include saturation in the model. Select a point at the
lower end of the curve, several points at, and just above the knee of the curve, and a point at the
upper end of the curve.
7. Convert these current versus voltage points to current versus flux points using the “L nonlinear
Data Function”.
8. Enter the obtained current-flux points into the non-linear inductor (Lnonl) in the model.
4-17
Figure 4-18
CT Characteristics
For normal conditions, the simulation of the test circuit yields the following plots for the currents
in CT primaries and secondaries (Figure 4-19).
Figure 4-19
CT Currents –Normal Conditions
4-18
It is seen from the plots that under normal conditions, the currents in secondaries get stepped
down by a correct ratio and retain the exact sinusoidal shape of primary currents. The current
plots for the fault condition are shown in Figure 4-20. It may be seen that currents in CT
secondaries are distorted due to saturation and the time-to-saturation is about half a cycle. This
circuit lets the user to evaluate the impact of parameters such as fault current levels and CT
burden on the CT saturation. These impacts for the test circuit are summarized in Table 4-2 and
Table 4-3.
Figure 4-20
CT Currents –Fault Conditions
Table 4-2
Impact of Primary Current Level on CT saturation
Primary Current (kA)
Time to saturation (cycles)
7.6
0.6
9.2
0.5
5.6
1.4
Table 4-3
Impact of Relay Burden on CT saturation
Primary Current (kA)
Time to saturation (cycles)
8
0.6
4
2.7
2
NA
4-19
10. Linders, J.R. et. al., 1995, “Relay
Performance Considerations with
Low-Ratio CT’s and High-Fault
Currents,” IEEE Trans. Industry
Applications, Vol. 31, No. 2,
March/April 1995, pp. 392-404
References
1. ANSI/IEEE. 1996. C37.110-1996,
IEEE Guide For The Application of
Current Transformers Used for
Protective Relaying Purposes
11. Lucas, J.R., "Representation of
Magnetisation curves over a wide
region using a non-integer Power
Series", IJEEE, Vol. 25, No. 4, 1988,
Manchester U.P. UK. pp. 335-340.
2. Annakkage, U.D. et. al. , 2000, “A
Current Transformer model based on
the Jiles- Atherton Theory of
Ferromagnetic Hysteresis,” IEEE
Trans. Power Delivery, Vol. 15, No.
1, January 2000, pp. 57-61
12. Lucas,
J.R.,
McLaren,
P.G.,
Keerthipala, W.W.L., and Jayasinghe,
R.P., “Improved Simulation Models
for Current and Voltage Transformers
in Relay Studies,” IEEE Trans. Poer
Delivery, Vol. 7, No. 1, January 1992,
pp. 152-159
3. Blackburn, J.L., 1998, Protective
Relaying Principles and Applications,
New York: Marcel Dekker
4. Elmore, W.A. (ed.), 1994, Protective
Relaying Theory and Applications,
New York: Marcel Dekker
13. NFPA 70-2005, National Electrical
Code, Quincy, MA: National Fire
Protection Association.
5. EPRI. 2004. Protective Relays:
Numerical Protective Relays, EPRI,
Palo Alto, CA: 2004. 1009704.
14. Power Systems Relaying Committee,
1976, Transient Response of Current
Transformers, IEEE Publication 76
CH 1130-4 PWR, New York: IEEE,
1976
6. Folkers, R., 1999, Determine Current
Transformers Suitability using EMTP
Models. SEL.
7. GE Multilin, 1997. Instructions Time
Overcurrent Relays IAC53A IAC53B
IAC54A IAC54B Form 800 and p,
General Electric Co. Publication
GEK-3054H
15. PSERC. 2005. “New Implications of
Power System Fault Current Limits”.
PSERC Publication 05-62, October
2005
16. Schweitzer. 2003. SEL-351-0, -1, -2,
-3, -4 Directional Overcurrent Relay
Reclosing Relay Fault Locator
Instruction
Manual,
20030908,
Schweitzer Engineering Laboratories,
Inc., Pullman, WA
8. Jiles , D. C. and Atherton, D. L.,
“Theory of ferromagnetic hysteresis”,
Journal of magnetism and magnetic
materials, vol. 61, pp. 48, 1986.
9. Kojovic, L. A., “Comparison of
Different Current Modeling
Techniques for Protection System
Studies”, IEEE PES Summer
Meeting, Vol. 3, 2002, pp. 10841089
17. Seveik, D.R., DoCarmo, H.J., 2000
“Reliant Energy HL&P Investigation
Into
Protective
Relaying
CT
Remanence”,
Conference
for
Protective Relay Engineers, Texas
4-20
A&M University, College Station,
Texas, April 11 – 13.
4-21
b. Provide low zero-sequence impedance
for return of the unbalanced fraction
of ac system currents
5
3. To ensure electrical safety, minimizing
energy by:
EFFECT OF HIGH
FAULT CURRENTS ON
GROUNDING GRIDS
a. Rapidly identifying system faults,
leading to reduce fault duration.
b. Limiting touch or step voltages to
levels that restrict body currents to
safe values.
Excessive Fault currents that enter a
grounding system for a substation may
translate into a reduction in electrical
safety due to increased step and touch
potentials.
4. To
contribute
to
electromagnetic
compatibility, eliminating some hazards
and reducing the energy of others.
All of these functions are provided by a
single grounding system. Some elements of
this system may have specific electrical
purposes, but all elements are normally
bonded or coupled together, forming one
system to be designed or analyzed.
This chapter reviews some of the basics
of ground grid designs as they relate to
high fault currents.
Introduction
The grounding or earthing system (EPRI,
2004a, sec. 5.2.2) is a total set of measures
used to connect the electrically conductive
components of a power system to earth. The
grounding system is an essential part of both
high and low voltage electrical power
networks, and has at least four important
roles:
When fault currents that are in excess of
design values
enter
a In order to prevent these
grounding
effects,
regular
system,
the recalculation of ground
following
grid parameters should
effects may be made, in addition to
occur:
normal ground grid
maintenance, to ensure
1. Reduction
that fault current levels
in
are not exceeded:
electrical
safety:
1. Ground resistance
increased
measurements.
step and
touch
2. Short
circuit
potentials
calculations.
1. To protect against lightning by:
a. Providing
an
electrically
and
mechanically robust path for current
to flow to ground
b. Limiting potential differences across
electrical insulation on stricken
towers
c. Reducing the number of flashovers
that occur.
2. Damage
or failure
of
grounding
equipment
2. For correct operation of the power
system, minimizing energy by,
a. Providing unambiguous identification
of faults, so that the correct protection
systems operate
Recalculation of step
and
touch
potentials.
a. Thermal damage due to excessive
short circuit currents
5-1
b. Mechanical damage due to excessive
short circuit stresses
and ground grid. (IEEE 2000, Sec.
17)
c. Drying of the soil resulting in
increased soil resistivity (IEEE 2000
Sec. 12.3)
9. Increasing thickness of upper layer of
crushed stone. (IEEE 2000, Sec. 12.5)
10. Use of soil treatment to lower resistivity
(IEEE 2000, Sec. 14.5)
d. Insulation failure due to high-induced
voltages (IEEE. 1996.)
Table 5-1
Critical Parameters in Ground Grid Design
3. Possible effects of grounding grid
degradation on the electrical power
system
a. Reduced lightning protection
b. Misoperation
protection
of
ground
Symbol
Name
Equation
Typical
values
units
IG
Maximum
Grid
Current
IG = D f × I g
0.5-10
kA
fault
c. Increased zero-sequence impedance
for unbalanced load currents.
d. Reduced
compatibility
Df = Decrement
factor
Ig = Sym. grid
current
electromagnetic
There are various measures that can be taken
to reinforce a ground grid. These include
(IEEE 2000, Sec. 16.6):
tf
Fault
duration
--
0.251.0
s
ts
Shock
duration
ts = tf
0.251.0
s
ρ
Soil
resistivity
From
measurement
10-10
4
Ω-m
ρs
Resistivity
of surface
layer
From
measurement
Dry:
4000-1
9
x 10
Ω-m
1. Adding ground grid conductors,
decreasing the spacing of conductors.
Wet:
21-6 x
106
2. Increasing the area of the grid.
3. Adding parallel conductors around the
perimeter of the ground grid.
Table 5-2
Procedure for Ground Grid Design
4. Adding ground rods, with closer spacing
at the perimeter.
5. Diverting fault currents to other paths.
(Popovi . 2000)
Step
Description
Result
1
Site survey, soil resistivity
test
Area, ρ
2
Conductor size, zero
sequence current, faultclearing time.
3I0, tc, d
3
Step and touch potentials
Estep, Etouch
4
Conductor loop design,
conductor spacing, ground
rod locations
Various
dimensions
5
Estimated resistance of
grounding system in uniform
soil
RG
6
Recalculate ground current
IG, tf
6. Limiting total fault current.
7. Barring access to hazardous areas.
8. Connecting overhead ground wires from
transmission lines.
a. Decreasing tower footing resistances.
b.Reinforcing or replacing connectors
between above-ground components
5-2
Step
Description
The critical parameters in the design of the
ground grid are listed in Table 5-1. The
ground grid procedure is listed in Table 5-2.
Result
and fault duration based on
current splits, worst-case
fault, and future expansion.
7
If GPR < tolerable touch
voltage, go to final step
IGRG<Etouch
8
If GPR > tolerable touch
voltage, calculate mesh and
step voltages
Em, Es,
If mesh voltage < tolerable
touch voltage, go to next
step. Otherwise go to step
11.
Em<Etouch
10
If step voltage < tolerable
touch voltage, go to final
step. Otherwise go to step
11.
Es<Etouch
If mesh or step voltage >
tolerable touch voltage,
revision of design is
required.
IEEE 2000,
Sec. 16.6
Equipment ground
conductors, additional grid
conductors, ground rods as
needed. Final design
review.
IEEE 2000,
Sec. 17
12
The shape and area of the substation are
determined,
and
ground
resistance
measurements (ANSI/IEEE. 1983.) (IEEE.
1991a.) are taken to determine the ground
resistivity.
various K
9
11
Site Survey
Conductor Sizing
Most important in terms of over-current
phenomena is the sizing of the ground grid
conductors. The design procedure makes
two assumptions:
1. Adiabatic heating of the conductor
2. Thermal capacity per unit volume
remains constant (usually true for
short fault durations)
Summary of Ground Grid Design
Procedures
The conductor size can then be determined:
Akcmil = ID f K f t c
Design of a ground grid is part of the overall
design of a substation. The ground rods will
be driven and the ground grid constructed
before the surface layer of gravel is poured
and the above ground portions of the
substation constructed. The design goals, as
listed in IEEE Standard 80-2000 (IEEE,
2000, Sec. 16.1) are:
Eq. 5-1
Where:
tc = fusing time of the conductor in s.
Akcmil = conductor cross sectional area in
kcmil.
I = rms symmetrical fault current in
kA.
1. “To provide means to dissipate
electric currents into the earth without
exceeding
any
operating
and
equipment limits.”
It is assumed that the symmetrical rms
fault current I f ≈ 3I 0 from the ground
fault calculations for the substation.
2. “To assure that a person in the
vicinity of grounded facilities is not
exposed to the danger of critical
electrical shock.”
A summary of the procedure is supplied in
(Keil. 2003.) The theory is discussed in
depth in (Meliopoulos. 1988. Ch. 5 and 8).
Kf = material-fusing constant. Typical
values are 7.00 for soft-drawn copper,
10.45 for 40% conductivity copper-clad
steel wire and 15.95 for steel conductor.
(IEEE. 2000. Table 2)
5-3
Df = Decrement Factor. Where fault
durations are less than 1 s or the X/R
ratio is greater than 5, the asymmetry of
fault current waveforms produces
additional heating, which must be taken
into account:
T
Df = 1+ a
tf
−2 t f
⎛
⎜
Ta
⎜1 − e
⎜
⎝
⎞
⎟
⎟
⎟
⎠
Copper Conductors. Clearing Time 0.5
seconds
Conductor
#4/0 AWG
500 kcmil
Symmetrical
42.7
101
X/R = 10
41.6
98
X/R = 20
40.6
96
X/R = 30
39.6
94
X/R = 40
38.8
92
Eq. 5-2
Where:
tf
= fault duration in s.
Zsys
If
Ta = time constant X/2πfR in s.
Vsys
This is illustrated in Figure 5-1 for several
X/R ratios.
Ig
RB
Ib
Decrement Factor
ETouch
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
Rf
Rf
Vth= Touch voltage
Zth= Rf/2
Ground Grid
Figure 5-2
Touch Voltage
0
10
20
30
40
50
60
Fault Duration Cycles
10
20
30
40
Zsys
Figure 5-1
Decrement Factor Versus Fault Duration for
Four Different X/R Ratios (IEEE 2000. Table
10)
If
Vsys
Ig
RB
Ib
Ib
EStep
The
A safety factor is usually
conductors
of
a applied in the design to allow
for future growth in fault
ground
grid have current magnitudes.
been
designed for a particular maximum fault
current, X/R ratio and clearing time.
Examples are shown in Table 5-3.
Rf
Rf
1 meter
Vth= Step voltage
Zth= 2Rf
Ground Grid
Figure 5-3
Step Voltage
Table 5-3
Fusing Currents in Symmetrical kA for
Annealed Soft-Drawn 100% Conductivity
5-4
ρs = ρ if there is no surface layer.
Step and Touch Voltages
Normally, ρs > ρ
If a person standing on a surface whose
potential has risen due to the flow of ground
current touches a grounded object, they
experience a touch voltage (Figure 5-2). A
Thévenin equivalent circuit of the person
exposed to the touch voltage is also shown.
Whether the touch voltage is hazardous can
be determined by comparison with the
calculated safe level of touch voltage for
that substation.
Kw = 0.116 for “50” or 0.157 for “70”.
ts = duration of the shock current,
seconds.
CS = surface layer-derating factor.
CS = 1.0 if there is no surface layer,
otherwise an approximate formula
(within 5% of computer models) may be
used:
Similarly, if a person is standing on the
surface, and the flow of ground current
causes a dangerous voltage drop to occur
between their feet, they are exposed to a step
voltage (Figure 5-3). A Thévenin equivalent
circuit of the person exposed to the step
voltage is also shown.
⎛
ρ ⎞
⎟⎟
0.09⎜⎜1 −
ρ
s ⎠
⎝
Cs = 1 −
2hs + 0.09
Where:
The safe levels of step and touch potentials
are defined based upon a person’s body
weight and the length of exposure. The usual
standards (IEEE. 2000. Section 8.3) used are
for 50 and 70 kg (110 and 154 lb) body
weights. The step and touch potentials are
calculated using:
E XW = (1000 + m ⋅ A ⋅ C s ⋅ ρ s )
Kw
ts
Eq. 5-4
hs = depth of the surface material (m).
Cs will approach 1.0 as hs increases and
as ρs → ρ. It will approach 0 as ρs → 0.
Ground Grid Layout
Using the shape and area previously
determined, a grid is laid out, at a depth, h,
with spacing D, and total length of buried
conductor LT. If ground rods, unequal
spacing, or shape other than square are used,
other parameters will apply as well.
Eq. 5-3
Where:
X = either “step” or “touch”.
Ground Resistance Calculation
W = either “50” or “70”.
The resistance of the grounding grid, Rg, can
be estimated using Sverak’s equation (IEEE.
2000. Section 14.2):
1000Ω
is the typical resistance of the
human body.
m = 0 for metal to metal touch voltage, 1
otherwise .
⎡ 1
Rg = ρ ⎢
+
⎢⎣ LT
A = 6 for “step” or 1.5 for “touch” A
derivation for these constants appears in
(Meliopoulos. 1988. p. 131.).
⎛
⎞⎤
1
⎜1 +
⎟⎥
20 A ⎜⎝ 1 + h 20 A ⎟⎠⎥⎦
1
Eq. 5-5
Where:
ρs = resistivity of the surface material,
Ω-m.
ρ = earth resistivity in Ω-m.
5-5
LT = total length of buried conductors in
m.
1
2π
Km =
A = total area of the grid in m2.
+
Calculation of Maximum Grid Current
The maximum current that is used in ground
potential rise calculations is not always the
same as the maximum current used for
conductor sizing. The maximum grid current
is:
I G = D f S f 3I 0
Ki = irregularity factor, defined as:
K i = 0.644 + 0.148n
The variables in Km and Ki are defined as
follows:
Eq. 5-6
D =
spacing
conductors, m.
Sf = current split factor determined
from the detailed substation short circuit
calculations (IEEE. 2000. Section 15.)
Calculation of Ground Potential Rise
(GPR)
Eq. 5-7
h
= depth of the grid, m.
d
= diameter of the grid conductor, m.
K ii =
1
(2n )
2
n
, if there are no ground rods
or only a few ground rods, with none
located on the corners or on the
perimeter.
Mesh Voltage
Kh is the grid depth factor, defined as:
The basis of the design procedure is to
minimize the “mesh voltage,” (Em) which is
the maximum touch voltage within the area
of the ground grid, which is taken to mean at
the center of the corner mesh, which is the
usual point of maximum.
LM
parallel
Kii=1, if there are ground rods along the
perimeter or in the corners and along the
perimeter and throughout the grid area.
This will be compared with the touch
potentials, Etouch50 and Etouch70. If it is larger, the
mesh voltage will be calculated.
ρ Km Ki I G
between
Kii = inner conductor factor, defined as:
The ground potential rise is calculated as:
Em =
⎤⎤
K ii ⎡
8
ln ⎢
⎥
K h ⎣ π (2n − 1) ⎥⎦ ⎦
Eq. 5-9
Where:
EGPR = Rg I G
⎡ ⎡ D2
(D + 2h )2 − h ⎤
+
⎢ln ⎢
⎥
8D d
4d ⎥⎦
⎢⎣ ⎢⎣ 16hd
Kh = 1+
h
1m
n = effective number
conductors in a grid:
Eq. 5-8
n = n a nb n c n d
Eq. 5-10
of
parallel
Eq. 5-11
Where:
Where:
ρ = earth resistivity in Ω-m.
na =
Km = geometrical factor, defined as:
5-6
2 LC
LP
Eq. 5-12
⎧1 for square grids
⎪
n b = ⎨ LP
⎪
⎩ 4 A
minimized, the step voltage is brought
within limits as well.
Eq. 5-13
Step Voltage
The maximum step voltage is assumed to
take place for a one-meter stride across the
perimeter of the ground grid at its most
extreme corner:
⎧1 for square and rectangular grids
⎪⎪
0.7 A
nb = ⎨ ⎡ L L ⎤ L x L y
x y
⎪⎢
⎥
⎪⎩⎣ A ⎦
Es =
Eq. 5-14
⎧1 for square, rectangular and L - shaped grids
⎪
n d = ⎨ Dm
⎪ L2 L2
⎩ x y
Eq. 5-15
Eq. 5-17
LS
Where:
LS = effective
conductor:
length
LS = 0.75LC + 0.85LR
and
of
buried
Eq. 5-18
LC = total length of conductor in the
horizontal grid, m
KS = spacing factor for step voltage:
LP = length of the perimeter of the grid,
m.
KS =
1⎡1
If
LM = effective buried length in m:
modified.
LM
1
(
)⎤
when 0.25 m < h < 2.5 m .
Ly = maximum length of the grid in the
y direction, m.
⎡
⎛
Lr
⎜
⎢
= LC + 1.55 + 1.22⎜
⎢
2
⎜ L x + L2y
⎢⎣
⎝
1
+
+ 1 − 0.5 n −2 ⎥
π ⎢⎣ 2h D + h D
⎦
Eq. 5-19
Lx = maximum length of the grid in the
x direction, m.
the
step
voltage is higher than
the design must be
E Step50 and E Step 70 ,
⎞⎤
⎟⎥
⎟⎥ LR
⎟⎥
⎠⎦
Detailed Design
After the safe design has been obtained, the
detail can be added such as equipment
ground
conductors,
additional
grid
conductors, ground rods as needed for surge
arresters and other equipment. A final
design review is performed.
Eq. 5-16
Where:
LR = individual ground rod length, m.
If
ρ K S Ki I G
Example Design from IEEE Standard 80
Dm = the maximum distance between
any two points on the grid, m.
Step 1.
the
A = 70 m x 70 m
mesh voltage is higher than
E Touch 50 and ETouch 70 , the design must be
modified. Once the mesh voltage is
ρ=400 Ω-m
5-7
may be aggravated by high fault currents,
such as:
Step 2. 30% Copper-clad steel wire 2/0
AWG
•
Step 3.
E Step 70 = 2687 V
•
ETouch 70 = 838 V
Step 4.
•
D=7m
•
•
LT = 1540 m
h = 0.5 m
•
Step 5. Rg = 2.78 Ω
Step 6. IG = 1908 A
•
Step 7. EGPR = 5304 V >> 838 V
Step 8. Em = 1002 V > 838 V
•
Step 9. Ground rods added to periphery and
steps 5-8 repeated:
•
Rg = 2.75 Ω
•
IG = 1908 A
EGPR = 5247 V >> 838 V
Em = 747 V < 838 V
Drying of the soil, increasing the
ground resistance
Excessive voltage drops in the
conductors and connectors due to
high currents
Fusing, melting and connector
failures
Arcing, burning and open circuits
Altered current flow paths, further
increasing voltage drops
Corroded or otherwise damaged
conductors and connectors (Lawson.
1988.)
Thinning of the protective surface
layer of crushed stone or gravel
Weeds and shrubs growing in
surface layer
Mixing of the surface layer with soil
and dust, decreasing its resistivity
Failure of static wire, ground or
neutral wire connections from
transmission and distribution lines to
substation.
Reduction in Electrical Safety: Increased
Step and Touch Potentials
Step 10. Es = 549 V < 2687 V
The mechanisms by which these and other
factors can cause injury or death are the step
and touch potentials.
Step 11. Not necessary
Step 12. Ready to proceed.
⎡ 1
Rg = ρ ⎢
+
⎢⎣ LT
Effects of High Fault Currents in Ground
Grids
Failure Mechanisms
⎛
⎞⎤
1
⎜1 +
⎟⎥
20 A ⎜⎝ 1 + h 20 A ⎟⎠⎥⎦
1
Eq. 5-20
There are
There are many possible
many
causes for increased Ground
possible
Potential Rise in substations,
causes for which may be aggravated by
increased
high fault currents.
Ground
Potential Rise (GPR) in substations, which
EGPR = Rg I G
Eq. 5-21
When exposed to high fault currents, ρ can
increase due to drying, LT can decrease due
to conductor damage, all increasing Rg.
5-8
fusing current of the conductors to which
they are attached, and heating from fault
currents for up to 90% of the fusing current
for 10 seconds.
Similarly, the mesh voltage
Em =
ρ K m Ki I G
Eq. 5-22
LM
The test specifications in IEEE Standard
837-2002 call for the construction of a
control conductor, Figure 5-4, of length LCC1
and resistance RCC1. If the connector joins
two different materials, a non-control
conductor, LCC2, RCC2 should also be
constructed. A test loop, containing up to
four connector assemblies under test is
assembled. Each connector assembly
contains of two conductor samples of
lengths LSample1 and LSample2. The overall
resistance at a temperature 20 °C is RTotal. (If
the resistance measurement is not at 20 °C,
then the measurement must be corrected to
20 °C.) If the conductors are stranded, then
equalizers must be used at each end where a
connector is not in place. The purpose of
the equalizers is to establish an equipotential
plane across the ends of the conductor
strands. All samples used in short circuit
tests must have a resistance at 20 °C of no
more than 110% of that of the control
conductor:
increases with IG, ρ and decreasing LM. The
step voltage
Es =
ρ K S Ki I G
Eq. 5-23
LS
also increases with IG, ρ and decreasing LS.
The constants K are geometrical and do not
change with increasing fault current.
Damage or Failure of Grounding
Equipment
Thermal Damage to Conductors Due to
Excessive Short Circuit Currents
An increase in fault current will decrease the
fusing time of the grid conductors:
⎛ Akcmil
tc = ⎜
⎜ I ⋅ Df ⋅ K f
⎝
⎞
⎟
⎟
⎠
2
Eq. 5-24
Examples of fusing current calculation
results are shown in Table 5-4. It is
recommended that the ground grid
conductor thermal limit be plotted as a point
on a time current curve, Figure 5-5, and an
2
i t line be extended upward from it to
represent the ground grid thermal damage
curve. If this curve is exceeded, the
conductors may be subject to fusing, melting
or other forms of thermal and mechanical
damage.
RTotal LCC1
≤ 1.10
RCC1 LSample1 + LSample 2
(
)
Eq. 5-25
If there are two types of conductor, then:
RTotal
≤ 1.10 Eq. 5-26
RCC1 LSample1 RCC 2 LSample 2
+
LCC1
LCC 2
The electromagnetic force test applies an
asymmetrical waveform with the following
specifications:
Connector Damage Due to Excessive Short
Circuit Stresses
1. Rms value equal to 1 second fusing
current for the conductor.
This
affects
primarily
permanent
connections. (IEEE 2002.) Grounding grid
connections should withstand short circuit
electromagnetic forces up to the 1.0 second
Peak value for first half cycle 2.7 times
rms value (fully offset)
5-9
duration fall under the two points defined by
these tests, as listed in Table 5-5, there
should be no failures of connectors due to
excessive fault currents. It is recommended
that the connector thermal damage limit be
plotted as a point on a time current curve,
Figure 5-5, and an i2t line be extended
upward from it to represent the damage
curve. A vertical line may represent the
mechanical damage curve.
Test circuit highly inductive, with X/R >
20.
Test current duration minimum 0.2
seconds (12 cycles @ 60 Hz),
maximum 1 second to avoid fusing.
Results after the first test are not to exceed
110% of RCC1, and 150% after three tests.
The fault current test applies a symmetrical
fault current of 90% of the 10-second fusing
current for 10 seconds, repeated three times.
The evaluation of the test consists of
disassembly and dissection of the
connection and inspection for signs of
melting or other damage.
Drying of the Soil Resulting in Increased
Soil Resistivity
A current density of less than 200 A/m2 for
1 s is recommended. (IEEE 2000 Sec. 12.3)
Heating of soil whose temperature is above
the freezing point has negligible effect on its
resistivity.
Fusing currents for both tests are calculated
using the cable ampacity equation (IEEE.
2002. Annex C):
⎛ T + Tm ⎞
⎟
ln⎜⎜ 0
T0 + Ta ⎟⎠
⎝
I=A
β tc
The effect of moisture content on soil
resistivity is given in the IEEE Green Book
(IEEE. 1991b. Table 11) for several soil
types. Resistivity is quite constant above
22% moisture content, but increases
dramatically below that. Figure 5-6 shows
this effect for three different types of soil.
Without performing three-dimensional
electro-magnetic simulations, there is no
easy way to calculate the drying of soil by
the passage of electrical current, and the
consequent increase in resistivity. Similar
calculations have been performed for HVDC
terminals (Villas and Portela. 2003a and
2003b).
Eq. 5-27
Where:
A = conductor area in mm2.
T0 = conductor material temperature
constant, 234 °C for annealed soft-drawn
100% conductivity Copper.
Tm = fusing temperature, 1083 °C for
annealed soft-drawn 100% conductivity
Copper.
RTotal
LSample2
LSample1
Ta = initial conductor temperature in
°C.
Test
Loop
β = material factor, 19.8 for annealed
soft-drawn 100% conductivity Copper.
tc = time of current flow in seconds.
Test
Loop
Values of the constants for other conductor
constants may be found in IEEE Standard
837-2002 and other references. Results of
example calculations are shown in Table
5-4. As long as the fault magnitude and
Conductor
Connection
Conductor
Equalizer
Equalizer
Equalizer
Equalizer
Control Conductor
LCC1
(RCC1)
Figure 5-4
Short Circuit Testing of Ground Grid
5-10
Connectors (IEEE 2002). The Test Loop
Contains a One Through Four Connector
Assemblies
Table 5-5
Fault Current Tests for Connectors
Table 5-4
Fusing Currents and Test Currents for
Annealed Soft-Drawn 100% Conductivity
Copper Conductors
Conductor
#4/0 AWG
500 kcmil
Fusing Current,
kA, rms,
symmetrical, 1.0
seconds
21.4
50.4
Electromagnetic
Force Test
Current, kA,
peak 0.2
seconds
57.6
136
Fusing Current,
kA, rms,
symmetrical,
10.0 seconds
9.5
22.4
Minimum Fault
Test Current,
kA, rms,
symmetrical,
10.0 seconds
8.6
20.2
Test
Force
Thermal
Fault current
duration
0.2 s
10.0 s
Fusing current
duration
1.0 s
10.0 s
Fault current
X/R
20
N/A
Fault current
peak/ rms
2.7
2
Fault current
rms/ Fusing
current rms
1.0
0.9
Figure 5-5
Typical Time-Current Curves Showing
Thermal and Mechanical Withstand for
Ground Grid Conductors (Assuming 0.5
Second Fault Clearing Time) and Connectors
5-11
blown debris/dirt infill and vehicular
traffic.
Resistivity Ohm-m
Effect of Moisture Content on Soil Resistivity
2000
1800
1600
1400
1200
1000
800
600
400
200
0
5. Corrosion, sometimes to the point of
disappearance of ground grid
conductors, especially steel or copperclad conductors.
0
5
10
15
20
25
6. Opening of underground connections
between grid conductors.
30
Percent by Weight
Sandy Loam
Top Soil
7. Loose or open connections from
equipment to the ground grid.
Red Clay
8. Undocumented changed to ground
grid design.
Most respondents did evaluate their
grounding grids only when there was a
problem or when an expansion was planned.
The methods used for ground grid
evaluations were:
Figure 5-6
Effect of Moisture Content on Soil Resistivity
(IEEE 1991b)
Case Studies
Survey of Substation Grounding System
Assessment and Refurbishment Practices
1. Current injection to verify continuity
An IEEE Task Force (IEEE. 2005.)
conducted a survey of North American
utilities practices regarding assessment and
refurbishment of substation grounding
systems. The reasons for assessment and
refurbishment of substation grounding
systems are given as:
2. Visual inspection
3. Resistance test
Most respondents either have engineering
standards in place for evaluating results, or
were developing standards. If the assessment
showed problems, the utility would either
upgrade or replace the grounding grid.
1. Original design based on inadequate
or
undocumented
assumptions,
obsolete design methods, earlier
standards.
High resistivity surfacing layers were used
by 77% of respondents, while only 15% test
the resistivity of this material at the time of
installation, and almost none retest it during
ground grid assessment or have a
maintenance procedure for it. The required
depth for the material is typically 3 to 6
inches (75 to 150 mm), with a typical wet
resistivity of 3000 Ω-m.
2. Increased available fault current due
to new generation or substation
expansion.
3. Ground resistance can increase if the
original design relied on distribution
neutral connections, some of which
were later removed, or replaced with
jacketed conductors or enclosed in
plastic conduit.
Ground grid upgrades and associated testing
were undertaken by 75% of respondents.
This includes upgrading of surfacing
material. The most common test instruments
used were:
4. Deterioration of surface rock layer
due to construction activities, wind-
5-12
1. EPRI Smart Ground Meter (EPRI.
2004b) (Meliopoulos et al. 1994.)
touch voltages. The required depth for
the material is typically 3 to 6 inches
(75 to 150 mm), with a typical wet
resistivity of 3000 Ω-m.
2. Biddle DET/2
3. Multi-Amp GTS-300
6. Samples of the rock used for
surfacing should be tested regularly,
and a visual inspection should be
performed. Asphalt should be
checked for cracks and holes should
be patched.
4. NGI Unilap Geox
A full 90% of utilities do not de-energize
substations for ground grid testing. Most
respondents do not take into account the
impact of remote grids or current splits in
their evaluation. Only 33% of respondents
use any analysis or modeling to evaluate the
substation grounding system assessment
results. Software which is used includes:
Safety Assessments of Transit Supply
Substations and 161/69-kV Substations
The safety assessment of substation designs
for the Taipei Rail Transit Systems (TRTS)
is presented, using one substation as an
example (Lee. 2004.) These calculations are
based on IEEE Standard 80 Touch and Step
voltage criteria for a 50 kg person.
Grounding grids are present in the following
locations:
1. CDEGS
2. WINIGS
3. EPRI (EPRI. 1992a.)
The following best practices were
recommended as a result of this survey:
1. Bulk Supply Substations (BSS)
a. Primary 161 kV
1. Evaluation of the effectiveness of
grounding grids should be done on a
regular basis, typically every 5-10
years.
b. Secondary 22.8 kV
Traction Supply Substation (TSS)
c. Primary 22.8 kV
2. A standard should be developed
which covers both the design and
maintenance of a grounding system.
d. Secondary 589 V
3. Knowledge of the conductor material
and the soil characteristics will help
to determine how often a grounding
system
should
be
inspected.
Conductivity
tests
and
visual
inspections should be performed if a
problem occurs or is suspected.
e. 750 V DC negative return
The second example (Lee and Chang. 2005.)
consists of a comparison between indoor and
outdoor 161/69-kV substations of the
Taiwan Power Company (TPC). The
outdoor substation has a 166 x 137 m
ground grid, rectangular shape, with a
diagonal of 200 m. The indoor substation is
smaller, 107 x 82 m, with a diagonal of 130
m, and some irregularities in its shape.
4. Test equipment should be easy to use
and yet capable of accounting for
complex factors such as system
neutrals, remote ground grids and
current splits.
These examples emphasizes the importance
of the current division factor Sf ,which is the
percentage of ground fault current assumed
to be flowing between the grounding grid
5. High resistivity surfacing layers
should be used to control step and
5-13
and the surrounding earth.
Several
comparison tables and charts are provided to
show the effects of varying depths of ground
grid, h, spacing between grid conductors, D,
and surface layer resistivity ρS and shock
duration ts on mesh and step voltages. In
general:
Grounding Systems for Electric Traction.
The safety assessment (Natarajan et al.
2001.) was performed for a 47.05-mile
section of the Amtrak Northeast Corridor
system west of New London Connecticut.
This 27.5 kV single-phase system is
supplied from a 115 kV three-phase utility
transmission system. There are several
grounding grids, at New London and at the
paralleling stations. The New London grid
is 130 x 40 ft., with 19 foot spacing. At the
paralleling station, the area is 40 x 80 feet,
with 10 foot spacing. Rails, train platforms
and bridges are also grounded. There are
also numerous buried pipes that must be
accounted for. The soil at New London was
two layers, with an upper 10-foot deep layer
of 51 Ω-m, and a lower layer of 921 Ω-m.
Soil resistivity measurements were different
for each location in the system.
•
•
•
Step voltage decreases as h increases
Step voltage decreases as d increases
Mesh voltage increases as h
increases
• Mesh voltage increases as d
increases
• Permissible step and touch voltages
decrease as ts increases
• Permissible step and touch voltages
increase as ρS increases
The tolerable shock durations for a 50 kg
person are:
⎤
⎡ 0.116
(1.5C S ρ S + 1000)⎥
t smesh = ⎢
⎦
⎣ E mesh
⎡ 0.116
⎤
(6.0C S ρ S + 1000)⎥
t sstep = ⎢
⎣⎢ E step
⎦⎥
The Integrated Grounding System (IGS)
software was used to study the system and
produce plots of GPR and touch voltages.
The results were all found to be well within
the allowable limits calculated from IEEE
Standard 80 for 50 kg persons.
2
2
Ground Current Measurement During a
Fault
Tolerable shock durations increase as ρS
increases.
Ground current distribution in a 150 kV 50
Hz system (Maarten et al. 2003.) was
measured by injection of 60 Hz current from
a 230 V generator. The transmission system
consisted of two 150 kV substations, 21 km
apart, 3.3 km of which was insulated
underground cables, and the overhead
transmission lines. The current was injected
into one phase, while the other two phases
were left open. At the other end of the line,
the other two phase conductors were
grounded.
Using
this
measurement
technique, the current distribution between
cable sheaths and ground, and between
skywires and ground can be found. The
results were verified with EMTP
calculations.
The minimum buried conductor length is
discussed as a criterion for assessing the
safety of a grounding grid:
Lm = ρK m K i I g
tS
116 + 0.174C S ρ S
This paper highlights the importance of
maintenance of the high resistivity of the
surface layer and the underground conductor
length through integrity of the conductors
and their connections.
5-14
petrochemical plant with a 230 kV - 13.8 kV
substation having 400 MVA capacity and 50
kA available symmetrical rms fault current.
In this case, the ground grid covers the
entire plant, and not just the substation.
This older article uses the 1976 edition of
IEEE Standard 80. Several useful points are
brought out:
Design of Ground Grid for a Transition
Station System
This substation design example (Villas et al.
1990) was conducted in a large city with
high short circuit currents where there was
not sufficient ground area for either earth
resistance measurements or for constructing
an adequate grounding grid. In addition, the
project was hampered by the presence of
metallic objects such as water and gas pipes
and building foundations in the vicinity.
The substation is a transition station, that is,
it is at the junction between incoming
overhead transmission lines and outgoing
underground cables.
•
•
•
There is strong inductive coupling between
the cable sheaths, which are spaced closely
together, while there is low inductive
coupling between the transmission line
shield wires.
•
Several
hypotheses
were
examined
regarding the use of ground wires and
counterpoises. The conclusion was to use
two 250 kcmil bare copper bondings
connected to cable sheaths between
substations. These cables are laid in ducts.
The high ground fault current would then be
partially dissipated by the ground grid in the
neighboring substation. In addition two
266.8 kcmil ACSR ground-wires were
added to the overhead circuits. GPR reduced
from 24.9 kV to 4.3 kV with addition of
these conductors. The ground grid design of
the transition substation was:
•
•
•
•
•
The typical values of surface soil
resistivity in the standards may not
be accurate.
Use of the entire ground fault
current, instead of that portion
entering the ground, will result in
overdesign of the ground grid.
The addition of ground wires to
overhead lines can reduce the size of
a ground grid.
In an industrial plant, all areas within
the plant fence should be
investigated for mesh voltages.
Deep-Ground-Well Method
The grounding resistance of a substation can
be decreased by the deep-ground-well
method. (He et al. 2005.) When excessive
ground fault currents occur, high GPR has
had the following effects on substations in
China:
•
•
•
•
•
Destruction of control cables.
High voltage in control room.
Threat to safety of operator.
Control device malfunction
Control device rejection of operator
instructions.
China has Fault current levels have
many
been increasing with the
substations
rapid expansion of the
in
urban power system.
areas with
low soil resistivity, in hilly areas. The
following mitigation methods have been
applied with varying degrees of success:
Area 81.5 m x 52.5 m,
Depth 0.6 m,
Conductors: bare copper 126.7 mm2,
Resistance 3.01 Ω
Touch voltage 840 V.
Large Industrial Plants
Ground grid designs for large industrial
plants (Zotos. 1988) also use IEEE Standard
80.
This ground grid concerned a
5-15
•
•
Enlarging the grounding grid
Adding a subsidiary external
grounding grid
Increasing the burial depth of the
grounding grid
Connecting with other metallic
objects such as steel foundations of
buildings.
Adding long vertical grounding
electrodes
Replacing the soils around the grid
with low resistivity soils.
decreased to 0.5 Ω, where it has remained
constant for several years.
It has been found that usually only some of
these methods are suitable for a particular
location and that two or more are needed to
produce the desired result.
Resistance measurements are taken using
the Wenner (four-probe) method, and
analyzed to provide the following results:
•
•
•
•
Comparison with vertical grounding
electrodes shows an improvement of 1.57 to
3.27 times more resistance for vertical
grounding electrodes.
Design Using Two-Layer Soil Model
Substation grounding grid design with the
two-layer soil model can be performed using
computer analysis methods. (Villas et al.
1988). A computer program originally
designed for electric field calculations was
used, rather than commercial software.
ρ1
resistivity of upper layer
3400 Ω-m
A proposed explosive grounding technique
was found to be too expensive. (Meng, et al.
1999)
ρ2
resistivity of lower layer
553 Ω-m
The alternative of the deep grounding well
was implemented in the reconstruction of
the 110 kV Luohu Substation. This is a 90
m x 90 m grounding grid with 1.79 Ω
resistance, which was enlarged in 1989 to 90
m x 120 m and the resistance decreased to
1.35 Ω.
h
boundary depth
3.5 m
The example was for a remote substation in
Brazil. The design data was:
The grounding system was rebuilt again in
1999 with the deep-well method, where ten
wells between 11 and 15 m were drilled
around the periphery of the substation into a
soil region saturated with groundwater. The
diameter of the wells is 50 mm. The
electrodes were constructed of 6 m segments
of 40 mm inside diameter 5 mm thick
galvanized steel tubes. The steel tubes have
many water percolation holes drilled in
them. Carbon powder was filled under
pressure between the well hole and the
outside of the steel tube to provide good
conductivity. After the deep-well grounds
were added, the grounding resistance
Maximum ground fault current
6 kA
Maximum fault current into grid
1.7 kA
Maximum fault clearing time
0.5 s
Surface layer resistivity
3000 Ω-m
Depth of ground grid
0.6 m
Conductor diameter
Ground grid area
5-16
0.0126 m
400 m x 400 m
Mesh dimensions
8mx8m
Mesh voltage
902 V
Total conductor length
15,840 m
Gas Insulated Substation Grounding Grid
Tolerable touch voltage
792 V
The design of the ground grid for a 400 kV
Gas Insulated Substation (GIS) in Greece in
presented (Georgantzis et al. 1998.) The
advantages of GIS substations are the
extremely small area they occupy compared
to conventional substations, their isolation
from environmental conditions and their
high reliability. The impact on ground grid
designs lies in the small area and in the
possibility of interference with control
systems during fault conditions.
Tolerable step voltage
2500 V
Depth of ground grid
0.6 m
Soil resistance measurements were taken
using the Wenner (four-probe) method, and
analyzed to provide the following results:
ρ2
resistivity of lower layer
60 Ω-m
Equivalent
68 Ω-m
13 m
soil
resistivity
References
1. ANSI/IEEE. 1983.
Standard 81.
IEEE Guide for Measuring Earth
Resistivity, Ground Impedance, and
Earth Surface Potentials of a Ground
System. New York, NY: IEEE.
The ground grid data is as follows:
Maximum ground fault current
40 kA
Maximum fault current into grid
5.4 kA
Maximum fault clearing time
1.0 s
Surface layer resistivity
2. EPRI. 1992a. Substation Grounding
Programs. Report no. TR-100622.
vol. 1-5. Palo Alto, CA
2000 Ω-m
Surface layer thickness
Conductor cross-section
2000 m
Excessive fault currents that enter the
grounding system in a substation have
several adverse impacts including physical
damage, reduced safety and mis-operation.
The
assessment
and
subsequent
refurbishment of grounding systems in
substations may be needed under excessive
fault current conditions.
resistivity of upper layer
160 Ω-m
ρE
Total conductor length
Summary and Recommendations
ρ1
boundary depth
8400 m2
From this data comes a calculated substation
resistance of 0.36 Ω and a GPR of 2 kV.
The mesh and step voltages are 422 and 478
volts, respectively, well within tolerable
limits. The IEEE Standard 80 design was
tested using a current injection test. It also
met the requirements of the European
Standard EN 50179 and the German DIN
VDE Standard 0141.
The 400 kV Lavrion substation is part of a
1200 MW power generation complex
occupying 400,000 m2, of which the
substation occupies 9000 m2.
h
Ground grid area
3. EPRI 1992b. Seasonal Variations of
Grounding Parameters by Field Tests.
Report. No. TR-100863. Palo Alto,
CA
0.3 m
220 mm2
Mesh conductor cross-section
Maximum shock current duration
150 mm
2
4. EPRI. 2004a. Guide for Transmission
Line Grounding: A Roadmap for
0.5 s
5-17
Design, Testing and Remediation.
Report no.1002021. Palo Alto, CA
Connections Used in Substation
Grounding. New York, NY: IEEE.
5. EPRI.
2004b.
Smart
Ground
Multimeter:
Enhancements,
Validation, Testing and Training.
Report no.1008683. Palo Alto, CA
13. IEEE. 2005. Task Force F0A of the
IEEE/PES Substations West Coast
Subcommittee.
“Current
North
American
Assessment
and
Refurbishment Practices of Substation
Grounding Systems. IEEE Trans. on
Power Delivery, Vol. 20, No. 3, pp.
1886-1889.
6. Georgantzis, G.J., N.G. Gagaoudakis,
and Th. Connor. 1998. “Design
Practice for the Earthing System of
the 400 kV Gas Insulated Switching
Station at Lavrion.” 8th International
Conference on Harmonics and
Quality of Power, ICHQP ’98,
Athens, Greece, 14-16 Oct.
14. Keil,
R.P.
2003.
“Substation
Grounding.” McDonald, J.D., editor,
Electric
Power
Substations
Engineering. Boca Raton, FL: CRC
Press.
7. He, J., G. Yu, J. Yuan, R. Zeng, B.
Zhang, J. Zou, and Z. Guan. 2005.
“Decreasing Grounding Resistance of
Substation by Deep-Ground-Well
Method.” IEEE Trans. on Power
Delivery, Vol. 20, No. 2, pp. 738-744.
15. Lawson, V.R. 1988. “Problems and
Detection of Line Anchor and
Substation Ground Grid Corrosion.”
IEEE
Trans.
on
Industry
Applications, Vol. 24, No. 1, pp. 2532.
8. IEEE. 1991a. Standard 81.2. IEEE
Guide for Measurement of Impedance
and Safety Characteristics of Large,
Extended
or
Interconnected
Grounding Systems. New York, NY:
IEEE.
16. Lee, C-H and C-N Chang. 2005.
“Comparison
of
161/69-kV
Grounding Grid Design Between
Indoor-Type
and
Outdoor-Type
Substations.” IEEE Trans. on Power
Delivery, Vol. 20, No. 2, pp. 13851393.
9. IEEE. 1991b. Standard 142. IEEE
Recommended
Practice
for
Grounding
of
Industrial
and
Commercial Power Systems. (IEEE
Green Book) New York, NY: IEEE.
17. Lee, C-H. 2004. “Safety Assessment
of Bulk and Traction Supply
Substations in Taipei Rail Transit
Systems.” IEEE Trans. on Power
Delivery, Vol. 19, No. 3, pp. 10781084.
10. IEEE. 1996. Standard 367. IEEE
Recommended
Practice
for
Determining the Electric Power
Station Ground Potential Rise and
Induced Voltage From a Power Fault.
New York, NY: IEEE.
18. Meliopoulos, A. P. S., Power System
Grounding and Transients, New
York, NY: Marcel Dekker. 1988.
19. Meliopoulos, A.P.S., S. Patel and C.J.
Cokkinides. 1994. “A New Method
and Instrument for Touch and Step
Voltage Measurements.” IEEE Trans.
on Power Delivery, Vol. 9, No. 4, pp.
1850-1860.
11. IEEE. 2000. Standard 80. IEEE
Guide for Safety in AC Substation
Grounding. New York, NY: IEEE.
12. IEEE. 2002. Standard 837. IEEE
Standard for Qualifying Permanent
5-18
20. Meng, Q., J. He, F.P. Dawalibi, and J.
Ma. 1999. “A New Method to
Decrease Ground Resistances of
Substation Grounding Systems in
High Resistivity Regions.” IEEE
Trans. on Power Delivery, Vol. 14,
No. 3, pp. 911-916.
of a HVDC System.” IEEE Trans. on
Power Delivery, Vol. 18, No. 3, pp.
867-873.
27. Villas, J.E.T. and C. M. Portela.
2003b. “Soil Heating Around the
Ground Electrode of an HVDC
System by Interaction of Electrical,
Thermal
and
Electroosmotic
Phenomena.” IEEE Trans. on Power
Delivery, Vol. 18, No. 3, pp. 874-881.
21. Natarajan, R., A.F. Imece, J. Popoff,
K. Agarwal, and P.S. Meliopoulos.
2001. “Analysis of Grounding
Systems for Electric Traction.” IEEE
Trans. on Power Delivery, Vol. 16,
No. 3, pp. 389-393.
28. Zotos, P.A. 1988. “Ground Grid
Design in Large Industrial Plants.”
IEEE
Trans.
on
Industry
Applications, Vol. 24, No. 3, pp. 521525.
22. Popovi , L.M. 2000. “Efficient
Reduction of Fault Current Through
the Grounding Grid of Substation
Supplied by Cable Line.” IEEE Trans.
on Power Delivery, Vol. 15, No. 2,
pp. 556-561.
23. van Waes, J., M. van Riet, F.
Provoost, and S. Cobben. 2003.
“Measurement of the Current
Distribution Near a Substation During
a Single Phase to Ground Fault.”
CIRED Barcelona 12-15 May.
Session 2.
24. Villas, J.E.T., J.A.A. Cassaagrande,
D. Mukhedkar, , and V.S. da Costa,
1988. “The Ground Grid Design of
the Barra do Peixe Substation Using a
Two-Layer Soil Model.” IEEE Trans.
on Power Delivery, Vol. 3, No. 4, pp.
1363-1641.
25. Villas, J.E.T., D. Mukhedkar, V.R.
Fernandes, and A.C. Magalhaes.
1990. “Ground Grid Design of a
Transition Station System – A
Typical Example of Fault Transfer.”
IEEE Trans. on Power Delivery, Vol.
5, No. 1, pp. 124-129.
26. Villas, J.E.T. and C. M. Portela.
2003a. “Calculation of Electric Field
and Potential Distributions Into Soil
and Air Media for a Ground Electrode
5-19
overhead lines and substation buses. The
major differences will be explained and
three major effects, namely, tension
increase, clearance problem and spacer
compression will be discussed. The CIGRE
brochures will be referred to. Further simple
formulation developed for the increase of
tension due to short circuit forces developed
by Lilien and Papailiou will be detailed and
its validation on short circuit tests will be
discussed. Also, the research on interphase
spacers loads due to short circuit forces will
be presented.
6
EFFECT OF HIGH
FAULT CURRENTS ON
TRANSMISSION LINES
This chapter is intended to supplement
the Increased Power Flow Guidebook
(EPRI. 2002), providing information on
the effects on transmission lines of the
increased fault currents which often
accompany increased power flows.
Effect of High Fault Current on NonCeramic Insulators
Three main types of NCI are used on
overhead transmission lines:
The standard reference works on
Transmission Lines such as the EPRI “Red
Book” (EPRI. 1975, 1982, 2004) and the
Westinghouse
Transmission
and
Distribution Book (Westinghouse. 1950.)
include very little data on short circuits, and
none on the effects of short circuit currents
on transmission lines themselves.
The
exception to this is the Compact Line
Design Reference Book (EPRI. 1978). To
compile data for this chapter, technical
papers have been referred to, and IEC
standards relating to flexible conductor
substation buses (covered elsewhere in this
report). Some information has been found
in an EPRI report on non-ceramic insulators
(NCI) (EPRI. 1998.).
• Suspension insulators
• Post Insulators
• Phase-to-phase insulators
Each of the above may be applied in
numerous ways:
•
•
EPRI is also sponsoring development of a
new edition of the Transmission Line
Reference Book: Wind-Induced Conductor
Motion, commonly known as the “Orange
Book,” which was originally published in
1979 (EPRI 2005). In this book, one of the
chapters will cover transient motions, which
include short circuit forces, bundle rolling,
ice drop, gust response, and wind action on
members. A section in this chapter will deal
with the impact of short circuit forces on
•
6-1
In general suspension insulators are
intended primarily to carry tension
loads. Suspension insulators can be
applied in I- string, Vee-string and
dead-end applications
Post insulators are intended to be
loaded in tension, bending or
compression. The most common
application is horizontal posts. These
post insulators may be applied either
alone or together with a suspension
insulator in a braced post
configuration. Post insulators are
also applied in substations as bus
support or in disconnect switch
applications.
Phase–to-phase
insulators
are
intended to be loaded in tension,
torsion, bending or compression.
Phase to phase insulators couple two
phases together in order to control
conductor spacing during galloping.
This document deals mainly with suspension
insulators although many areas may be
relevant to both post and phase-to-phase
insulators.
during a power arc the galvanization of the
end fitting may be damaged, making the
fitting susceptible to corrosion. The longterm effects of localized heating on the end
fitting, FRP rod and weathershed system due
to power arcs still require further research.
Modern suspension and strain insulators
generally consist of the following main
elements:
If a flashover occurred across the insulator
because a transient event such as lightning,
the rubber weather-shed system or FRP rod
may or may not have sustained damage. If
the weather-shed system or FRP rod has
been damaged, that damage is usually
obvious, and the insulator should be
removed (see Figure 6-2). In some cases the
flashover may have caused no obvious
damage to the insulator apart from areas of
degalvanization of the end fitting hardware.
This may be especially true if the power arc
terminated on the grading rings. Insulators
whose external appearance indicates they
have only sustained this type of damage are
still of concern. Due to the nature of grading
ring attachments and end fitting design,
power fault currents usually flow through
the end fitting, which is in direct contact
with the FRP and rubber weather-shed
system. Whether the increased end fitting
temperatures associated with the power arc
current result in long-term degradation of
either the polymeric rubber material or the
FRP rod is unknown. Hence, it is advisable
to remove such insulators from service.
•
•
Energized metallic end fitting
Energized end grading ring (need
depends on application)
• Fiberglass reinforced plastic rod
(FRP)
• Polymeric weather shed system,
consisting of weather sheds and
sheath
• Grounded end grading ring (need
depends on application)
• Grounded metallic end fitting
Grading rings, also called corona rings, are
not installed at all voltage levels or on all
applications, and often only a single grading
ring is installed at the energized end. The
method of construction and materials used
depends on the manufacturer and
application.
Concern has been raised over end fitting
damage caused by fault currents flowing
during a flashover.
Although
tests have Concern has been raised over
end fitting damage caused by
indicated
that the fault currents flowing during
a flashover.
critical
tensile
strength of an NCI may be reduced to 80%
of its specified mechanical load (SML)
rating during a power arc test, the insulator’s
tensile strength recovers to a level above its
SML after the test. Since NCI are usually
applied at less than 50% of their SML, this
is not a significant concern. However,
6-2
Conductor Motion Due to Fault Currents
While
normal
transmission
line
construction, with widely separated phases,
does not appear to be significantly impacted
by conductor motion due to fault currents, it
is an important consideration for compact
transmission lines. Two parallel conductors,
each carrying current, will be subject to a
force of attraction or repulsion, depending
on current direction. The magnitude of the
force on each conductor is:
F=
µ0 I 2 l
2π d
6-1
Where:
I
= current in each conductor.
l
= length of each conductor.
d
= distance between conductors.
For short
If the current flow in each
circuit
conductor is in the same
currents,
these forces direction, the force will
may
be cause attraction; and,
sufficient to conversely, if the current is
in opposite directions, the
cause
force will cause repulsion.
significant
conductor
movement, particularly where conductors
are closely spaced (such as in EHV
conductor bundles or in adjacent phases of a
compact line), because of the inverse
relationship of conductor spacing and the
resultant force. The actual movement of a
conductor, considering inertia, is a function
of both the magnitude of the current and the
time it is applied, and is therefore dependent
on circuit breaker interrupting time.
Figure 6-1
Cross Section of a Typical Transmission Non
Ceramic Insulator (EPRI 1998)
Figure 6-2
Flashover Damage Sustained by an NCI
(EPRI 1998)
A phase-to-phase fault will cause current in
the two affected phases to flow in opposite
directions. The two conductors will then be
6-3
conductors can be shown to give results well
within line design accuracy requirements.
repelled and, on interruption of the fault
current, will swing together.
If the current
is due to a Even though such currents
fault on the on the compact line may
be less than the maximum
line
section
fault currents attainable
under
on the system, they may be
consideration,
sufficient
to
be
the electrical
determining
in
the
consequences
selection of phase-toof conductor
phase spacing or in
motion (even establishing the need for
clashing) are insulating spacers.
generally
unimportant
since that section will be tripped out
anyway. However, if the fault is on an
adjacent line section, the motion may be
serious since it might cause interruption of
the unfaulted section.
“Through-fault”
currents, or those supplied through a
unfaulted line to a fault elsewhere on the
system, can be an important design
consideration
for
compact
circuits.
Calculation of Fault Current Motion for
Horizontally Spaced Conductors
Figure 6-3 illustrates the conductor
configuration used as a basis for
calculations. It is assumed that the forces to
which each catenary span of the conductor is
subjected will cause the span to swing in a
plane, as shown in (a) in Figure 6-3. The
plan projection of each catenary is then also
a catenary. This assumption, supported by
experimental
results,
simplifies
the
calculation technique. The most severe fault
is phase-to-phase on adjacent phases, which
impresses
a
cyclic
separating
electromagnetic force.
Since all spans of a line contributing to a
through-fault will behave similarly, the net
pole-top force along the span, and therefore
motion, will be zero. Consequently, each
span can be assumed to be rigidly
terminated.
Effect of Conductor Shape
The true shape of each conductor is a
catenary, as shown in Figure 6-4. For a
catenary, y = k cosh (x k ) − 1 .
[
]
At x = S 2 , where S is the span length, (i.e. at
the conductor support), let y = y 0 , where y0
is the conductor sag, to define k.
Figure 6-3
Horizontal Conductor Motion During
Through-Fault (EPRI 1978)
⎡
⎛ S ⎞ ⎤
y 0 = k ⎢cosh⎜ ⎟ − 1⎥
⎝ 2k ⎠ ⎦
⎣
S2
=
8k
The motion of conductors subjected to
electromagnetic forces is similar to that of
weighted, stretched strings, with the
complication that the string is usually a
compound conductor (ACSR). Relatively
simple analyses of conductor motion of both
vertically
and
horizontally
spaced
Eq. 6-2
(For detailed derivations see (EPRI. 1978.)
or
k=
6-4
S2
8 y0
Eq. 6-3
d0 = initial conductor spacing.
yt = conductor sag at time t.
then the average force on one conductor is:
Favg =
µ0
I 2l
2
2π
d 0 + y0
Eq. 6-8
3
Figure 6-4
Conductor Geometry (EPRI 1978)
The conductor mass is modeled as a
pendulum, which swings at a distance
2
y 0 below
3
the support points.
Conductor Equations of Motion
From the previous equations, the conductors
are represented in Figure 6-5, where:
F = electromagnetic force of repulsion.
Figure 6-5
Forces on Conductor (EPRI 1978)
W = conductor weight.
And
yx =
2
S
8 y0
y avg =
⎡
⎛ 8 y0 x ⎞ ⎤
⎢cosh⎜ 2 ⎟ − 1⎥
⎝ S ⎠ ⎦
⎣
S
4 y0
⎡ S2
⎛ 4y ⎞ S ⎤
sinh⎜ 0 ⎟ − ⎥
⎢
8
y
⎝ S ⎠ 2⎦
⎣ 0
Eq. 6-4
2
y0
3
Eq. 6-5
at
Pt = actual conductor
midspan, time t.
position
at
Fc = normal component of conductor
tension.
Resolving
tangentially,
the
accelerating conductor swing is:
Ftan g = F cos θ − W sin θ
Eq. 6-6
force
Eq. 6-9
The conductor acceleration point Q is then:
is an almost exact solution
⎛g⎞
Q tan g = Ftan g ⎜ ⎟
⎝W ⎠
Since the force between two parallel
conductors is given by (8.3-1), and
d t = d 0 + 2 yt
position
Q = effective position of mass and
forces.
By substitution in the above equation, it can
be shown that for transmission lines
y avg =
P0 = actual conductor
midspan when at rest.
Eq. 6-10
where g is the gravitational constant.
Eq. 6-7
Using a step-by-step analysis, the conductor
velocity
Where:
dt = conductor spacing at time t.
6-5
v tan g (t ) = v (t −δt ) +
a (t ) − a (t −δt )
2
Calculation of Fault Current Motion for
Vertically Spaced Conductors
Eq. 6-11
dt
As a conductor span moves upward due to
fault current forces, the tension is reduced
and the acceleration is restrained by the
increase in the effective conductor weight.
Conversely, as a conductor moves
downward due to these forces, the
acceleration is inhibited by an increase in
conductor tension. Because of these effects,
it is important that the modulus of elasticity
be considered in calculations.
then
θ (t ) = θ (t −δt ) +
v tan g (t ) + v tan g (t −δt )
2r
dt Eq. 6-12
where
2
y0
3
r=
and
the
true
horizontal
displacement of point P is then
z t = y 0 sin θ
Eq. 6-13
midspan
Eq. 6-14
Effect of Conductor Stretch
As the conductor deflects under load, the
effective weight per unit length changes.
Resolving perpendicular to the conductor in
the conductor plane, using the terminology
of Figure 6-5:
Fc = W cos θ + F sin θ
Figure 6-6
Typical Vertical Conductor Arrangement
(EPRI 1978)
Eq. 6-15
If θ = 0 (i.e. conductor in vertical rest
position),
Fc = W
Figure 6-7
Vertical Displacement during Fault (EPRI
1978)
Eq. 6-16
and
WS 2
sag = y0 ≅
8H
Eq. 6-17
Where:
H = conductor tension.
If θ ≠ 0 , then:
Fc S 2
8H
W cos θ + F sin θ 2
S
=
8H
y0 =
Eq. 6-18
Figure 6-8
Conductor Angle at Support (EPRI 1978)
6-6
Figure 6-6 illustrates a typical vertical
conductor arrangement. As a simplifying
approximation, it is assumed that the forces
to which each conductor is subjected will
cause an increase or reduction in sag, but
that the conductor will retain a catenary
shape. This assumption is supported by
experimental results for low currents applied
for long durations. The assumption is even
more accurate for high fault current levels
and short durations, where most of the
kinetic energy is imparted to the conductor
before the conductor can move appreciably.
acceleration and the period of oscillation of
each conductor need not be the same.
Figure 6-8 illustrates the basis on which the
angle of the conductor at the support is
calculated.
For the conductor catenary:
y=
∂y
Wx
= sinh
∂x
H
∂y
WS
4D
= sinh
= sinh
∂x
S
2H
From the rest position, D0 = D1 = D2 , i.e. all
sags are equal. For any other position,
assuming both conductors are a catenary, the
average separation distance can be
expressed:
Eq. 6-19
L2 − L1 =
H1 − H 2
L
aE
Eq. 6-26
or
Eq. 6-20
δH =
δL
L
aE
Eq. 6-27
Where the conductor length, L, can be
approximated as:
L=S+
Eq. 6-21
8D 2
3S
Eq. 6-28
δL can be expressed as:
and for the top conductor:
Fnet = 2 H 1 sin θ1 − S (W − F )
Eq. 6-25
If an initial conductor length L1 and an initial
conductor tension T1 are assumed, then for
any subsequent motion resulting in L2 and
T2:
Note that the electromagnetic force has the
effect of changing the effective conductor
weight. The net accelerating force on each
span of the bottom conductor is:
Fnet = 2 H 2 sin θ 2 − S (W + F )
Eq. 6-24
Calculation of Tension Change With Motion
so that the average electromagnetic force is:
µ I 2l
F= 0
2π d avg
Eq. 6-23
At x = S 2 (i.e. at the conductor support),
Calculation Procedure
2
(D2 − D1 )
3
⎡
⎛ Wx ⎞ ⎤
⎢cosh⎜ H ⎟ + 1⎥
⎝
⎠ ⎦
⎣
i.e.,
The terminology used in analyzing the
vertical case is the same as for the horizontal
case. The configuration used as a basis for
calculations is illustrated in Figure 6-7.
d avg = D0 +
H
W
Eq. 6-22
δL =
Using Fnet = ma , conductor motion can be
expressed as a function of time. The rates of
8D22 8D12
8
(δD − 2 D1 )
−
=
3S
3S
3S
Where:
6-7
Eq. 6-29
δD = D1 − D2
Eq. 6-30
D=
The change in tension can be expressed in
terms of the change of vertical displacement
as follows:
WS 2
8H
Eq. 6-37
the maximum spacer force is
2
8aE
δH =
δD (δD − 2 D1 )
3SL1
Fspacer = 16 H
Eq. 6-31
this will be
both tension
case where
maximum
occurred.
Calculation of Mechanical Loading on
Phase-to-Phase Spacers
The electromagnetic forces during a phaseto-phase fault will act to move the
conductors apart, placing phase-to-phase
spacers in tension. After the fault is cleared,
the conductors will swing together,
compressing the spacers. These forces can
be analyzed using the diagram of
Figure 6-9. Using the previously defined
terminology, for any subspan swing angle,
θ,
Fc = F sin θ + W cos θ
Eq. 6-32
Fspacer = 2 Fc cos θ
= 2(F sin θ + W cos θ )sin θ
Eq. 6-33
Effect of Bundle Pinch on Conductors
and Spacers
Transmission line spacers, Figure 2-6, are
designed to withstand the compressive force
on bundled conductors caused by short
circuit forces. Spacer compression may be
calculated with the Manuzio formula (Lilien
et al. 2000.), originally for flexible bus
substation design:
Eq. 6-34
⎛z ⎞
cos θ = 1 − ⎜ t ⎟
⎝D⎠
2
Eq. 6-35
In the simple case where Fout reduces to zero
before maximum swing is reached (i.e. the
fault clears):
Fspacer = 2 SW
zt
⎛z ⎞
1− ⎜ t ⎟
D
⎝D⎠
the maximum spacer force in
and compression for the usual
the fault has cleared before
conductor
deflection
has
Figure 6-9
Derivation of Forces on Spacers (EPRI 1978)
and
z
sin θ = t
D
16 Hz t
zt
⎛z ⎞
1− ⎜ t ⎟ ≅
S
S
⎝D⎠
⎛a
Pmax = 1.45 I ′′ Fst log⎜⎜ S
⎝ dS
⎞
⎟⎟
⎠
Eq. 6-39
Where:
2
Eq. 6-36
Pmax =
Compression force on the
spacer in N.
Since
Fst =
Initial static tension on the
conductor bundle in N.
6-8
as =
ds =
Conductor spacing in mm.
2 F pi
Fc =
1 + (l nc (a s − d s ))2
Conductor diameter in mm.
Eq. 6-41
and
Tests by Lillien, et al., showed that the
Manuzio formula underestimated the stress
by 50%. Better results (within ±10%) were
obtained using finite element analysis. These
results were extended (Lilien and Papaliou.
2000) to transmission line design. The
Manuzio approach neglects:
Fc =
µ0 2
I
2π
l nc
∫
0
cos(ϑ (x ))
dx
2 y (x )
Eq. 6-42
where
1. The pinch effect, which results in an
increase
in
tension
of
the
subconductors during the short
circuit.
cos(ϑ ) =
1
Eq. 6-43
1 + (dy dx )2
2. The asymmetry of the fault current.
and
ϑ
is the deviation from the
horizontal (Figure 6-11).
3. The length of subspan between
spacers, l S
I
is the time-average short
circuit current (CIGRE. 1996)
All of these causes result in higher
compression forces on spacers during faults.
Lilien and Papaliou stress that the Manuzio
formula should no longer be used, especially
as fault current levels are increasing in
transmission systems. After an extensive
series of tests and computer simulations,
they recommend a new calculation method
based on the IEC 865-1 approach (IEC.
1993.) The subconductor, Figure 6-11, is
assumed to take a parabolic shape between
the spacer and the point where the
subconductors all touch:
y (x ) =
as − d s
2
⎛ x
⋅ ⎜⎜
⎝ l nc
2
⎞
⎛ x
⎟⎟ − (a s − d s ) ⋅ ⎜⎜
⎠
⎝ l nc
Solution of these equations results in values
for FC, the compression force on the spacer
and the l nc length of subconductor before the
pinch occurs.
Finite element method
simulations can also be used for this
problem (Kruse and Pearce. 2000.)
Spacers
ds
⎞ as
⎟⎟ +
⎠ 2
as
ls
Eq. 6-40
n sub-conductors
Where, l nc is the non-contact length, which
must be calculated.
Figure 6-10
Details of Transmission Line Conductor
Bundle With Spacers
Simultaneous numerical solution of the
following two equations is required:
6-9
8. EPRI 2005. Updating the EPRI
Transmission Line Reference Book:
Wind-induced Conductor Motion
(“The Orange Book”). Progress
Report, Palo Alto, CA EPRI. 1010223
9. IEC. 1993. International Standard
865-1: 1993. Short-circuit currents—
Calculation of effects. Part 1:
Definitions and calculation methods.
Genève: CEI.
Figure 6-11
Parabolic Model of Subconductor Pinch
Forces (Lilien and Papaliou 2000)
10. Kruse, G. C. and H. T. Pearce, 2000.
“The Finite Element Simulation of
Bundle Pinch of a Transmission Line
Conductor
Bundle,”
IEEE
Transactions on Power Delivery, Vol.
15, No. 1, January 2000, 216-221.
References
1. CIGRE. 1996. “The mechanical
effects of short-circuit currents in
open-air substations (rigid and
flexible bus-bars),” CIGRE, Paris,
CIGRE brochure no. 105, vol. 1 and
2, Apr. 1996.
11. Landry, M., R. Beauchemin and A.
Venne, 2000. “De-Icing EHV
Overhead Transmission Lines using
Electromagnetic Forces Generated by
Moderate Short-Circuit Currents,”
ESMO –2000 IEEE 9th International
Conference on Transmission and
Distribution Construction, Operation
and Live-Line Maintenance. pp. 94100.
2. EPRI 1978. Transmission Line
Reference Book: 115-138 kV
Compact Line Design, Palo Alto, CA:
EPRI. EL-100-V3.
3. EPRI. 1975. Transmission Line
Reference Book 345 kV and Above,
Palo Alto, CA: EPRI. EL-2500.
12. Lilien, J-L. and Papailiou, 2000.
“Calculation of Spacer Compression
for Bundle Lines Under ShortCircuit,” IEEE Transactions on
Power Delivery, Vol. 15, No. 2, April
2000, pp. 839-845.
4. EPRI. 1982. Transmission Line
Reference Book: 345 kV and Above-Second Edition, Revised, Palo Alto,
CA: EPRI. EL-2500-R1.
5. EPRI. 1998. Application Guide for
Transmission Line Non-Ceramic
Insulators. Palo Alto, CA: EPRI. TR111566.
13. Lilien, J-L., E. Hansenne, K. O.
Papailiou, and J. Kempf, 2000.
“Spacer Compression for a Triple
Conductor
Bundle,”
IEEE
Transactions on Power Delivery, Vol.
15, No. 1, January 2000, pp. 236-241.
6. EPRI. 2002. Increased Power Flow
Guidebook -- Overhead Transmission
Lines, Palo Alto, CA: EPRI. 1001817.
14. Miroshnik,
R.,
2000.
“The
probabilistic model of the dynamic of
the cables under short-circuit
current,” Computer Methods in
Applied Mechanics and Engineering,
Vol. 187/1-2, p. 201-211 (July 2000).
7. EPRI. 2004. EPRI AC Transmission
Line Reference Book - 200 kV and
Above, Third Edition, Palo Alto, CA:
EPRI. 1008742.
6-10
15. Westinghouse,
1950.
Electrical
Transmission
and
Distribution
Reference Book, Fourth Edition, East
Pittsburgh,
PA:
Westinghouse
Electric Corporation.
6-11
are bent beyond their elastic limit but
no electrical fault occurs.
• The second occurs when insulation is
damaged by the buckled conductors.
This may lead to a turn-to-turn or a
turn-to-ground contact resulting in
the internal short-circuit. In this
scenario, the current in the windings
can be considerably greater than the
winding current caused by a through
fault current. The result will be
serious damage with the burning of
insulation and melting of conductors.
Typically, a power transformer can
withstand around three full short circuits in
its terminals.
7
SHORT CIRCUIT
FORCES IN
TRANSFORMERS
The effects of short-circuit currents in
transmission and distribution networks for
electric energy are great, both for the
equipment and the networks. Mechanical
forces are produced in transformer windings
whenever a current is flowing in them. If
excessive, such forces can distort
transformer windings, or cause other
physical damage.
Typical Values of Mechanical Forces in
Transformers
Calculating
short-circuit
forces in a
transformer
using
analytical
methods is
very
difficult.
This section will deal with the effects of
short circuit forces on transformers, which
can lead sometimes to the mechanical
damage of the transformers, and sometimes
fires which will render the transformers
useless without a rebuild and rewind.
With an
Increased short circuit
increase
of
the currents create concerns to
transformers,
especially
connectiv
ity of the when they are aged.
different
grids, and the increases in the short-circuit
power in them, the short circuit currents
seen by the transformers have increased.
This creates a concern to transformers,
especially when they are aged. These shortcircuit currents have to be withstand without
impairing the transformer.
For a two-winding transformer carrying
normal load current, the current in the
primary winding flows in a direction
opposite to the direction of the current in the
secondary winding, and the total ampereturns in the primary are equal to and
opposite from the total ampere-turns in the
secondary. Because there is mutual stray
flux between the windings, the primary and
secondary windings tend to repel each other.
The repulsive forces are greatly magnified
for the transformer under fault conditions,
(greater than normal load currents by up to
as much one or two orders of magnitude
times).
These forces may distort the winding
conductors and lead to two kinds of failure.
•
Even though that normally
the duration of fault current
is short to cause thermal
damage to a transformer,
the
corresponding
d
mechanical forces could be
of damaging nature for
large power transformers.
The first is a purely mechanical
event in which winding conductors
The mechanical forces in a system are
proportional to the square of the peak
7-1
current, hence, the peak let-through current
may be important. For the typical case of
through fault current of about 25 times
normal rated current, and
since the
mechanical forces in the windings increase
proportionally to the square of the current,
the through fault mechanical forces can
become more than 600 times the forces at
rated current.
Transformer Categories
Category
Single-phase
(kVA)
Three-phase
(kVA)
I
5 to 500
15 to 500
II
501 to 1667
501 to 5000
III
1668 to 10000
5001 to 30000
IV
Above 10000
Above 30000
The design of a power transformer with
respect to the short-circuit current withstand
capability is centered on the control of the
forces inside the transformer. A low
impedance transformer, sometimes needed
for voltage concerns, will have a higher
through short circuit current compared to
“normal” impedance transformer. The low
impedance transformer will hence require a
more stringent mechanical design than it
normal counterpart.
Short Circuit Currents in a Transformer
For
short
Transformers, like series
circuit
reactors, have the quality
calculations,
to limit the short-circuit
the
currents
to
values
equivalent
predominantly
determined
circuit of a
by
the
transformer's
transformer
is basically impedance
the leakage
impedance as determined by factory tests.
Hence, it is relatively easy to calculate the
short-circuit currents resulting from various
types of short circuits or faults. One type of
fault is usually more severe than the others
in terms of generating damaging currents.
This is called the limiting fault, because
fault current can go no higher for any other
fault. The limiting fault depends on the
transformer
connection
and
system
impedances.
Effect of the System Impedance on Short
Circuit Currents
Although the transformer's characteristics by
itself are the parameters dominating the
amplitudes of the short-circuit currents,
some network parameters (such as stiffness
of the system), system Xs/Rs ratio, and
network conditions have to be taken into
consideration as well.
More important than the systems Xs/Rs ratio
is the transformers Xtr/Rtr ratio. The overall
time constant can be based on
(Xs+Xtr)/(Rs+Rtr). As modern transformers
are optimized for the no-load and load
losses, the copper losses are very low and
so, by design, is the resistance Rtr, this has
the effect of also increasing the fault current.
The Xtr/Rtr ratio can reach value of 40 or 50
or more, especially for large transformers.
When the value of Rtr is relatively low the
influence of Rs can not be neglected in the
simple calculations
The general short-circuit requirements for
liquid-immersed distribution, power, and
regulating transformers are specified in
Section 7 of IEEE Std. C57.12.00-2000.
Transformers may be categorized depending
on their rating as shown in Table 7-1. For
Category I and II transformers, symmetrical
short circuit current shall be computed using
transformer impedance only. For Category
III and IV transformers, system impedance
should also be considered along with
transformer impedance in order to compute
symmetric short circuit currents.
Table 7-1
7-2
Short Circuit Tests of Power Transformers
Through fault Currents in a Transformer
During their life, power transformers have to
face several short circuit conditions: many
small short-circuit currents, and hopefully a
smaller number of large ones. The
transformers have to withstand these short
circuits events without impairing the
transformer.
Through fault
The mechanical effects of
currents seen
through fault currents are
by
the cumulative. The extent of
transformer
the damage is dependent on
have thermal the fault frequency, fault
and
duration
and
most
mechanical
importantly magnitude of
impacts. For fault currents.
low
fault
currents, thermal effects are more
predominant but high fault currents near the
design capability of transformer have more
significant mechanical impacts. Through
fault capability of liquid-immersed and drytype transformers is presented in IEEE
standards C57.109-1993 and C57.12.592001 respectively.
Real tests
The short-circuit withstand
are
a
capability is defined as the
prerequisite
ability of the transformer to
for
withstand
several
full
transformer
asymmetrical short-circuit
manufacture currents in each phase and
rs
to in each representative tap
understand
position without impairing
withstand of the transformer suitability
the
for
normal
service
transformer
conditions.
and
to
improve the
calculation and simulation techniques, to
verify construction details and to experience
the difference between practice and theory.
Liquid-Immersed Transformers
ANSI C57.92-1962 provided the thermal
load capability of such transformers but it
did not account for the mechanical impacts.
Table 7-2
Transformer short-time Thermal Load
Capability
Normally transformers are tested in the
nominal tap position, in the highest tap
position (maximum voltage) and in the
lowest tap position (minimum voltage). In a
well designed system, and proper
transformer application, the probability of
reaching the extreme tap positions in service
are rather small and, if so, then the shortcircuit power in the network is small due to
either a situation with very light loads
(leading to extreme high network voltages)
or a situation with a lag of apparent power
(leading to extreme low network voltages).
The Standards (IEEE Std C57.12.00-1993,
IEEE Std C57.12.90-1993, Part I and II)
dictate that the short-circuit withstand
capability in the extreme tap positions is
specified with a network voltage equal to the
rated voltage of the tap under consideration
and with a short-circuit power available in
the network equal to the power specified for
the nominal tap.
Time
Times Rated Current
2s
25
10 s
11.3
30 s
6.3
60 s
4.75
5 min
3.0
30 min
2.0
Through faults in power transformers result
in impact forces that result in compression
and wear of insulation and friction-induced
displacement in the windings. These effects
are cumulative and should be considered
over the life of the transformer. Throughfault capability limit curves that takes into
consideration both thermal and mechanical
damages are shown here. For category I
7-3
transformer, I2t limit of 1250 is applicable
for fault currents in the range of 25-40 times
the transformer base current (See Figure
7-1). Through-fault capability limit curve for
Category II and III transformers that
experience infrequent faults (not exceeding
10 for Category II and 5 for Category III) is
shown in Figure 7-2. This curve is limited to
2 seconds. For transformers that experience
frequent
faults,
standard
considers
mechanical duty of fault currents higher than
70% of maximum fault current for category
II transformers and those of higher than 50%
for category III and IV transformers. I2t
curve for these transformers is calculated
based on the transformer impedance. For
example, the curve for category II
transformer having 7% impedance is shown
in Figure 7-3.
Figure 7-2
Through-fault Capability Limit Curve for
Category II and III Transformers with
Infrequent Faults (SEL 2005)
Figure 7-1
Through-fault Capability Limit Curve for
Category I Transformers (SEL 2005)
Figure 7-3
Through-fault Protection Curve for Category
II Transformers – Frequent Faults (SEL 2005)
7-4
Dry-Type Transformers
Through-fault protection curve that takes
into consideration both thermal and
mechanical damages for Category I and
Category II transformers is shown in Figure
7-4 and Figure 7-5 respectively. No
information has been provided for category
III transformers as they are not commonly
manufactured.
Figure 7-5
Through-fault Protection Curve for Category
II Transformers (IEEE Std C57.12.59-2001)
Protection Considerations
Based on the transformer type (Liquidimmersed or dry) and its category, the
protection with proper time-overcurrent
characteristic needs to be selected and
coordinated with the through fault capability
limit curve. An example of the TOC
coordination with through fault capability
limit curve of a Category IV transformer is
shown in Figure 7-6 (SEL 2005).
Figure 7-4
Through-fault Protection Curve for Category
I Transformers (IEEE Std C57.12.59-2001)
The logic that may be used for monitoring
and cumulative recording of the through
fault currents is shown in Figure 7-7 (SEL
2005) .The recorded values can be compared
against threshold values to trigger an alarm
that can be used for scheduling maintenance
and testing to check for any potential
mechanical damage.
7-5
Mechanical Forces in Transformers
The sections to follow are mostly
extracted from the Help file of the
EPRI
Transformer
Expert
Program, XVisor. [29]
The short-circuit forces can be broken down
into their
1. Radial components
2. Axial components
Through fault forces, or short circuit forces,
(both terms are used interchangeably
throughout this document) result from faults
external to a transformer. Faults in a power
system can cause currents substantially
greater than rated currents to flow through a
unit. Such currents, which are asymmetric
and can be large in magnitude, produce
significant mechanical forces within a
transformer. Transformer manufacturers
must be aware of the forces, and their
locations
throughout
a
transformer.
Provided with this information, the designer
must provide the mechanical construction
details to successfully withstand these forces
during the course of long-term operation.
Figure 7-6
TOC Coordination for a Category IV
Transformer (SEL 2005)
The mechanical forces within a unit are the
product of the interaction of current flowing
in the windings, its geometry, and the
magnetic field in which they are located. In
all situations, the forces produced can be
calculated based on the work of Ampere and
Biot Savart. Their experimental work,
expressed in vector form, determined that
the differential force (dF) on a differential
current element (Idl) was equal to the cross
product of the current and the magnetic flux
density (B).
Figure 7-7
Cumulative Through Fault Logic (SEL 2005)
Eq. 7-1
7-6
Current does not flow in differential
elements as indicated in the equation above,
but in a complete current path or loop.
Therefore, a more useful expression can be
obtained by integrating this expression
around a closed loop yielding the following.
Eq. 7-3
The calculation of the force of interest
requires knowledge of both the current
flowing and the magnetic flux density. The
current is readily determined from the short
circuit conditions of the system, and its peak
value is used to determine the maximum
force. The magnetic flux density is typically
more difficult to identify, but can in general
be calculated from Maxwell's equation,
Eq. 7-2
The force calculated will always be
perpendicular to the plane formed by the
current and magnetic field vectors.
A mnemonic that is frequently used to
represent the relationship between these
vectors is the left hand rule. The fingers,
represented in the order of the equation
above, are used to point in the direction of
their representative vector. The thumb
points in the direction of the force; the
forefinger points in the direction of the
magnetic field (perpendicular to the thumb);
and the middle finger points in the direction
of the current (perpendicular to the thumb
and forefinger pointing toward the inside of
the hand).
Eq. 7-4
and the relationship between magnetic flux
density and magnetic field strength.
Eq. 7-5
It is important to note that the magnetic flux
density is also a function of current;
therefore, calculations of force will always
be proportional to the product of the current
squared. Proportional to the current
squared, mechanical forces increase rapidly.
A doubling of the current increases the
forces by a factor of four.
Eq. 7-6
This relationship applies for all calculations
needed to assess short circuit forces in
transformers. The most difficult challenge
is to calculate the magnetic field at all the
points of interest. In a large percentage of
cases, it can be approximated, but in others,
finite element analysis will be necessary.
Although this device is useful, and many
times needed to understand conditions at a
particular location in a unit, it is frequently
easier to remember that axial flux will
produce radial force, and radial flux will
produce axial force.
7-7
If the movement is significant or persists,
the winding will loosen, the core will
provide inadequate support, and the unit will
fail.
Critical Forces in Shell Form
Transformers
Axial Forces
In a shell form transformer the core is an
integral part of the structure designed to
resist short circuit forces.
The predominant forces in a shell form
transformer are axial resulting from the main
radial flux. Geometrically the main radial
flux is parallel to the pancake windings;
therefore, the axial forces due to this flux are
perpendicular to the pancake windings.
Figure 7-8
Generated Forces in a Shell Form
Transformer
The axial forces will compress the windings
and produce large forces on the core and end
walls. The high and low voltage windings
will repel each other, and the highest axial
forces will be in the coils adjacent to the
high-low voltage space. The forces will be
strongest within the core window as the flux
density is higher as to be expected.
Investigation of shell form short circuit
failure mechanisms that are generated by
axial forces should consider the following:
support of coil/pancake edges, beam
bending of conductors between spacers,
winding bending, core movement/end
support collapse, hydraulic pumping and
conductor tipping.
Radial forces
Radial forces are not a major concern in
shell from transformers due to the minimal
axial flux present (perpendicular to the
pancake windings). Radial forces will have
a tendency to compress a pancake winding
into a tighter radius, and in infrequent cases
vise a versa, splitting or pulling the winding
apart.
Investigation of radial forces should
consider inward forces, which will typically
be strongest at the corners of the pancake
winding, and should also look for any
unusual flux patterns which could generate
radial outward forces for instance in tapped
regions.
During short circuit conditions, the core legs
are in tension in the axial direction, and the
core yoke is loaded as a beam in bending.
Constructed of numerous steel laminations
the core is an effective structural member, as
long as it acts as a homogeneous member.
To prevent movement and looseness, the
laminations and associated core joints, have
to be tightly assembled and clamped.
Shell-Form Failure Mechanisms
Although
not
all The failure mechanisms
resulting from short circuit
inclusive,
forces are somewhat unique
the
for core-form and shell-form
following
i
is
a
prioritized list of failure modes most
frequently encountered in shell-form
transformer designs. The prioritized list is
Use of core joint overlap and clamping in
the corners is essential. Inadequate clamping
can lead to core joint opening, bending of
the laminations, and in general movement of
the core.
7-8
adequate insulation strength to ground. As a
result, the bracing clamping and blocking
design must adequately transfer forces
across the windings to core and tank
structure.
in rank order of occurrence, i.e., inadequate
support of coil ends is most likely to occur.
Inadequate Support of Coil Ends
In most shell form units, the diameters of the
pancake winding assemblies differ across a
winding. The first pancake winding in an
HV winding
Loose
windings
or
is usually the
conductors
precipitate
smallest, so
failures in all types of
that adequate
transformers, due to their
insulation can
excessive
movement
be provided
during fault conditions.
from
that
winding
to
ground. Inherent in the design of a shell
form unit is the difference in diameter of
pancake coils because of the need to supply
adequate insulation to ground.
Moreover, in these units, it is critical that the
tightness of the winding is maintained both
inside and outside the core window. The
winding structure within the window can be
stiffer due to the presence of the large mass
of core to support the windings, but requires
attention to assembly detail within the core
window to maintain clamping pressure.
However, outside the core windows the
assembly of pancake windings is supported
by the tank structure alone. The tank
structure, which is much less stiff than that
of the core, requires additional design
considerations to resist short circuit forces.
A comparison of the support structures
found the core assembly to be approximately
40 times stiffer than the tank assembly,
however, it is also dependent on tank end
structures to avoid movement [17].
During fault conditions, mechanical forces
may be such that the coil ends of a larger
diameter pancake coil are forced toward a
smaller one. If the space beyond the outer
diameter of the smaller disk is not
constructed with adequate support strength,
or the collars over the ends of the larger
pancake winding are inadequate, the larger
winding will bend over the smaller one. The
bending will lead to the rupture of
insulation, and an attendant dielectric
failure.
The outer turns of a loose winding can
literally be pushed off the edge the winding
package. Windings must be tight so that the
short circuit forces are transmitted to the
clamping structure, preventing extreme
movement of the windings. When this is not
accomplished, mechanical damage can lead
to the collapse of a winding or to the
bending or tipping of conductors. The
resulting turn-to-turn or section-to-section
failures will compromise the dielectric
integrity of the unit with catastrophic results.
A subset failure mechanism of loose
windings in shell type transformers is
hydraulic pumping as described herein.
Loose Winding/Conductors
The design
of pancake The design of pancake
windings for windings for a shell form
a shell form transformer is subject to
the voltage and capacity
transformer
is subject to ratings specified for a unit.
the voltage
and capacity ratings specified for a unit.
Because of these design requirements,
pancake coils of various sizes will be
developed. For instance, the first HV coil
will typically be the smallest to provide
Winding Bending
The design of pancake windings for a shell
form transformer is subject to the voltage
and capacity ratings specified for a unit.
For instance, the first HV coil will typically
be the smallest to provide adequate
7-9
insulation strength to ground. Tertiaries, or
buried stabilizing windings frequently have
reduced MVA requirements, but must still
maintain adequate mechanical strength to
resist short circuit forces. Windings must be
developed to transfer the axial forces to the
clamping or end structure without any
intermediary bending of the windings.
corner of the core can be in the order of
millions of pounds.
The remainder of the winding, above and
below the core, will impact its force on the
end frames, making the calculation of the
stress they can withstand of high
importance.
Furthermore,
the
methodologies to insure tightness between
the windings and the end frame, wedges,
packing, etc., and quality control in their
assembly should be examined.
For example, if there were substantial
differences between the radial build of the
low voltage versus tertiary voltage sections
of a unit, this could lead to the low voltage
winding section actually bending over the
outer diameter of the tertiary during a fault.
As with any winding movement, if it is
substantial enough it will lead to the rupture
of insulation, leading to strand to strand,
turn to turn, or winding to winding dielectric
failures.
Critical Forces in Core Form
Transformers
Radial Forces
Radial
forces result Radial forces in a core
from
the form transformers are the
main axial predominant contributors
leakage flux to through fault failures.
parallel to
the length of the concentric windings.
Geometrically, radial forces are diametric to
the circular windings and will compress the
inner winding, and expand and stress the
outer winding outward. As expected, the
forces will be highest in the high-low space
and this area of the winding requires close
review.
Core Movement End Support Collapse
The core is
well suited In a shell form transformer
to
this the core is an integral part
application of the structure designed to
because of resist short circuit forces.
its
large
mass and stiffness. During short circuit
conditions, the core legs are in tension in the
axial direction, and the core yoke is loaded
as a beam in bending.
Investigation of core form short circuit
failure mechanisms that are generated by
radial forces should consider the following:
inward radial buckling, helical winding
spiral, and outward hoop stress. Specific
attention to winding detail is required, since
core form windings may be constructed in
disk, layer, or helical formats. Each format
has particular strength and weaknesses
relative to radial forces and corresponding
failure mechanisms.
The core is constructed of numerous steel
laminations is an effective structural
member, as long as it acts as a homogeneous
member. To prevent movement and
looseness the laminations and associated
core joints have to be tightly assembled and
clamped. Use of core joint overlap and
clamping in the corners is essential.
Inadequate clamping can lead to core joint
opening, bending of the laminations, and in
general movement of the core. If the
movement is significant or persists, the
winding will loosen, the core will provide
inadequate support, and the unit will fail.
One must appreciate that the forces in the
For example, in a winding built with four
layers, the forces in each layer will be
different. The axial flux and resulting force
for each layer must be calculated separately,
7-10
structure support, and the physical
displacement of primary and secondary
windings during assembly. [2] [10] [11] [12]
[14].
although the layer closest to the high low
space will exhibit the highest forces. [2] [10]
[11] [12] [13] [14].
Failure Modes in Core Type Transformers
In the following some of the failure modes
for core type transformers are discussed.
Although not all inclusive, the following is a
prioritized list of failure modes most
frequently encountered in core-form
transformer designs. The prioritized list is in
rank order of occurrence, i.e., inward radial
buckling is most likely to occur.
Inward Radial Buckling
If
the
The net result of inward
loading
produces a radial force is to compress
stress that the inner winding towards
exceeds the the core.
expected
design capability of the winding, it will
become unstable and deform. The winding
typically deforms in one of two modes,
described as either free or forced buckling.
The buckling mode exhibited will depend
on the stiffness of the winding support
structure as compared to the stiffness of the
winding. Moreover, it will be influenced by
the stiffness of axial spacers that may be
placed between the winding, and the
insulating support cylinder. It is best to
assume that a winding needs to be self
supporting and should be evaluated on that
basis. Buckling failures will be found in
layer, disk, or helical winding coil
configurations.
Figure 7-9
Generated Forces in a Core Type
Transformer
Axial Forces
Axial forces
in a core Axial forces in a core form
transformer are the result
form
of the radial flux present,
transformer
are
the and tend to be stronger
result of the towards the end of the
radial flux winding.
present, and
tend to be stronger towards the end of the
winding where the leakage flux is bending
as it returns to the core. It can further be
exasperated by vertical misalignment of the
magnetic centers of the high and low voltage
windings, or by unbalanced flux distribution
due to tapped sections of a winding. The
resulting axial forces during through faults
are substantial and must be supported by the
clamping and end structures. Axial forces
further challenge design of this assembly,
because of the unequal distribution of these
forces between the high and low voltage
windings.
Forced
Buckling
If
the
stiffness of
the winding
support
structure,
Investigation of axial forces should consider
winding beam bending, clamping and end
7-11
Forced buckling will be
exhibited by the bending of
the
winding
between
supports
without
deformation of the support
structure.
typically consisting of a winding insulating
cylinder and axial spacers, exceeds the
buckling strength of the winding, forced
buckling will result. Forced buckling will
be exhibited by the bending of the winding
between supports without deformation of the
support structure. This is essentially another
type of a beam bending failure.
Figure 7-11
Example of Winding Free Buckling
Critical Hoop Buckling Stress / Calculations
Inward radial buckling is the result of the
radial forces generated by the main axial
leakage flux. Typical hoop compression ( h )
in psi can be calculated as follows:
As indicated in the Figure 6-2, the length of
the buckle LB is equal to the distance
between the supports LS in forced buckling
failures.
Eq. 7-7
The hoop buckling stress is equal to a
constant times the number of turns (N) in the
winding; times the peak magnitude of the
fault current (I) squared (including offset
factor K); times the mean diameter of the
winding (Dm). This is divided by the height
of the winding (h); times the area of a strand
(Ac); times the number of strands comprising
the conductor (Ns) (all dimensions are in
inches). For a small 5 MVA transformer
with a layer winding, a stress of 5600 psi
was calculated [2]. For a medium power
transformer, 40 MVA, with relatively high
impedance this stress was calculated to be
1740 psi with a helical winding; and for a
336 MVA autotransformer calculations
result in a stress of 20103 psi for a disk
winding with CTC bonded conductor.
Figure 7-10
Example of Winding Forced Buckling
Free Buckling
Free
Free buckling of the winding
buckling,
is manifested in a wavelike
see
bulge.
Figure
6-3, is exhibited by the deformation of both
the winding and its support structure,
creating a wavelike bulge in the winding
assembly. This mode of buckling is
observed when the winding support
structure does not reinforce the stiffness of
the winding coil, which is most typically the
case. As shown, the length of the buckle LB
is not equal to the distance between the
supports LS in free buckling failures, and LB
is defined as the half wavelength of the
buckle.
One
approach
is based
on
the
work of
Thompso
n,
et.
al.[19].
From his
7-12
The difficulty comes in
determining what stress is
acceptable for a given
conductor, and is still subject
to
discussion
from
manufacturer
to
manufacturer.
resists tipping since it is a ring of copper that
resists being deformed. The applied force is
opposed by the work done in increasing the
strain energy in the copper. In addition, the
copper is held in place by spacers, radial
spacers in the case of a disk winding, which
provide friction on the edges of the
conductor as it is forced to tip. Empirical
equations to evaluate the critical tipping
stress have been developed to reasonably
predict the critical load that can be
supported [2], [23], but care must be
exercised in calculating the maximum axial
force that will be applied.
test efforts and other published data W. J.
McNutt developed the following critical
hoop buckling stress curves to determine
acceptable stresses, see Figure 7-12.
Figure 7-12
Critical Hoop Buckling Stress (W. J. McNutt)
There are two curves on the graph, one for
annealed copper, and one for hard copper.
In both cases a conductor is adequate if its
thickness supports a hoop stress less than
calculated stress, i.e., below each curve is a
safe zone, and above each curve is a not safe
zone. For example, if the calculated hoop
stress was 6500 psi, the conductor hard
copper and .15" thick, it would be adequate
(in the safe zone) since it is less than the
8450 psi indicated.
Figure 7-13
Example of Conductor Tipping
As expected, the critical load is a function of
the number of spacers, the spacer material,
the dimensions of the conductor, and the
strength of the conductor in the winding.
Thicker and/or stiffer (i.e., higher yield
strength) conductors or an increased number
of spacers can all be used to control this
failure mechanism. The strength of the
conductor can also be increased by using
bonded conductors, which has been
recognized by the industry as a means to
essentially eliminate tipping failures.
However, if the forces are high enough,
even with bonded conductors, a failure can
still occur by rupturing the radial spacers.
A review of the manufacturer's calculations,
models, and testing of this failure mode will
provide
valuable
information
for
determination of acceptable design forces.
Conductor Tipping
Conductor
tipping, see Conductor tipping will most
Figure 7- frequently appear in disk
13 results windings, as illustrated, but
from large can also develop in other
winding configurations.
axial
forces
applied to the narrow edge of a conductor
causing it to tip over. The conductor itself
Beam
Bending
The
solution
7-13
Beam bending, a classic
mechanics
problem,
describes a load or force
applied to a beam rigidly
fixed at both ends.
to this problem applies to any structure, i.e.,
a wood or steel beam in a building, or in this
case to a copper conductor between two
spacers or supports.
high, the winding will compress towards the
core and tighten to a smaller radius. Due to
the spring like helical shape of the winding,
it will actually twist itself into a smaller
diameter. Conductor insulation can be torn,
and the twisting can be severe enough to tear
the leads or connections from the winding.
This problem has been solved by numerous
methods, producing the following result.
Compressive forces in the area of 7000 psi
are
of
concern,
however,
each
manufacturer's design must be evaluated
according to the design's margins.
The bending stress b is equal to the load
applied (w) times the length of the beam (L)
squared, divided by twice the thickness (t) of
the conductor, times the height (ha) of the
conductor squared (all dimensions are in
inches).
Clamping Structure
Axial
Axial forces of varying
forces
will be magnitudes will be produced
up and down the length of a
highest
in areas winding, due to the radial flux
present.
with
inconsist
ent distribution of amp turns, such as tap
locations, or due to the displacement of the
magnetic centers of the windings. When all
the forces are resolved, a net axial force will
be presented to the clamping structure. The
clamping structure must be designed to
support the maximum force produced, from
the highest force producing combination of
taps in the transformer. Once the short
circuit force is identified, the clamping
structure can typically be evaluated as a
beam type structure. Single piece clamping
structures, versus multi-segment clamps,
provide the best performance.
Calculations of beam bending stress become
increasingly difficult when multi-strand
conductors, CTC, or bonded conductors are
used. The beam equation still applies, but it
becomes a challenge to determine the
maximum bending stress allowed for the
composite conductor.
A beam bending failure occurs when the
applied load produces a bending stress, b,
greater than the yield strength of the
conductor considered. The conductor will
permanently deform by bending, and may
rupture both its own and adjoining
insulation. With the integrity of the
insulation structure compromised, a
dielectric failure would be imminent.
Bending stress can easily be reduced by
increasing the number of spacer columns,
i.e. reduce the unsupported length. In
average impedance designs an unsupported
spacing of 5 inches is typical, and in high
impedance designs (lower fault currents)
unsupported spacings can approach 8 inches.
Typical failures of clamping structures result
either from loss of clamping pressure on the
winding; non-uniform clamping pressure on
the winding (axial forces on inner and outer
winding will be different), or structural
failure of same. Forces can be of the order of
multiples of a hundred thousand pounds.
Helical Winding Spiral
If
the
forces are
sufficiently
Helical windings, which
are frequently applied as
high current low voltage
windings, are subject to
inward radial forces during
short circuit conditions.
Impact of Fault Currents on Transformer
Life
7-14
in Table 7-3 . It can be seen that out of the 6
samples, one failed between 35,600 and
50,400 cycles and another one failed
between 50,400 and 64,400 cycles. The
complete set of test results can be found in
(McNutt
and
Patel.
1976).
McNutt and
Patel did an Impact forces created by
high magnitude through
experimental
fault currents have an
study
to
adverse impact on the
demonstrate
transformer life
the impact of
thermal aging
and mechanical stresses on the transformer
insulation. They used statistical analysis on
experimental results to quantify the
relationship between insulation wear life and
thermal and mechanical parameters.
Table 7-3
Life Test data – Disk Winding
As per their hypothesis, cumulative thermal
aging over the years reduces the mechanical
strength of the insulation. Then, repeated
mechanical stresses resulting from shortcircuit forces contribute to the mechanical
weakening and eventual rupture of all or a
portion of the weakened insulation. Finally,
the reduced dielectric strength of the
ruptured insulation allows a dielectric failure
during a period of transient overvoltage.
Number of
Samples
Cycles
Passed
Cycles Failed
6
10,200
-
1
35,600
50,400
1
50,400
64,400
4
64,400
-
These experimental results have been used
to come up with a mathematical model of
the life relationship of the insulation with
thermal aging and mechanical stresses. The
mathematical expression that was used to
represent the relationship is given by:
Two types of winding geometries were
modeled for the experimental study. The
first was a layer winding employing
machine transposed cables (MTC) with
annealed copper conductor, film strand
insulation, and paper turn insulation. The
second was a disk winding using annealed
rectangular copper conductor with paper
insulation. Three degrees of aging was
selected for the paper-turn insulation in the
winding models to represent new, 2.4 years
and 7.2 years of continuous service at 110º
C. Tests were conducted in a short circuit
laboratory by using force generation coil to
generate the axial stresses representing the
mechanical impact of fault currents. As per
the failure criterion adopted, the complete
destruction of the conductor insulation
represented the failure.
L = C1σ − C2 ε − C3 y
Eq. 7-8
Where,
L = Functional life in number of cycles
of short circuit current.
= Mechanical stress on conductor
insulation during short circuit, in psi.
y = per unit thermal age.
C1, C2, C3= constants to be determined
by curve fitting techniques.
Weibull distribution was assumed for the
purpose of statistical analysis of life test
data. The method of maximum likelihood
was then used for curve fitting to determine
the values of the constants (See Table 7-4).
Sample test data for the disk winding (6
samples) having thermal age of 7.2 years
and subjected to stress of 2270 psi is shown
The derived mathematical relationship was
used to perform life prediction of 5MVA
and 50MVA transformer (See ). Results
7-15
may not be very accurate due to the various
assumptions involved and the difficulty of
replicating the actual conditions in the
experimental setup. But it can be deduced
that from the point of view of insulation
wear failure, larger transformers have lower
life expectancy.
Summary and Recommendations
Impact forces created by high magnitude
through fault currents have an adverse
impact on the transformer life. The resultant
short-circuit mechanical stresses can
physically damage the conductors and
insulators resulting in transformer failure.
Therefore, current standards take into
account both thermal and mechanical impact
of fault currents while defining the through
fault
protection
curves
of
power
transformers.
The life relationship is not accurate enough
to predict the actual short-circuit life of a
transformer, but it can be used to measure
the relative effects. For example, if the short
circuit current is reduced to 50% of
maximum, the mechanical stress is reduced
to 25% resulting in increase in wear life by a
factor of 140.
References
Table 7-4
Constant values from Curve Fitting
Winding
Type
Constant
C1
Constant
C2
Constant
C3
Disk
8.5e18
3.57
3.99
Layer
1.3e24
5.47
4.58
1. K. Karsai, D. Kerenyi, and L. Kiss,
Large Power Transformers, Elsevier:
New York, 1987.
2. M. Waters, The Short-Circuit
Strength of Power Transformers,
Macdonald: London, 1966
3. R. L. Bean, N. Chackan Jr., H. R.
Moore, and E. Wentz, Transformers
for the Electric Power Industry,
McGraw-Hill: New York, 1959.
4. L. F. Blume, A. Boyajian, G. Camilli,
T. C. Lennox, S. Minneci, and V. M.
Montsinger, Transformer Engineering
A Treatise on the Theory, Operation,
and Application of Transformers,
John Wiley & Sons, Inc.: New York,
1951
5. J. D. Fyvie - "Design Review to
Determine
The
Short
Circuit
Capability of Power Transformers,"
VA Tech Transformers for WG12.19
Task Force #3, CIGRE, Transformer
Colloquium, Budapest, booklet 2,
June1999.
6. H. Moore, "The Short-Circuit
Problems of Power Transformers,
Calculation of Forces and Stresses,
Figure 7-14
Example Life Prediction (McNutt and Patel.
1976)
7-16
Consequences on Design and
Construction, Full Scale and Model
Testing…" Third Part, CIGRE Report
no.12-00, 1980.
13. R. Boersma and J. Wildeboer, "The
Short-Circuit Strength of the Inner
Windings of Transformers Against
Radial Forces," CIGRE, Report no.
147, 1962.
7. T. M. McCauley, "Through-Fault
Capability Requirements for Unit
Auxiliary
Transformers,"
IEEE
Transactions on Power Apparatus and
Systems," vol. PAS-96, no. 5,
September/October 1977, pp.16391647.
14. "Calculation of Short-Circuit Forces
in Transformers", Working Group 1204, CIGRE Study Committee No. 12,
ELECTRA, no. 67, 1979, pp. 29-75.
15. M. Boutteau, J. Verdon, P. Hofer, B.
Hochart, Y Tournier, and G. Roge,
"Short-Circuit Behavior of Large
Power Transformers, CIGRE, Report
no. 12-07, 1972.
8. W. J. McNutt, C.J. McMillen, P. Q
Nelson, J. E. Dind, "Transformer
Short-Circuit Strength and Standards
- A State-Of-The-Art Paper," IEEE
Transactions on Power Apparatus and
Systems," vol. PAS-94, March/April
1975, pp.432-443.
16. G.
Thompson,
and
R.
A.
Schwarzmeier, "Designing a shellform transformer for maximum shortcircuit strength," The Line, McgrawEdison Company, Issue 82/1, 1982.
9. A. Bossi, G. Caprio, A. Inesi, L.
Giannuzzi, A. Babare, and G.
Sigaudi,
"The
Short-Circuit
Withstand
of
Large
Power
Transformers Contribution to Design
Improvement and Test Criteria,"
CIGRE, Report no. 12-12, 1980.
17. R. L. Bean and E. C. Wentz,
"Mechanical Forces in Interleaved
Rectangular Pancake transformer
Coils," AIEE Transactions, Pt. 3, vol.
73, 1954, pp. 962-971.
18. D. Girardot, and G. Robert,
"Mechanical Failure Modes in Shell
Type
Transformers,"
CIGRE,
Transformer Colloquium, Budapest,
booklet 2, June 1999.
10. W.J. McNutt, W. M. Johnson, R. A.
Nelson, and R. E. Ayers, "Power
Transformers Short-Circuit Strength Requirements,
Design,
and
Demonstration," IEEE Transactions
on Power Apparatus and Systems,"
vol. PAS-89, November/December
1970, pp.1955-1969.
19. H. A. Thompson, F. Tillery, and D.
U. von Rosenberg, "The Dynamic
response of Low Voltage, High
Current, Disk Type Transformer
Windings to Through Fault Loads,"
IEEE
Transactions
on
Power
Apparatus and Systems," vol. PAS98, May/June 1979, pp1091-1098.
11. E.W. Tipton, "Mechanical Problems
Involved in Short Circuits on CoreForm Power-Transformer Coils," The
American Society of Mechanical
Engineers, Semi-Annual Meeting,
San Francisco California, June 9-13,
1957.
20. R. B. Steel, W. M. Johnson, J. J.
Narbus, M. R. Patel, and R. A.
Nelson, "Dynamic Measurements in
Power Transformers Under Short
Circuit Conditions," CIGRE, Report
no. 12-01, 1972.
12. E. T. Norris, "Mechanical Strength of
Power Transformers in Service," IEE
Proceedings, vol. 104A, 1957, pp289300.
7-17
21. H. Kojima, H. Miyata, S. Shida, and
K. Okuyama, "Buckling Strength
Analysis of Large Power Transformer
Windings
Subjected
to
Electromagnetic Force Under Short
Circuit," IEEE Transactions on Power
Apparatus and Systems," vol. PAS99, no. 3, May/June 1980, pp.12881297.
92, September/October 1973, pp.
1558-1576.
28. M. Kozlowski, W. Marciniak, W.
Pewca, and W. Weretynski, "Selected
Short-Circuit Strength Problems in
Power Transformers," CIGRE, Report
no. 12-05, 1980.
EPRI, XVisor Transformer Expert
System Software, Electric Power
Research Institute.
22. R. M. Del Vecchio, B. Poulin, and R.
Ahuja, "Radial Buckling Strength
Calculation and Test Comparison for
Core-Form Transformers," CIGRE,
Transformer Colloquium, Budapest,
booklet 2, June1999.
SEL 2005. “Protecting Transformers
from Common Adverse Conditions”
23. L. Torske, E. Stenkvist "Short Circuit
Problems in Large Transformers,
Appendix II," CIGRE, Report no.
142, 1962.
24. W. J. McNutt, M. R. Patel, "The
Combined Effects of Thermal Aging
and Short-Circuit Stresses on
Transformer Life," IEEE Transactions
on Power Apparatus and Systems,"
vol. PAS-95, no. 4, July/August 1976,
pp.1275-1283.
25. Y. Tournier, G. Ebersohl, A. Ciniero,
S. Yakov, A. B. Madin and J. D.
Whitaker, "A Study of the Dynamic
Behavior of Transformer Windings
Under Short Circuit Conditions,"
CIGRE Report no., 143 and 143a,
1962.
26. D. O. Swihart, and L. S. McCormick,
"Short Circuit Vibration Analysis of a
Shell Form Transformer," IEEE
Transactions on Power Apparatus and
Systems," vol. PAS-99, no. 2,
March/April 1980, pp.800-810.
27. M. R. Patel, "Dynamic Response of
Power Transformer Under Axial
Short-Circuit Forces, Parts I and II,"
IEEE
Transactions
on
Power
Apparatus and Systems," vol. PAS-
7-18
8
FAULT CURRENT
LIMITING METHODS
This chapter discusses the fault current
methods.
Both
conventional
and
advanced technologies will be discussed.
Technologies to be considered will
include:
•
Conventional technologies, such
as Current Limiting Reactors
(CLR) and high resistance
grounding.
• Solid-state, such as the Solid State
A fault condition
may result
in an electric
Current Limiter
(SSCL).
power transmission system from events such
• Superconducting,
the
as lightning
striking a power such
line, orasdowned
Superconducting
Current
Limiter
trees or utility
poles shorting
the power
lines
(SCCL).
to ground.
The fault creates a surge of
current through the electric power system
The can
material
is basically
that
cause shown
serioushere
damage
to grid
extracted
from
the
equipment. Switchgear, such available
as circuit
manufacturer
materials.within transmission
breakers,
are deployed
substations to protect substation equipment.
Figure 8-1
Overview of fault Current Limiting Measures
Conventional Methods
Some of the
conventional
solutions are
listed below:
An overview of fault current limiting
measures is given in Figure 8-1 (Schmitt and
CIGRE WG A3.16. 2006). The “Passive”
measures make use of higher impedance
under all the conditions whereas “Active”
measures introduce higher impedance only
under fault conditions. The measures may
also be classified as “Topological” and
“Apparatus” measures. Further, measures
may also be classified into “Conventional”
and “Emerging/Novel” depending on the
technology used. The individual measures
are explained in the subsequent sections.
Conventional solutions to
fault current over-duty
problems are often costly.
1. Construction of new substations.
Fault current over-duty coupled along
with other factors may result in a
utility selecting this solution, which
will correct immediate problems, as
well as providing for future growth.
However, this is the most expensive
of all the conventional solutions.
2. Introducing a higher voltage level.
The choice of a particular voltage
level for (new) transmission and
distribution systems is governed
primarily by the desired power
8-1
ratings. The objective is to keep rated
current levels within the standard
brackets of commercially available
equipment, especially circuit breakers
(e.g. IEEE C37). These brackets
typically provide enough margin with
respect to short circuit power at any
given voltage level. Nevertheless,
depending on the constraint of the
applications (e.g. grid density, nearby
generation) the short circuit power
may exceed the ratings of the
available equipment. This may
require choosing a higher voltage
level based on the short circuit
capability of the equipment. All these
considerations will play a role when
designing new systems.
divided into smaller portions which
are then fed separately from the next
higher voltage level. The splitting
reduces the fault current level in each
of the sub grids to the allowable level.
5. Multiple circuit breaker upgrades.
When a fault duty problem occurs,
usually more than one breaker will be
affected. Upgrade of these breakers
has the disadvantage of not reducing
available fault currents and their
associated hazards, as well as the
often-prohibitive
expense
of
replacement of all, or nearly all, of the
switchgear within a substation.
6. Current limiting reactors (CLRs)
and high impedance transformers.
Fault current limiting reactors, Figure
8-2, limit fault current due to the
voltage drop across their terminals,
which increase during the fault.
However, current limiting reactors
also have a voltage drop under normal
loading conditions and present a
constant source of losses. They can
interact
with
other
system
components and cause instability, as
well as an increase in Transient
Recovery Voltage (TRV). When
conditions are right, air core CLRs
can be an economical solution to high
fault currents (Amon, et al. 2005).
Also, the installation of a suitable
capacitor across the reactor and from
the sides of reactor to ground can
solve the problem of severe TRV.
In case of existing systems, increasing
the voltage level is more likely a
viable option for medium voltage
levels where the increase in system
voltage can be accommodated more
easily within the same or similar
geometrical constraints by simply
installing more modern equipment. In
high voltage systems, increasing the
voltage level often is associated with
major investments and thus not a
preferred option in many cases.
3. Bus splitting. This entails separation
of sources that could possibly feed a
fault by the opening of normally
closed bus ties, or the splitting of
existing buses. This effectively
reduces the number of sources that
can feed a fault, but also reduces the
number of sources that supply load
current during normal or contingency
operating conditions. This may
require additional changes in the
operational philosophy and control
methodology.
4. Splitting into sub-grids. This term
refers to a measure whereby a grid
(with one common voltage level) is
8-2
(a)
Figure 8-3
CLR as Bus Coupling
(b)
Figure 8-2
Air Core Current Limiting Reactor, (a) Under
Testing, (b) in a Substation (Areva T&D)
The three possible locations of CLR at
the substation bus bar are shown in
Figure 8-3, Figure 8-4 and Figure 8-5.
The configuration in Figure 8-3 that
involves installation of CLR as coupling
is effective in reducing the short-circuit
capability of the system but is unable to
control the individual contributions of
the incoming feeders. Individual limiting
from the feeding sources is achievable in
configuration in Figure 8-4 by providing
a CLR for each incoming feeder but it
suffers from high losses and poor
regulation. These disadvantages can be
overcome by using configuration in
Figure 8-5 where a CLR is installed in
each outgoing feeder.
Figure 8-4
CLRs in series with Incoming Feeders
Figure 8-5
CLRs in series with Outgoing Feeders
8-3
two sources (A and B). All the
breakers (A1 through D2) can handle
the maximum fault current, for
example in case of a fault at feeder C
with both sources contributing. When
a new source (N) shall be added, the
total fault current may exceed the
ratings of the existing breakers. If
only those interrupting ratings are of
concern while all the equipment can
carry the increased fault current a
sequential tripping scheme may be
applied to avoid upgrading of
breakers A1 through D2. In that case,
the new breaker (N1 or N2) is tripped
first while the tripping signals for the
existing breakers are delayed until
source N is disconnected. This
measure may be considered a fault
current limiting measure since it
reduces the fault current duty of the
interrupting device (circuit breaker)
by changing the system topology
during the fault. From that viewpoint
it is a mix between a topological and
an apparatus measure. However,
sequential tripping does not reduce
the overall fault level on the system.
Furthermore, it poses an increased
reliability risk and may overstress
equipment over a longer period of
time.
7. Impedance Grounding. When the
high fault currents are ground fault
currents, solidly grounded systems
can be converted to impedance
grounded systems, such as lowresistance grounding, inductance
grounding,
or
high-resistance
grounding. (Figure 8-6)
Figure 8-6
Neutral Grounding Reactor (Trench Electric)
8. Sequential
breaker
tripping.
Sequential tripping of circuit breakers
is a special measure occasionally used
in substations to manage high fault
currents without replacing all circuit
breakers. A sequential tripping
scheme prevents circuit breakers from
interrupting excessive fault currents.
If a fault is detected, a breaker
upstream to the source of fault current
is tripped first. This reduces the fault
current seen by the breaker within the
zone of protection at the location of
the fault. This breaker can then open
safely. A disadvantage of the
sequential tripping scheme is that it
adds a delay of one breaker operation
before final fault clearing. Also,
opening the breaker upstream to the
fault affects zones that were not
originally impacted by the fault.
Figure 8-7 illustrates the method. In a
substation with two busses the two
feeders (C and D) are connected to
Figure 8-7
Illustrating the Sequential Tripping
Scheme (EPRI 2005a)
9. Stand alone HV fuses. A stand-alone
HV fuse is a device which carries the
load current directly through the
8-4
explosive charge that opens a link,
diverting the fault current to a parallel
current limiting fuse, Figure 8-9.
Triggering is by an electronic module,
which senses the di/dt rise typical of a
high level of fault current.
IsLimiters are typically employed as
bus tie limiters when the interrupting
capability of the circuit breakers of a
substation have been exceeded. A 38
kV version was developed in a 1996
EPRI project, (EPRI. 1996), (Das.
1997.). Figure 8-10 shows the active
parts of the so-called “Is-limiter”
from ABB Calor Emag (Germany) (a
similar device is available from G&W
Electric in the USA). The load current
does not flow through the fuse but
through a parallel path (9), which is
opened by means of an explosive
charge (10) when a fault is detected
by the trigger unit. Therefore,
thermodynamic requirements of the
stand-alone fuse do no longer govern
the design of the fuse element in the
Is-limiter. Consequently, this allows
for significantly higher rated currents.
For example, the Is-limiter is
available up to 4 kA rated current (at
17.5 kV) and up to 2.5 kA (at 40.5
kV). For higher load currents, two or
more Is-limiters can be installed and
operated in parallel provided equal
current sharing is guaranteed by
proper
special
arrangements.
Interrupting ratings range from 140
kA (at 40.5 kV) to 210 kA (at 17.5
kV). The external trigger circuitry
allows for a more flexible setting of
the tripping characteristic. Tripping
can be blocked or the tripping level
can be changed if system conditions
require a new setting. Because of their
higher current ratings commutation
fuse-based limiters are often used in
substation bus ties and to protect
generators, which are tied directly
into the medium voltage network.
active (melting) element. The design
is somewhat more complex than any
low voltage fuse in order to develop
high enough arcing voltage for the
high voltage application. Otherwise,
the basic principle is the same.
Resistive heating of the fuse element
due to over currents of fault currents
causes it to melt within a certain time
interval. This pre-arcing time is a
function of the fault current with a
characteristic determined by the
design (e.g. the voltage level).
Therefore, the tripping characteristic
of a specific HV fuse cannot be
changed in the field.
Some HV current limiting fuses are
fitted with a so-called striker
mechanism. This not only provides a
visual indication that the fuse has
operated, but can also be used to
operate other switchgear. In this way,
a fuse on a single-phase system can
cut off all three phases if a fault
occurs.
Due to the fact that the load current is
carried directly by the active
(melting)
element,
the
thermodynamic design requirements
only allow for rated currents of
typically less than several hundred
amps. For currents in excess of that
stand-alone HV fuses are not
available.
Interrupting
ratings
typically range up to 63 kA.
Stand-alone HV fuses are widely used
to protect feeders and apparatus such
as transformers in medium voltage
distribution systems and motors in
industrial systems.
10. Pyrotechnic devices.
Sometimes
called “Is-Limiters,” these devices
(Figure 8-8) are activated by a small
8-5
Commutation fuse-based limiters can
also be installed in parallel to a current
limiting reactor (bypass). In such an
application the impedance of the reactor
is inserted only during a fault (when the
fuse-based limiter has triggered). This
improves voltage regulation and
stability during normal operation
(reactor bypassed) while maintaining
the supply after the fault has been
cleared elsewhere in the system.
Table 8-1 summarizes these solutions and
their respective pros, cons and relative cost.
This table primarily considers the initial
capital installation cost in the cost
comparison. In the cases of multiple circuit
breaker upgrades, the cost of bus work
reinforcement must also be considered, since
the level of fault current is not being
reduced.
Limiter (G & W Electric)
Table 8-1
Pros and Cons of Conventional Solutions
Figure 8-8
Current Limiting Protector, EPRI (1996)
Figure 8-9
Operation Sequence of Pyrotechnic Current
Solution
Advantages
Disadvantages
Relative
Expense
Relative
Expense to
SSCL/SCCL
New
Substation
Solves all fault
issues and
accommodates
future growth
Expensive and
lengthy to install
Very high
SSCL/SCCL
less expensive
Bus
Splitting
Separates
source of fault
current from
over-duty
breakers
Also separates
sources of load
current from load
centers and
undermines
system reliability
High, if
breakers
to split bus
are not
already
installed
SSCL/SCCL
less expensive
Multiple
Breaker
Upgrades
Most direct
solution to
problem with
no adverse
side affects
Difficult to
schedule outages,
due to numerous
breakers involved;
Bus work
reinforcement also
needed
High to
medium,
depending
on number
of
breakers
involved
Expected to
be competitive
with most
multiple
breaker
replacements
Current
Limiting
Reactors
Easy to install
Voltage drop and
power loss;
potentially cause
instability and the
need to install
compensating
capacitors
Medium to
Low
SSCL/SCCL
cost higher
Impedance
Grounding
Easy to install.
Only limits ground
fault current.
Requires changes
in protective
relaying.
Medium to
Low
SSCL/SCCL
cost higher
Sequential
Breaker
Tripping
No major
hardware
installation
involved
Expands impact of
fault to wider
range of the
system and
undermines
reliability
Low
SSCL/SCCL
cost higher
Pyrotechnic
Devices
Easy to install
Element must be
replaced after
operation
Low
SSCL/SCCL
cost higher
Figure 8-10
Is-Limiter Construction (EPRI 2005a)
8-6
Emerging Technologies
The
Novel technologies are
disadvantages
of
the in various stages of
development but are
conventional
methods
for expected to be gridlimiting
the ready in near future
fault
current
have been documented in the previous
section. Utilities are seriously re-assessing
fault current mitigation methods. They are
considering emerging new technologies
(solid-state, superconducting etc.) as vital
alternatives to existing methods, provided
these technologies prove to be the most cost
effective
means
of
fault
current
management. Recently, there has been a
phenomenal increase in R&D activities
towards the development of technically
feasible
and
economically
viable
technologies to design a range of medium
voltage and high voltage devices for fault
current limiting applications in distribution
and transmission.
Figure 8-11
Equivalent Circuit and Current Waveforms
during a Fault (CIGRE 2003)
In the figure,
i1
= Fault current in absence of any
current limiting action which is interrupted
by conventional breaker at time t3
FCL Definitions
A fault current limiter can limit a fault
current passing trough it within the first half
cycle. The principle of FCL operation is
shown in Figure 8-11.
i2
= Fault current due to the limiting
action of the FCL
= Fault current due to the limiting
i3
action of the FCL that also gets interrupted
at time t2
Some definitions related to fault current
limiting are:
Ir
= Peak value of rated current
Imin
= Minimum current that initiates
FCL action
8-7
Imax
= Maximum limited current
Ip
= Peak short circuit current
Ifol
= Peak value of follow current
•
High reliability
•
Low maintenance
•
No risk for personnel
•
Low impact on transformer
tr
= Recovery time (time between
current interruption and return of FCL to its
initial operation state)
•
Low weight and small size
•
No or only few auxiliaries
η0
= Follow current ratio (Ifol/Ir)
•
Low total cost of ownership
η1
= Peak current limiting ratio (Imax/Ip)
η2
= Current limiting ratio (Ifol/Ip)
ta
= Action time (time required from
fault initiation to maximum limited current)
td
= Fault duration time (time required
from fault initiation to fault current
interruption)
The background on the novel FCL
technologies and the status of the various
projects is presented in the following
sections.
η2
= Dynamic current limiting ratio
(Imax/Ir)
ηi
Solid-State Current Limiter (SSCL)
=Initiating ratio (Imin/Ir)
High power Although
the
power
solid-state
industry
has
been
power
interested in the concepts
components
of
solid-state
circuit
have
breakers and in limiting
continued to of fault currents for
improve in decades, it only now
their
appears that the time has
performance
arrived when the selling
while their price can be low enough
cost
has to justify significant sales.
continued to
decline. The
aim is to use these evolving trends, together
with some focused thinking about optimum
circuits and arrangements, to achieve a
practical, useful and cost effective solution
to some of the utility industry’s technical
problems. The Solid-State Current Limiter
(SSCL) offers a viable solution to the
transmission and distribution system
problems caused by high available fault
current. By providing instantaneous (subcycle) current limiting, the SSCL alleviates
the short circuit condition in both
downstream and upstream devices by
limiting fault currents coming from the
sources of high short circuit capacity. The
Requirements for Fault Current Limiters
Some of the requirements on a FCL from the
utilities perspective (Duggan.2006) and
CIGRE WG (CIGRE 2003) can be
summarized as follows:
•
Low impedance
operation
•
Low losses
•
Adequate
current
limiting
performance in terms of maximum
limited current and action time being
less than the time to first peak.
•
Automatic and quick recovery to
initial operating state
•
Ability to withstand the magnetic
and mechanical forces of repeated
mitigation operations
•
Compatibility with existing
planned protection schemes
•
No deterioration of current limiting
performance during the design life
during
normal
and
8-8
advantages of added functions that a
conventional circuit breaker cannot offer
help to justify the higher cost associated
with a solid-state system. To interrupt the
current, the SSCL must rapidly turn off its
conducting components and insert an
energy-absorbing resistor into the circuit to
limit the fault current. In addition to limiting
the fault current, the SSB (Solid State
Breaker) can also be beneficial on closing
by limiting the inrush current (soft start
capability), even for capacitive loads, by
gradually phasing in the switching device
rather than making an abrupt transition from
an open to a closed position. A solid-state
current limiter can offer the following
advantages
•
Normal operation where no limiting
action takes place
•
Fault condition during which the
FCL is active
•
Recovery period while the FCL
resets and regains normal operating
condition.
I Normal
operation
II Fault condition
Fault Inception
current
Fault Clearing
W ithout FCL
Îm ax
Î fol
Îp
În
III Recovery
W ith FCL
Îm in
time
ta
tr
td
•
•
td
Limited fault current
Limited inrush current (soft start),
even for capacitive loads
• Repeated operations with high
reliability and without wear-out
• Reduced switching surges
This extra
performance By limiting the current,
comes with one can achieve fault
and
better
no
extra isolation
effort as an network protection, taking
aspect of the care of most of the
distribution
system
speed and
situations that result in
limiting
voltage sags. Thus the
qualities of
SSCL can substantially
the SSCL.
improve the power quality
High fault
through fault current
currents are
limiting
and
inrush
known to be current reduction.
a factor in
reducing
transformer life, so it is expected that an
advantage from the use of a current limiting
breaker will be longer life with higher
reliability for nearby transformers.
rated system
voltage (Un)
(î n ):rated current
(peak)
îm in : minimum initiating current
îm ax : maximum limited current
î p : peak (prospective) short circuit current
î fol : peak value of the follow current
ta : action time: from t = 0 until î m ax
td: fault duration time
recovery time tr
time between
current interruption
and return of the
FCL to its (initial)
low impedance
state
Figure 8-12
Generalized fault current trace with FCL
activated (EPRI 2005)
The switch turn-off operation without
having to interrupt current immediately or to
limit the fault current can be delayed until
zero crossing.
Figure 8-13 shows the
schematic circuit diagram and the waveform
associated with the switch operation. In this
case, the silicon-controlled rectifier is used
as the switch. If the immediate fault clearing
is needed, then the switch needs to be gatecontrolled devices such as GTO or IGBT or
"Super" GTO. The schematic circuit
diagram and the waveform associated with
the switch operation are shown in Figure
8-14. The use of GTO or IGBT may allow
current limiting, but their conduction voltage
drop is too high that it is not practical to use
these gate-controlled devices alone in the
100% continuous conducting duty.
Figure 8-12 depicts principal waveforms and
indicates the three basic operating regions of
a FCL:
8-9
"All Solid-State" Based Designs
Thyrsitor-based
Solid-state
circuit breaker
Sss
Vs
The most straightforward solid-state fault
current limiter (SSFCL) is the solid-state
fault current limiting circuit breaker. Figure
8-15 depicts the basic phase module of such
a device built by SIEMENS using turn-off
devices such as IGBTs or IGCTs (Kunde, et
al). These devices are placed in the DC
branch of a full-bridge diode rectifier circuit.
Therefore, only one unipolar turn-off device
is required for AC line current operation
(iLINE). The second device shown in this
figure is for increased voltage withstand
capability and adequate reliability to meet
the “N-1” failure mode criteria. In addition
to the turn-off device there must exist an
over voltage protection element such as a
metal oxide varistor (MOV) in order to limit
the voltage build-up caused by the AC line
inductance during the hard turn-off by the
IGBT. Typically, such a SSFCL-CB is
designed to develop 2-3 times the rated
system voltage during turn-off which forces
the fault current very rapidly (within 1 ms)
down to zero. One module may typically
develop up to 6 kV and turn-off up to
5.6 kA. A medium voltage SSFCL-CB may
consist of several modules connected in
series. Similar systems have also been
developed by other companies but no
economically viable solution could be made
available for the commercial market.
Vs
Sss
tclear
Fault
occurs <8.3ms
Figure 8-13
Solid-State Switch Operation without Having
to Interrupt Immediately or Fault Current
Limiting
GTO-based
Solid-state
circuit breaker
Sss
Vs
Vs
i LINE
Sss
Fault
occurs
Snubber
circuit
Balancing
resistors
tclear <1ms
Figure 8-14 Using GTO-Based Solid-State
Circuit Breaker Allows Instant Fault Current
Clearing.
Turn-off
devices
Over-voltage
protection (MOV)
Figure 8-15
Principle of a solid-state fault current limiting
circuit breaker based on turn-off devices
8-10
An alternative circuitry for a solid-state fault
current limiter (circuit breaker) based on
SCR thyristors with commutation circuitry
rather than turn-off devices is currently
under development for EPRI by Powell
Electronics Inc (EPRI 2004b). The project
status is presented later in the chapter.
In contrast to the GTOs where the current
can be interrupted at any point in the cycle,
SCRs can interrupt currents when the
current waveform goes to zero. Thyristor
breakers, unlike GTO breakers, can be
designed to maintain fault current to satisfy
the required time-current characteristics for
typical overcurrent protection coordination
schemes. The SCR section will be able to
conduct fault currents for a period of time
(10 to 15 cycles), repeatedly.
Westinghouse in association with EPRI
developed a prototype (see Figure 8-16) of
an "all solid-state" distribution breaker
where a SCR-GTO combination is used. As
the figure suggests, this design consists of
two parallel connected circuit branches: a
solid-state switch composed of GTOs (and
their associated snubber and over-voltage
protection metal oxide arresters (ZNOs))
and another solid state switch using SCRs
(and their associated components). A unit
was built for 13.8KV by series grounding of
six GTO modules per phase each one rated
for 3000A and 4.5KV.
The advantage of the GTO switch is its
capability to interrupt current with
negligible delay. The advantage of SCR
switch compared to GTO switch of the same
wafer size is its ability to handle
considerably higher currents. SCRs are also
available commercially with higher nominal
current rating (required for distribution
voltages above the 15-kV class).
One major drawback for "all-solid state"
designs is that thyristor based design will
have substantial losses during normal
operation serving the load. In EPRI's SCR
based fault current limiters the thyristors
will have per phase conduction losses of
about 14.4KW, and Three Phase Losses =
2V(drop)*6(in series)*1200A*3(phase) =
43.2KW, when carrying the rated current of
1200 amperes. Another problem with this
circuit and the SCR-GTO combination
circuit is the component counts and their
associated reliability issues. Also, thyristors
are a mature end technology. It will be
difficult to drive the component cost down
even with the wide spread potential market
for distribution switchgears.
Pairs of anti-parallel connected GTO
devices are used in series in the GTO section
of the SSB. The GTO switch is the main
circuit breaker and it is conducts load
current in the steady state. The GTO switch
is used to clear source-side faults. It is rated
for the maximum normal line current, but
not rated for fault currents. It is normally
closed and conducts current until the
magnitude of the current reaches a pre-set
level at which point it opens rapidly
interrupting the current flow. To achieve the
required SSB voltage for application to the
utility 13.8-kV primary distribution voltage,
six GTO modules are required in series for
each phase.
The SCR switch is normally open and has
no continuous current rating. Its function is
to conduct fault current to facilitate
operation of conventional protective devices
on the load side of the SSB. For this purpose
it is rated for short duration fault surge
currents. Its operation is coordinated with
the GTO breaker.
8-11
driven along metallic rails of high resistivity
in order to produce a voltage drop across the
switch large enough to commutate the fault
current into a parallel resistor. The device,
rated 7.2 kV/400 A has been tested
successfully in the field for over 2 years in
Japan. No information is available on any
further developments of this technique,
especially not with respect to higher voltage
applications.
A hybrid solid-state switch that can also
perform current limiter function was
proposed (Steurer, et al) at the Swiss Federal
Institute of Technology and patented by
ABB Switzerland. Figure 8-17 shows the
hybrid switch that utilizes an ultra fast
mechanical switch Sm1 for the normal
conducting path, an IGCT-based solid-state
switch for a short conducting period to
prevent arcing, and a positive-temperaturecoefficient (PTC) resistor for current
limiting. A 10kV/ 1kA prototype was tested
successfully in ABB Switzerland. The main
problem with this circuit is the need for a
very high-voltage high-current PTC, which
requires a substantially stacking effort with
the commercially available low-power PTC
products. Furthermore, the use of three
mechanical switches indicates the problem
of economical design issue. In fact, if the
mechanical switch Sm1 is fast enough, and the
PTC is available, the solid-state switch Sss
and its associated mechanical switch Sm2 can
be eliminated.
Figure 8-16
Solid-State Breaker Proposed by
Westinghouse
Hybrid Designs
The operating characteristic of solid-state
switchgear is primarily dictated by the
capabilities of the semiconductor devices
used. Voltage and current ratings of the
breaker define the number of power
semiconductors required and, consequently,
the cost and the operating losses of the
breaker. Since a closed mechanical contact
still exhibits the least amount of conduction
losses amongst all “switching” elements it is
most desirable to utilize mechanical contacts
in fault current limiters for carrying the
continuous operating current. However,
mechanical contact systems alone will not
develop enough arcing voltage drop to limit
fault currents in medium or high voltage
systems.
A wide range of materials, mostly ceramics,
exhibit a highly non-linear positive
temperature coefficient (PTC) of the
resistivity above room temperature. A sharp
increase in resistance can be used for fault
current limitation. In fact, PTC resistors are
commonly used for fault current limitation
in (low power) electronic circuits. To extend
this functionality into the medium voltage
range for possible applications in power
systems was the goal of a project
(Strumpler, et al) by ABB. The project
One possible solution to achieve sufficiently
high arcing voltage at least for a medium
voltage (distribution) class FCL is the
method of a “driven arc” described in a
recent reference (Fukagawa, et al). Similar
to the technique used in low voltage current
limiting circuit breakers the switching arc is
magnetically driven into a special chamber
where it is divided into a large number of
sub-arcs. Subsequently, these sub-arcs are
8-12
concluded with the successful testing of a 12
kV class stack of PTC elements (at very low
rated current of only 10 A, however)
Sm1
Design Description
Design Specifications
The initial SSCL design specifications were
established in December 2001 after a
number of EPRI meetings with input from
electric utilities. The key requirements of the
SSCL were:
ultrafast switch
mechanical
switch
Sm2
Sss
GTO
MOV
PTC
Sm3
•
Meet Peak Voltage Withstand
Capability – twice the phase to
neutral voltage plus a 10% allowance
for voltage regulation.
• Minimum fault current to limit &
withstand 63 kA.
• Be able to perform circuit breaker
functions, but not meet all the ANSI
requirements for a circuit breaker.
• Provide current pulses with up to 10
kA instantaneous peaks following a
limiting action –let– through current
for downstream coordination.
• Limit inrush current for a reclosing
operation and soft start up. An
example is the inrush current of an
energizing transformer.
• Reclose after a fault.
• Close into a capacitor.
• Keep size to a minimum. Utilities
requested to keep the first model
SSCL, rated 1200 A, 3-phase 15 kV
to the standard 15 kV metalclad
switchgear
dimensions:
approximately 36” wide X 96” High
X 102” Long.
• Meet all ANSI dielectric-testing
requirements for circuit breakers.
The original specification called for limiting
only 40 kA fault currents, but in December
2001 the input from a major utility related
that they were already purchasing equipment
rated 63 kA due to a continuously raising
fault level on their system. This is a good
example of the growing need for the SSCL.
load
switch
Figure 8-17
A Hybrid Solid-State Switch Using
Mechanical Switch for Regular Conducting
and PTC for current limiting
EPRI SCR Based FCL
The progress to date on the development of
a Solid State Current Limiter (SSCL)
sponsored by EPRI and being developed at
Powell’s Watsonville, CA facility is being
summarized here. The goal is to create a
three-phase, 15 kV, medium voltage device
rated at 1200A that will limit system fault
currents and the damage that could incur on
downstream devices. This is a continuation
of the development program for Solid State
Current Limiters at voltages from 15 to 138
kV. The unique design of the Solid State
Current Limiter (SSCL) would enable it to
limit the current based on its rate of rise as
opposed to its amplitude. This feature would
provide an instantaneous (sub-cycle) current
limit so that the system never has to sustain
the full impact of a fault current. The SSCL
would interrupt the current by rapidly
turning off its conducting components and
inserts an energy absorbing resistor into the
circuit to limit the fault current and
eventually stop the current flow.
8-13
Power Semiconductor Selection
There were basically four different power
semiconductors that were evaluated – GTO,
IGBT, IGCT and the thyristor. A number of
circuit topologies were examined to provide
the optimum cost/performance ratio. The
key factors were availability, proven
reliability, application demands and cost.
The thyristor was selected since it met all
these parameters.
The Power Circuit
The power circuit design was one that was
conceived and tested at 480 volts in the late
1980’s. The testing was done at a major
Canadian utility test laboratory. It performed
as predicted. Since that time, patents have
been issued in the U.S. and the top eleven
industrial nations. The power circuit
diagram is as shown in Figure 8-18.
Figure 8-19
Gate Drive Board
In normal operation mode, the current flows
through TH1 and TH2 or TH3 and TH4
depending on the direction of the current’s
flow. Initially as current flows, the
capacitors charge, and the “Y” terminal
holds a positive charge while the “X”
terminal holds a negative one. When a fault
current is detected, a “Turn On” pulse is sent
to fire the commutating SCRs (TH5, TH6,
TH7, TH8). The discharge current through
TH 5 & 6, C1, and L1 build up, in order to
exceed the load current, IL, on the main
SCRs. The current on the main SCRs, IL, is
eventually reduced to zero, and the excess of
the commutating impulse current, IC, flows
through TH2. When TH2 turns ON, the
voltage across TH1 appears as an inverse
voltage, and it turns OFF. After reaching its
peak, the commutating current, IC, starts to
decay, and the capacitor charges with the
opposite polarity (“X” positive with respect
to “Y”). The commutating current then
switches into the diode and resistor where it
Figure 8-18
Single Section’s Power Components
The four main thyristors, or Silicon
Controlled Rectifiers (SCRs), are controlled
by four corresponding gate driver boards
(Figure 8-19). Each gate driver is fiber
optically connected to each other and uses
firing pulses to turn each main SCR on. A
current transformer monitors the bus’
current and sends the status of each SCR
back to the main phase controller. This
controller sends signals to the gate driver
boards to fire on the SCRs, depending on the
current output.
8-14
dissipates energy. Within a half-cycle the
current crosses 0 and the commutating SCRs
turn off.
controls (SCADA). The fault detector
generates a tripping signal to activate current
limiting action. Section controllers provide
firing pulse distribution both for main and
commutating thyristors, supervise cooling
and
pre-charge
systems,
monitor
temperature and provide auxiliary power
distribution. The se general functions are
illustrated in Figure 8-20.
The re-closing process begins by slowly
phasing back the main SCRs to allow a
short, limited pulse of current through the
system. High firing angles are sent to the
SCRs to allow a minimal amount of current
to flow through the circuit. The limiter
determines the line impedance from the
current flow and if it is zero, or virtually
zero, the fault current still exists and the
main SCRs remain off to prevent the current
from flowing. If there is impedance on the
line, the limiter calculates the current and if
the fault clears, the firing angle is slowly
decreased to let more current through until
eventually the four main SCRs are
conducting in normal operation mode.
Figure 8-20
SSCL general Control Scheme
Functional Description of the SSCL
Operation
Controls
Normal Operation
The control system provides both control
and supervisory functions for the SSCL. It
generates
firing
signals
for
all
semiconductor devices, determines when the
fault
occurs,
provides
temperature
monitoring, implements protective relaying,
interfaces with external controls (system
control and data acquisition or SCADA). In
other words, it forms the brain and nervous
system of the SSCL. The major components
of the control system are: current sensor,
voltage sensor, main or central controller,
fault detector, digital input and output (IO)
interface and section controllers. The current
sensor provides proper current signal both
for normal operations and for fault
conditions operations. In order to provide
fast current limiting, it should have a fast
response and not be effected by
asymmetrical currents. Voltage sensors are
required for providing firing signals and
determining system conditions. The main
controller processes voltage and current
signals, generates firing signal for the main
thyristors, implements over current relaying
functions and interfaces with external
Under normal operating condition only the
main thyristors are fired. Thus the SSCL
acts as a giant voltage regulator; the firing
angle of the thyristors controls its output.
However, that type of voltage/current
regulation
mode
has
an
inherent
disadvantage of having high harmonic
content. So it is used only during soft start,
controlled let through current after fault
interruption and, possibly, during shutdown.
The main controller monitors the load
current and thyristor temperature and sends
out a fan speed command to the section
controllers to provide adequate cooling. The
main controller also performs overcurrent
protection (50/51 functions) and shuts the
SSCL if an overcurrent condition is present.
Soft Start
Upon receiving a start command, the main
controller issues a firing signal with varying
delay angle α. It starts with α close to 180
degrees. Therefore, even for a short circuit
condition, the current will be small initially.
8-15
into the current limiting mode. The firing
pulses are removed from the main thyristors
and commutating thyristors are fired.
Then the firing angle is gradually decreased
while current is monitored. By knowing the
relationship between the firing angle and the
current, the controller determines the
impedance and if it is below a safe value, the
controller aborts the starting procedure and
blocks flow. Otherwise, the controller
advances the firing angle to 0 degrees,
which corresponds to continuous conduction
(“closed switch”) and switches to the normal
operating mode.
The current switches into commutation
thyristors and, at first it flows through the
capacitor. The capacitor begins to charge up
from the fault current and after its voltage
changes polarity and becomes significant,
the current gradually switches into the
resistor. Current eventually drops down to
zero shortly after the voltage crosses zero.
At this point commutation thyristors turn
off.
Running Mode
While the current is below a safe level, the
system continues to run in the normal mode.
During this time the controller monitors load
current and adjusts cooling fan speed
accordingly. If the current exceeds a safe
level, the controller will shut itself down
(implementing overcurrent protection). The
controller also monitors each section
checking the temperatures, fan speed,
commutating capacitors voltage and
condition of the commutating thyristors
among other parameters. If an abnormal
condition is found, an alarm is produced. If a
serious malfunction is found (normally this
would be a second contingency) the unit will
shut down.
Next stage is conducting let-through or
coordination current using delayed firing of
the main thyristors. Again, we start with a
delayed firing pulse, determine fault
impedance and then reduce firing to provide
the required let-through current for a
specified number of cycles to give
downstream protective relays time to detect
and isolate the fault.
During the next stage we recharge the
commutating capacitors (it can take from 10
to 30 sec) and we are ready to re-close the
unit.
Harmonic Distortion
Any sold state equipment can produce
harmonics distortion. In the case of the
SSCL there are two separate modes of
operation: normal ON mode and current
limiting let through current mode.
Stop Command
Upon getting a stop command, the main
controller will shut down. It can be either a
soft stop (the firing angle is gradually
increased) or a hard stop, when the firing
pulses are removed from main thyristors.
Soft stop is advantageous when operating
capacitor banks since it allows the discharge
of capacitors and thus eliminates the
customary discharge time before re-applying
voltage.
Normal ON Mode
In a normal ON mode the main thyristors are
conducting all the time. The voltage drop
across them has two components: threshold
voltage whose polarity is determined by the
current direction and resistive voltage drop.
The latter does not produce any harmonics
(it is a linear component). To estimate the
effect of threshold voltage we can assume
that typical threshold voltage is 1.5 V, we
have six thyristors in series per phase and
Current Limiting (Fault Operation)
Upon detecting a fault condition by the fault
detector (either a current or a combination of
current and di/dt too high,) the system enters
8-16
selected for this function. This was verified
with a test circuit made expressly for this
project. A design for the medium voltage
device indicated that it could be fitted in a
standard cubicle for metalclad switchgear.
This is an important achievement for both
size and for meeting challenging cost goals.
In 2003 a single-phase prototype was
constructed and subjected to voltage tests,
which were successful. In 2004 some
modifications were made to the design to
avoid a materials problem, to reduce partial
discharge levels and to improve the
mechanical strength of the assembly.
the line to neutral voltage is 8 kV (13.8 kV
line to line voltage). The Nth harmonic due
to thyristor threshold voltage is
AN =
6 Vth
Eq. 8-1
2 N
The biggest value is the third harmonic; its
value is 2.1 VAC, so harmonic distortion in
the normal operating mode is below 0.1%
and for practical purposes we can neglect it.
Let Through Current and Soft Start
During let through current and soft start
phases, the thyristors conduct during only
part of each half cycle and therefore
harmonic content is much higher. The
harmonic contents of the current depends on
conduction angle β, which in turn is
determined by the ratio of required let
through current to the maximum available
short circuit current.
I RMS
2
=
I SC
π
2 + cos 2 β
3
β − sin 2 β
2
4
Eq. 8-2
Figure 8-21
Single-phase Prototype of the Solid-State
Current Limiter (SSCL)
Where:
β changes from 0 to 90 degrees,
The mechanical design of the single-phase
prototype of the SSCL was modified in
several ways as a response to tests and to
further thinking about support. One of the
reasons for modification was the corona
level, which was read. Some of the extra
activity was traced to the corners of the heat
sinks for the power thyristors. Some small
modifications reduced the partial discharge,
but it was found that the cooling fans were
the biggest source. These have been
replaced with slightly different models.
IRMS = required let through current,
ISC = available short circuit current.
Though the harmonic content in this mode
can be substantial, we need to keep in mind
that duration of both soft start and let
through current modes of operation is just
few cycles, so no serious harm (additional
losses and heating, harmonics torque in the
motors, etc.) can be done.
Progress Report
Because of the heavy weight of the SSCL
and the fact that the three phases are to be
stacked vertically, it was decided to improve
the strength by using the stacks of power
thyristors and heat sinks as support columns
to carry weight. By the end of 2004, a
Single-Phase Prototype
A breadboard of the standard module,
Figure 8-21, has succeeded in interrupting
63 kA RMS Symmetric using components
8-17
were very enlightening on one phase of the
limiter, and now the next step in the
development of the SSCL is to thoroughly
test the performance of the three-phase unit.
Test plans are already in place to conduct a
three-phase BIL and partial discharge test
within the metal enclosure. These design
tests will help verify that the power
electronics, fans, and other devices inside
the limiter will not adversely affect the
dielectrics of the three phase device. The
temperatures of the SCRs and various
energized components will be measured
during the upcoming continuous current
tests which will be performed on the threephase prototype in early 2006. Additionally,
the three-phase continuous current tests will
provide more information on the air flow
and temperatures within the device and
determine if the enclosure design or
ventilation equipment will have to be
altered. Furthermore, additional capacitor
discharge current limiting tests are being
designed and planned to prepare the limiter
for high power lab testing at KEMA.
Extensive high power tests at KEMA are
planned this year in order to verify the
current limiting performance in the presence
of an actual fault as opposed to a capacitor
simulated fault. In October 2006, EPRI has
reported that the prototype has successfully
passed all the scheduled tests. Sothern
California Edison (SCE) with California
Energy commission (CEC) funding is
planning a field demonstration of this
refurbished unit in 2007.
single-phase prototype was constructed,
modified and successfully tested.
Three-Phase Prototype
In 2005, a three-phase prototype (Figure
8-22) was designed, assembled and some
initial tests were performed to check the
operation of power electronics (EPRI
2005b). Design tests including BIL and
partial discharge have been successfully
implemented to assess the dielectric stability
of the structure. The power electronics
within the device have proved to conduct up
to 1200 amps of continuous current, while
the control system has successfully fired all
12 SCRs.
Superconducting Fault Current Limiters
A
Fault
Current
Limiter (FCL)
can be applied
(EPRI. 2004a)
to reduce the
available fault
current to a
lower,
safer
level where the
Figure 8-22
Three-phase Prototype of the Solid-State
Current Limiter (SSCL)
The plan this year was to validate the
functionality of the Solid State Current
Limiter’s through additional tests in order to
simulate real world industrial and utility
applications. The tests performed in 2005
8-18
FCLs employing High
Temperature
Superconductors can
provide the necessary
current
limiting
impedance during a
fault condition, but
have essentially zero
impedance
during
normal grid operation.
existing switchgear can still protect the grid.
FCLs employing High Temperature
Superconductors (HTS) provide the
necessary current limiting impedance during
a fault condition, but have essentially zero
impedance during normal grid operation.
Therefore, HTS FCLs have no negative
impact on overall system performance, in
contrast to other conventional current
limiting devices, such as a current limiting
reactor (CLR) that produces large voltage
drops, circulating currents in transformers
and substantial energy loss.
Superconducting Current Limiter (SCCL)
Operation
A superconducting state is a state where an
electrical conductor exhibits no electrical
resistance if the current flow through the
material is below a certain threshold (the
“critical current level IC”), when operating
below certain temperature and external
magnetic field (the so called “critical
temperature TC” and “critical field HC”
range). Figure 8-23 shows a measured I-V
curve of a typical superconductor.
Figure 8-23
IV Curve of a Superconductor
As
The quenching of a
illustrated
and
in
the superconductor
subsequent
recovery
to
a
figure, the
superconducting
state
supercond
corresponds
to
a
“variable
uctor
shows no resistance” effect is ideal
for
current
limiting
electrical
applications.
resistance
when the
current is below the critical current level. If
the current exceeds this critical level
however, the superconductor will undergo a
transition from its superconducting state to a
resistive state. This transition is termed
“quenching”. As long as the heat generated
(I2R loss) during the resistive stage does not
damage
the
superconductor,
the
superconductor can be brought back to its
superconducting state if sufficient cooling is
8-19
provided to dissipate the heat quickly to
lower the temperature of the superconductor
to within its critical temperature range.
In addition, superconductor quenching can
occur if one or any combination of the
following three factors exceeds their
corresponding “critical level:”
•
•
•
Operating current level
External magnetic field
Operating temperature
The surface
high-temperature
plots shown in A
superconducting
(HTS)
Figure 8-25
material
operates
near
give a more
the liquid nitrogen
complete
picture of the temperature (77K), as
compared to a lowinterdependen
cy
among temperature
these
three superconducting (LTS)
factors for a material that operates
near liquid helium
typical
temperature (4K).
superconducti
ng material.
As long as those three factors are within the
“critical surface,” the superconductor is in
its superconducting state. If any of the three
parameters goes above that surface, the
superconductor transitions to a resistive
state. A superconductor, once quenched, can
be brought back to its superconducting state
by changing the operating environment to
within the boundary of its critical current,
temperature and magnetic field range,
provided that no thermal or mechanical
damage was done during the quenching of
the superconductor. Manipulating properties
of an HTS material is much easier because
of its higher and broader operating
temperature range, and because HTS has
much higher tolerance to thermal instability.
Figure 8-24
Using Superconductors as a Fault Current
Limiter
Figure 8-24 shows a potential fault current
limiter
incorporating
the
“variable
resistance” feature of superconductors. Such
a device can be designed so that under
normal operating conditions, the peak of the
AC current level of the power transmission
and/or distribution network is always below
the critical current level of the
superconductors, therefore no I2R loss will
result during the process and no or very little
voltage drop across the device (if the device
is designed as non- or near non-inductive).
This device is then essentially “invisible” to
the grid. When the fault occurs however, the
fault current level exceeds the critical
current level of the superconductors,
creating a quenching condition. The
superconductors are forced to transition to
their resistive states, thereby introducing the
necessary current limiting impedance Z0 into
the grid to limit the fault current. In the case
shown in Figure 8-24, the parallel-connected
inductor provides most of the impedance,
since the HTS element is in a high resistive
state and most of the fault current passes
through the parallel inductor.
8-20
Even though there are
no industry standards
regarding HTS FCLs,
they can be put into
three major categories
based
on
the
characteristics of the
fault current limiting
impedance the device
can provide to the
electric power grid
during fault, namely
resistive, inductive and
resistive/inductive
hybrid type of HTS
FCL.
wires,
while
others utilize
mainly
bulk
HTS materials
or films. The
yttriumbarium-copper
oxide (YBCO)
film
on
sapphire wafer
design is an
example of a
resistive type
FCL. The twincoil shielded-core FCL concept is in a
resistive and inductive hybrid nature.
Figure 8-25
Critical Surface of a Superconductor
As described earlier, manipulating current
change can create quenching and subsequent
recovery of superconductors. By the same
token, mechanisms altering the operating
temperature and/or magnetic field level can
be put in place either as a catalyst or an
assistant to achieving fast quenching and
recovery of a superconducting device to
accomplish a “variable impedance” effect.
All these principles can be utilized to design
a comprehensive superconducting fault
current limiting device.
Shielded Core
The very first SCFCL ever installed in the
field in 1996 was manufactured by ABB
(Chen, et al. 1997.). It was of the “shielded
core” type, where the superconducting
element is not physically connected into the
power circuit but coupled into it by means of
a series transformer. The principle is
depicted in Figure 8-26. In particular, the
secondary side of the coupling transformer
is a single turn of SC material. Its advantage
is two fold: 1) No current leads are required
into the cryogenic environment which
substantially reduces the refrigeration
requirements, and 2) with the additionally
free parameter of the turns ratio between the
line side winding and the SC side single-turn
the SC material is better utilized as a highcurrent device which reduces the hot-spot
problem. Although the ABB device worked
very well during a one-year endurance test
in a Swiss power plant the concept was
finally abandoned since it requires
approximately four times the size and
weight compared to the pure “resistive” type
SCFCL (Paul, et al. 2000.). There are still a
few small, mostly university based academic
projects active that utilize the “shielded
core” type, but the prospect of this type of
SC limiter to be economically competitive is
HTS Fault Current Limiter
Developments
Ever since the discovery of hightemperature superconductivity in the mid1980s, HTS fault current limiting devices
for power transmission and distribution
systems have been a major research and
development area of interest throughout the
world, especially in Europe and Japan.
Prototypes based on various HTS FCL
designs have been attempted. The
participants of those prototype projects have
usually included HTS conductor vendors,
academic institutions, government agencies,
and major switchgear and power equipment
manufacturers. (Leung. 2000, Leung et al.
2000, and Teklesadik et al 1999) Among
different HTS FCL designs, some use HTS
8-21
still extremely small1. When a fault occurs
and the current rises above the critical
current RSC increases rapidly and the fault
current commutates nearly completely into
the parallel impedance. During that
transition the voltage across the device may
be somewhat higher than the voltage after
the quench depending on the loop
inductance of the parallel circuit.
very low. Therefore, this type will not be
discussed further in this report.
Figure 8-26
Principle of the “Shielded Core” Type SCFCL
(EPRI 2005)
Resistive Type
The most compact SCFCL design up to date
can be achieved by the so-called “resistive”
type design. It employs the superconducting
material as the main (load) current carrying
conductor under normal operation. In AC
applications, the superconductor is therefore
subject to AC losses which, together with
the losses in the current leads, are the major
loss components in normal operation (plus
the no load heat losses through the cryostat).
When the fault occurs, the superconductor
quenches which increases its resistance by
several orders of magnitudes. This highly
non-linear increase in resistance requires an
impedance element to be provided in
parallel to the superconductor in order to
avoid its thermal destruction. This element
also avoids excessive over voltages from the
power network (line inductance) since the
superconductor alone would, like an ideal
switch, effectively “turn-off” the fault
current within the first half cycle.
Figure 8-27
Resistive Type SCFCL Principle With Shunt
Element Completely in the Cold Environment
It shall be noted that the impedance
characteristic of the resistive type SCFCL
after a quench is essentially governed by the
shunt element. Therefore, it is possible that a
“resistive” type SCFCL may introduce
significant inductance into the power system
during a fault if the shunt element is highly
inductive. Various “resistive” type SCFCL
projects and some of their specifics are
discussed below.
Fault Current Controller (FCC)
Semiconductor devices with only one p-n
junction diodes exhibit the smallest on-state
voltage drop of all semiconductor switches
(amongst devices of the same semiconductor
base material such as silicon, of course).
The principle of a resistive type SCFCL is
depicted in Figure 8-27. The superconductor
is represented by R in parallel with the
resistive and/or inductive shunt element
RP/LP. During normal operation the line
current flows entirely through the
superconductor, i.e. iSC = iLINE, while RSC is
1
Typical values for the voltage per unit length in the
superconducting stage (≤ 1 µVcm-1) versus the fully
quenched state (0.1…15 Vcm-1) together with an
estimate of the voltage drop across the SCFCL
required for limiting the current (0.2…1 times the
line voltage) yields a very small value of (4x108…10-5) pu for RSC during normal operation (in the
superconducting stage)
SC
8-22
the inductor several modified versions of the
circuit have been, and are currently,
investigated which utilize thyristor phase
angle control for maintaining the DC bias
current. A 15 kV class single-phase
experimental setup utilizing this method and
using a non-superconducting inductor was
also successfully tested at Los Alamos
National Laboratory (Boenig et al. 2002.).
However, a diode cannot be turned off in
forward direction. This limitation can be
overcome when the diode is biased with a
constant (DC) current flow in forward
direction in the circuit depicted in
Figure 8-28 and described in (Boenig and
Paice. 1983.). The DC voltage source VB
keeps all four diodes biased in forward
direction by I0/2. Therefore, the AC line
current (iLINE) can pass through the parallel
diode path formed by D1-D3 and D4-D2 for
both AC half cycles. However, if the
instantaneous value of iLINE exceeds I0 then
D3 and D4 will block during the positive
half cycle and D1 and D2 during the
negative half cycle. Therefore, the AC
current has to flow through D1-L-D2 during
the positive half cycle and through D3-L-D4
during the negative half cycle, which
effectively inserts the impedance of the
inductor L in to the AC, circuit. In order to
minimize losses in the inductor L, it is
desirable to make it a superconducting coil.
In addition, the fact that the inductor only
exhibits DC current (plus some small AC
ripple, probably) makes superconductivity
the ideal choice for the coil technology since
no AC losses occur like in any of the
resistive type SCFCLs.
Figure 8-28
Principle of the Diode-Bridge FCL With DC
Biased Coil and External DC Voltage Source
SuperPower participated in a U.S.
Department of Energy Superconductivity
Partnerships with Industry (SPI) program
led by General Atomics in building and
testing a 15kV 30MVA HTS fault current
controller (FCC). This device is based on a
thyristor-bridge concept that combines
power electronics and HTS coils to achieve
a “variable impedance” effect, Figure 8-29
and Figure 8-30. This design is an example
of an inductive FCL.
The disadvantage of the circuit shown in
Figure 8-28 is that the power electronic
devices cannot interrupt the AC fault
current, thus requiring a circuit breaker in
series. This limitation can be overcome by
using thyristors instead of diodes in the
circuit. Since the fault current can be
adjusted by means of thyristor phase angle
control, such an arrangement is called a fault
current controller (FCC). A 15-kV/1.2 kA
class three phase device utilizing HTS
superconducting DC bias coils was built and
tested successfully at Los Alamos National
Laboratory (only at single phase operation
after repair of dielectric failures that
occurred during initial three phase tests).
While this device still used a DC power
supply to provide the bias current through
In Japan, TOSHIBA has successfully tested
a 66 kV class superconducting DC coil.
Although no FCL or FCC project is directly
associated with the coil development it is
clearly aimed for a power electronic based
FCL system (Yazawa, et al. 2004.).
8-23
melt cast process (MCP) for BSCCO-2212,
these are manufactured by cutting
superconducting tubes to bifilar coils. At the
operation temperature of 65K a current
density of 4000A/cm2 was achieved. In order
to protect the superconductor during fault
current limitation, the component was
equipped with an electrical shunt contacted
on its entire length. Single-phase tests were
first completed with nine of these
components in series, corresponding to a
protected load of 1.2MVA. These initial
tests include different types of short circuits
as specified by the utilities within the
project, and included lightning surge loads
up to 75kV. In May 2004, it was announced
that a successful implementation of the
10kV, 10MVA SFCL into the energy grid
for RWE Energy was completed in Netphen
near Siegen, Germany.
Figure 8-29
Schematic Showing Keys Components of
FCC and Location in Typical Network. V is
the Voltage Source, L is the Source
Inductance, Br is the Breaker, and BPS is the
Bias Power Supply. From Waynert, et al.
(2003)
The field test was stopped after one year of
Reliable operation and successful test (but
no short circuit during that time) in March
2005. Up to date the CURL10 device has
been the most powerful SCFCL ever tested
in the field. Figure 8-31 depicts the device
and the cold mass.
Figure 8-30
An Example of the Response of the FCC to a
Fault Applied at 0.18 s When the Phase Delay
Angle is Set for 120 Degrees. For Times Less
Than 0.18 s, the FCC Was Responding to a
1 MW Load (175 A Peak Current). From
Waynert, et al. (2003)
CURL10
The German government recently funded a
resistive superconducting fault current
limiter project known as CURL 10. This
FCL is based on bulk material and aims at
the development of a three-phase prototype
for the medium voltage level (10kV,
10MVA). (Bock et al. 2004) The key
element of the project is the development of
suitable
robust
superconducting
components. On the basis of the well-known
Figure 8-31
CURL10 Device (a), With Cold Mass
Removed From the Cryostat (b)
Amongst the partners involved were (Bock
et al. 2005):
8-24
•
order to avoid hot spots during the
current limitation phase. The loop
inductance between the HTS element
and the shunt is negligible and does
not cause any measurable transient
over voltage during the fast
transition of the current from the
HTS material to the shunt. However,
the metallic shunt material limits the
maximum electric field strength
during the quench to approximately
0.6 V/cm.
ACCEL as the project coordinator
was responsible for cryostat, cooling
and system integration.
• Nexans SuperConductors developed
and
manufactured
the
superconducting components used in
the test device (ATZ Adelwitz
developed components based on an
alternative material option YBCO).
• The two largest German utilities,
RWE and E.ON defined a set of
specifications,
organized
the
laboratory tests at FGH Mannheim,
and RWE installed the demonstrator
in the grid.
• Forschungszentrum
Karlsruhe
supported
the
component
development
by
testing
and
characterizing material and single
components and contributed to the
important electrical insulation.
• In addition there was ATZ (YBCO
development), EUS (power system
simulation) and ACCESS (FEM
simulation) as project partners. This
project was funded by BMBF, the
German ministry of education and
research.
As
shown
in
Figure
8-32
the
superconducting elements employed in the
CURL10 device exhibit the following
distinct features:
•
•
Figure 8-32
CURL10 HTS Elements (EPRI 2005)
Despite the very successful tests of the
CURL10 device the developers conclude
that “the challenge remains to develop a
limiting system for the transmission level
(100 kV) where alternative technologies are
completely missing and an economically
viable application is very likely. Although
technically probably feasible, a system
based on the presently developed
components (bifilar coil) will not be viable
from the economic point of view (3000
components for a 110 kV/350 MVA system
and high voltage issues).” (Bock et al.
2005). But, very promising novel concepts
based
on
magnetic
switching
of
superconducting properties are in the
development stage.
The current flows in a bifilar coil
structure, which is machined from
the raw BSCCO 2212 tubes. This
guarantees long conductor length for
voltage build-up within a highly
compact geometry. It also ensures a
low-inductive design of the device
required to minimize the voltage
drop during normal operation.
Along the entire length of the
superconducting material it is
bonded directly to a sheet-like shunt
resistor made of a Cu/Ni-alloy in
ABB’s Resistive SCFCL
For over 10 years ABB Switzerland has
pursued R&D on superconducting FCL
8-25
devices based on their own BSCCO 2212
material development. After the successful
test of a shielded iron core type SCFCL in
1996 (see project description above) the
R&D focused on the development of a
resistive type FCL. Finally, in 2001 ABB
tested its 6.4 MVA single-phase resistive
SCFCL device at 8 kV in the ABB Power
lab in Baden, Switzerland (Chen et al.
2002). Instead of tubes (as in the CURL10
device) this design used BSCCO 2212 HTS
elements cast into plates and machined into
a meander form. Stainless steel is used as
the shunt material bonded to the HTS
material for hot-spot prevention. Figure 8-33
depicts one of many single BSCCO 2212
meander plates (a), which compose a
complete SCFCL module (b). Since 2001
no major development on SCFCL has been
reported by ABB.
for FCL applications. Apparently, the high
cost for the YBCO thin film HTS elements
make
this
technology
economically
unattractive.
Figure 8-34
Siemens 1.25 MVA SCFCL
CESI Project in Italy
In 2003, an Italian R&D project named
LIMSTAT and sponsored by CESI was
started. It involved design, manufacturing
and testing of superconducting FCL
prototypes (See Figure 8-35). It is a resistive
based design and the HTS elements are
composed of BSCCO-23 and MgB2. They
plan to develop a 400kVA 1-phase and
1MVA 3-phase prototypes. Short circuit
testing has been carried out at CESI facility.
Prospective symmetrical and asymmetrical
fault currents 40 times larger than the
nominal current value has been applied for
40-160 ms on single phase devices. As a
result prospective short circuit current of 15
kA peak has been reduced to 800-2200 A in
testing.
Figure 8-33
ABB’s BSCCO 2212 Meander Plates (a) Are
Stacked to Compact Modules (b) (EPRI 2005)
SIEMENS Resistive SCFCL
A 3-phase superconducting fault current
limiter with 1.25 MVA protected power at a
rated voltage of 7.2 kV and a rated current
of 0.1 kA was built and tested by Siemens in
2003.
Figure 8-34 depicts the device and its HTS
components (Neumueller. 2004.). This
device is based on YBCO thin film
technology (Kraemer, et al. 2003.).
Although this technology has proven to be
very effective for FCL applications, Siemens
has discontinued this project and does no
longer pursue YBCO thin film technology
8-26
Figure 8-35
CESI SCFCL prototype
Super-ACE Project in Japan
One of the items included in the Superconductive AC Equipment (Super-ACE)
project in Japan was development of a 66
kV/1 kA class high-Tc superconducting
(HTS) fault current limiter (FCL) magnet
(Figure 8-36). The design of the magnet and
test results may be found in Yazawa et al,
2005. The magnet mainly consists of a
vacuum vessel, a nitrogen bath, a pair of
current leads, cryocoolers, and six sets of
coils wound with Bi2223 tape. The rated
current of the magnet is 750 A. The
insulation voltage of the magnet is of the 66
kV class. In the final year of the project, all
six sets of the coils are set connected in the
cryostat and evaluation tests have been
implemented. In the cooling test, sub-cooled
nitrogen of 65 K was obtained, with
homogenous temperature distribution in the
cryogen. The rated current of 750 A was
successfully obtained for both direct and
alternating current tests. In addition, the
magnet passed the simultaneous current and
voltage application test. Finally, the
dielectric test results showed that the magnet
satisfied the insulation for 66 kV apparatus
in the Japanese Electrotechnical Committee
Standard (JEC) standard.
Figure 8-36
66kV/750 A Magnet
Matrix Fault Current Limiter
Concept
To address the market need for an economic
transmission-level HTS FCL, the Matrix
Fault Current Limiter (MFCL) was
developed by SuperPower, in conjunction
with Nexans. By using Melt-Cast Processed
(MCP) BSCCO-2212 HTS elements, the
MFCL provides a solution, which is more
economical than many conventional
solutions to breaker over-duty problems.
Numerous utilities have expressed the need
for a device that can economically address
breaker over-duty problems in an overstressed transmission network.
The MFCL employs “Matrix” technology
that offers modular features to scale up to
transmission voltage levels of 138kV. The
MFCL consists of individual modules that
contain High Temperature Superconducting
(HTS) elements and an inductor connected
in parallel that carries the current during the
fault. The HTS elements consist of bulk
BSCCO-2212 material, and are fabricated
using Nexans’ Melt-Cast Process (MCP). A
number of these modules are arranged in an
M x N array to form the current limiting
8-27
matrix. The milestone driven program
includes the fabrication of three prototypes,
a proof-of-concept design, the Alpha
prototype, and the Beta prototype, each with
progressively higher ratings to scale up to
the transmission voltage level. The Beta
prototype will be designed to meet a specific
utility (AEP) application and will be
installed on a 138kV transmission grid for
demonstration in 2007.
Figure 8-37
Most Basic Configuration of MFCL
Electrical Configuration of an MFCL
As shown in Figure 8-37, the most basic
form of an MFCL device includes a trigger
matrix arranged between node A and node B
in series with a current limiting matrix
arranged between node B and node C. The
primary function of a trigger matrix is,
during a fault situation, to create a magnetic
field that is sufficient enough to “trigger”
the
quenching
of
superconducting
components in the current-limiting matrix.
The primary role of the current-limiting
matrix is to provide the majority of the
overall required current-limiting impedance
during the fault.
Figure 8-38
Block Diagram of MFCL Matrices
Figure 7-24 shows a schematic diagram of a
current-limiting matrix that includes “m”
number of current-limiting modules
electrically connected in series between
nodes B and C of an MFCL. Each module
further includes “n” number of currentlimiting matrix elements (F-11 through F1n, ., F-m1 through F-mn) electrically
connected in parallel. The current-limiting
matrix is therefore an m x n matrix. Each
current-limiting matrix element includes a
parallel electrical arrangement of a
superconductor R and an inductor L. For
example, current-limiting matrix element F11 of module 1 includes a superconductor R
arranged in parallel with an inductor L . As
another example, current-limiting matrix
element F-mn of module m includes a
superconductor Rmn arranged in parallel with
an inductor Lmn. A most important feature of
the MFCL design is the passive way
magnetic field is generated to “trigger” the
quenching of the superconducting elements
Figure 8-38 illustrates a high-level block
diagram of the MFCL comprising of a “1 x
N” (column x row) trigger matrix and an “M
x N” current-limiting matrix. The trigger
matrix includes “n” number of trigger matrix
elements (T-1 through T-n) while the current
limiting matrix contains “m” number of
current-limiting modules. Each currentlimiting module consists of “n” number of
current-limiting elements (F-11 through F1n... F-m1 through F-mn). Each trigger
matrix element is to trigger “m” number of
current-limiting elements that have the same
row number. For example, trigger element
T-1 is to “trigger” current-limiting elements
F-11 through F-m1 in modules 1 through m.
11
11
8-28
in the current-limiting matrix. This feature is
embedded in the physical relationships of
the elements relative to one another, of
superconductors and inductors within the
trigger and current-limiting matrices.
The superconducting components in an
MFCL device are in the form of noninductive tubes. The inductors are fabricated
from
non-superconducting
electrically
conductive materials, in this case copper,
and are formed solenoid coils. The MFCL
matrix assembly is housed in a cryostat that
contains
liquid
nitrogen.
The
superconductors in the MFCL device are
maintained at the superconducting state by
being cooled to below their critical
temperature, with cooling provided within
the cryostat supported by external
cryocoolers.
•
•
•
Proof-of-Concept MFCL Design
Figure 8-39
Schematic of MFCL Current Limiting Matrix
Design Requirements
Several other observations can be made
regarding the features and design concerns
of an MFCL device:
•
operation. This makes fast recovery
of an MFCL device to its
superconducting
state
more
attainable. It also protects the
superconductor
elements
from
thermal damage.
The simultaneous triggering of all
superconducting components in the
current-limiting matrix ensures that
the voltage is evenly distributed
across
each
superconducting
component. Overcurrent allowance1
The number of rows in an MFCL is
determined by the peak value of the
normal operating current level (plus
whatever level of overcurrent is to be
considered). More rows can be added
to increase the redundancy of the
design so that if one or a few
superconducting components fail, it
will not cause a failure of the whole
device, making an MFCL device
highly reliable. In addition, the
current limiting impedance required
for a specific network primarily
determines the number of columns in
the current limiting matrix, making
an MFCL design highly scalable.
The overall impedance of the trigger
matrix of an MFCL device during a
fault also contributes to the currentlimiting impedance of the whole
device.
As discussed, the MFCL development
program is divided into the demonstration of
a series of hardware prototypes. The first
demonstration, the proof-of-concept preprototype, is focused on demonstrating the
current limiting performance of the matrix
concept. Table 8-2 summarizes the electrical
design requirements for the pre-prototype.
The following design requirements were
The parallel-connected inductors in
the current limiting matrices act as
shunts. The partial divergence of the
surged current to the inductors serves
to reduce the thermal energy the
superconductors absorb during the
current limiting phase of the MFCL
8-29
selected
for
demonstration:
the
proof-of-concept
outer vacuum vessel. Heat is removed from
the bath with two cryocoolers. The external
connection is made to the matrix through a
set of lead assemblies. The lead assembly
penetrates the vacuum vessel and traverses a
vacuum region and then penetrates the
pressure vessel to connect to the matrix in
the Liquid Nitrogen (LN2) bath. The singlephase cryostat is approximately 60 inches in
diameter and 109 inches high. There is an
additional 9 inches of height contributed by
the bushings. The overall assembly weighs
about 3 tons when filled with liquid. The
cryostat is large enough to be refitted later
with higher rated bushings up to 52kV, to
support future tests as needed. The preprototype is presently fitted with 15kV
bushings, which are small relative to the
overall size of the device. Other than the
lead assembly design, it is expected that the
basic approach shown here will be followed
in the subsequent Alpha and Beta designs.
Table 8-2
Pre-Prototype MFCL Specifications
Property
Magnitude
Line-to-line voltage
15 kV
Load Current
800 Arms
Prospective Fault
Current
(Symmetrical)
10 kA
Prospective Fault
Current
(Asymmetrical)
25 kA
Phase-to-ground
voltage
8.66 kV
Overcurrent
allowance1
20 %
Limited Fault Current
(Symmetrical)
8 kA
Fault duration
3 cycles
For a nominal operating current of 800A rms, the
corresponding peak current is 800 * •2 = 1130A
peak. The critical current level of the HTS is
specified to allow for 20% over-current or 1.2 * 1130
= 1350A peak
1
An overview
of the selected When a fault occurs, the
design
is MFCL is expected to
the
provided here demonstrate
capability
to
provide
with
current limiting before
additional
detail
to the first peak of the
fault
follow on the asymmetrical
current.
respective
subsystems in
the
sections
below.
Figure 8-40 shows the main internal
components of the pre-prototype device. The
matrix of HTS elements and parallelconnected inductors are contained in the
matrix assembly immersed in the liquid
nitrogen bath in the cryostat. The cryostat
consists of an inner pressure vessel and
Figure 8-40
Pre-Prototype MFCL for Proof of Concept
Demonstration
Matrix
Design Considerations
In order to meet the system requirements for
the MFCL pre-prototype, the MFCL matrix
design must address several major issues:
•
8-30
HTS
element
design
characterization, including:
and
which amounts to about 1130A. The critical
current level of the HTS tubes was specified
as greater than 1350A to allow for 20% of
over-current that may occur in the grid.
During a fault, the current exceeds the tube’s
critical current level and the tube transitions
to its resistive state, creating a voltage drop
across the element and achieving a certain
level of current limiting performance.
Therefore, to accurately characterize the
voltage development across the HTS tube
under various transport-currents is of most
importance to an HTS FCL design. Such a
relationship is usually expressed in terms of
electric field vs. current density of the HTS
component, which is dependent on the
critical temperature of the HTS material, the
operating temperature, and the self- and
external- magnetic field. Figure 8-41 shows
such a relationship measured by Nexans, for
a MCP BSCCO-2212 rod with a 5mm
diameter.
– Voltage development of HTS
element under fault current and
magnetic field
– Mechanical strength of HTS
element
Electrical design of the HTS current
•
limiting matrix
•
Thermal design of the matrix
assembly, including:
– Minimization of heat generation
from the components of the
matrix
Mechanical design of the matrix
•
assembly, including:
– Electromagnetic force due to
fringe magnetic field and
coupling within the multicomponent assembly
– Thermal-mechanical stresses due
to application of materials
having
different
thermal
characteristics
The following section of this report will give
brief
descriptions
of
the
design
considerations given to these issues.
HTS Element Design and Characterization
HTS elements in tube configuration
fabricated from bulk BSCCO-2212 material
were selected for the MFCL pre-prototype
design. Nexans SuperConductors, GmbH
manufactured these tubes in Germany, using
a Melt-Cast Process (MCP). A description
of the process and the elements are given in
a later section. Consideration must be given
to the voltage development across the HTS
element under fault current and magnetic
field. An HTS tube acts as a “variable
resistance” in an FCL application, as
described above. During normal operation,
when the current in the tube is below its
critical current level, it is fully
superconducting, without any voltage drop
across the element. For a normal operating
current of 800A rms for the pre-prototype,
the corresponding peak current is 800A*•2
Figure 8-41
Magnetic Field Impact on U(l) Characteristics
of the Elements
The individual HTS tube elements were
tested extensively and their current limiting
performance characterized in detail under
various and repeated fault conditions by
Nexans. The elements exhibit excellent
characteristics in areas such as fast transition
from superconducting state to resistive state,
substantial voltage development during
initial fault current rise, and consistent
voltage development during the course of
8-31
current limiting well beyond the 3 cycle
period specified. Figure 8-42 shows one
example of the test results for one typical
element sample. In addition, extensive
testing was conducted using a small mockup assembly incorporating multiple HTS
element and shunt modules to obtain the
voltage development characteristics under
various fault current level and magnetic
field, details of which are described below
under CAPS tests.
where Tc is the critical temperature of
BSCCO 2212 in K. With a = 39.23668, b = 0.856427 and c = 0.004673. E is in mV/cm,
J in A/cm2, and T in K. Figure 8-43 shows
the results of this work plotted against the
actual test data. Further work was carried
out to obtain a much more accurate
analytical description of the test data, the
plot of which is shown in Figure 8-44.
Figure 8-42
Quench Tests on Elements
To assist the design of the MFCL matrix
assembly, an accurate analytical model of
the behavior of the HTS tube is also of
critical importance. One important aspect of
such an analytical model is to simulate the
HTS element voltage development at
different current levels. This relationship, as
pointed out earlier, is usually expressed in
terms of the electric field E (the voltage
across the HTS element divided by the total
length) vs. current density J (total current
divided by the cross section of the HTS
element). Some work in this area has been
done. The following is an example of the
analytical formula in published literature by
Cha:
(
E( J ,T ) = a + b ⋅ T + c ⋅ T
2
)⋅ 10
2 (T −89 )
Figure 8-43
BSCCO-2212 E-J Curve of Published Results
vs. Measured Data
During the
The mechanical strength of
normal
HTS elements is also a
operation,
the current consideration in the design.
limiting
phase and subsequent re-cooling period after
the fault, the HTS element will be subjected
to substantial stress. This comes mainly
from two sources. One is the thermalmechanical stress associated with the HTS
element being heated up during the fault and
being cooled down afterwards. This creates
a cycle of thermal expansion and contraction
causing stress to the element. Another
source of stress comes from the
electromagnetic force due to electric current
⎛T ⎞
1+ 40 ln ⎜ c ⎟
⎝T ⎠
⋅J
Eq. 8-3
8-32
matrix assembly is the trigger circuit, which
is represented by the block. All of the other
components shown in the diagram are part
of the external test circuit. The matrix is
physically connected to the external test
circuit via the lead assembly described
earlier. The matrix assembly contains all
36 current limiting modules and the trigger
circuit. The selection of component values
for this final design of this matrix for the
pre-prototype incorporates the current
limiting behavior of individual HTS element
and considerations from the multi-module
mock-up matrix test results.
in the HTS element interacting with the
surrounding magnetic field.
To address these stress issues, mechanical
stabilization techniques were developed. A
specially designed connection mechanism
and soldering techniques were also
developed to enable reliable interconnects
between the HTS tube, its mechanical
stabilizer and the copper connector. The
main challenge in developing such an
interconnection is to ensure good electrical
contacts among all components, reduced
stress on the HTS element, and also to
achieve very low contact resistance to
reduce the burden to the overall cryogenic
subsystem of the pre-prototype.
Figure 8-45
Simplified Electrical Schematic for PrePrototype Matrix
Development of MFCL Elements
Description of Elements for the Proof-ofConcept Prototype
Nexans started designing and manufacturing
the tubes in August 2003. It was decided to
select tubes with an OD of 26 mm as the
basic element. The requirements were thin
walls, long HTS length, low ohmic contact
resistance, high Jc, excellent Jc - uniformity.
Moreover,
the
tubes
would
need
stabilization in order to withstand the strong
short circuit requirements with respect to
Lorentz forces.
Figure 8-44
Simulation of E-J Curve vs. Test Data
Electrical Design
The current limiting matrix provides the
bulk of the current limiting impedance
required to achieve the performance
specification listed in Table 8-2. The matrix
for the proof-of-concept pre-prototype
contains 36 current limiting modules
connected in series as shown in the
schematic in Figure 8-45. Each module
consists of an HTS element with a shuntconnected inductor. Also included in the
A machining process for the inner and outer
surface of the tubes could be developed
within a very short timeframe. This process
of thinning the ceramic down to a wall
thickness of 1.5mm is very sensitive from
the mechanical point of view. First, shorter
8-33
tubes with a length of 10 cm were
developed.
After
designing
and
manufacturing the suitable copper contacts,
a first approach for stabilization was made.
About thirty elements were fabricated for
the first series of tests at Center for
Advanced Power Systems (CAPS). As a
consequence of these tests the stabilization
had to be redesigned. Nexans developed
new stabilization based on a completely new
concept during November 2003. More then
thirty samples were produced for the second
round of tests at CAPS in December 2003.
The tests were 100% successful and
confirmed in an impressive way the validity
of the new approach for stabilization. In the
beginning of 2004 this design was scaled up
to tubes of 21cm length, which were needed
for the proof-of-concept prototype.
Figure 8-46
Elements for the Proof-of-Concept PrePrototype
The typical contact resistance is 0.3 µΩ per
contact at this development step. The
material inhomogeneity (differences in the
critical current Ic) within one sample and the
scatter between different samples was, in
any case, less than 10%.
The actual design of the elements, which
meet the requirements for the proof-ofconcept prototype, is shown in Figure 8-46.
The design is easy to adapt for other rated
currents and voltages since it is simple to
manufacture and simple to adapt the specific
application requirements. The normalconducting coil around the HTS tube
generates the parallel magnetic field, which
drives the superconductor in a uniform way
into the normal-conducting state during a
short circuit event. This normal conducting
trigger coil is part of the MFCL matrix
element.
The high mechanical forces during the short
circuit and different thermal expansion of
the BSCCO ceramic, the stabilization and
the copper contacts during cooling are a
challenge for the element design. The design
meets all requirements, at least in single
element tests, as verified during numerous
short circuit tests on small and longer
samples. Resistivity was measured at up to
470K on a 10mm rod, providing information
about the material resistivity up to higher
temperatures. The material must withstand
high temperatures during the limiting event.
Figure 8-47 shows that the resistivity of
BSCCO-2212 is linear up to 470K.
As shown in Figure 8-46 the critical current
density of the MCP-BSCCO-2212 material
increases by a factor of approximately three
between 77K and 66K. The critical current,
and also the rated current, can be adapted by
changing the cross-section of the tube.
Considering a safety factor of 20%, the
critical current for one element must be
around 1400A for a rated current of 800A
rms. The same design could be used up to
2400A rms just by reducing the operation
temperature.
8-34
•
Figure 8-47
Resistivity vs. Temperature for a 10mm Rod
•
Magnetic Field Impact on the MCP-2212
Material
Electromagnetic
between coils
HTS
material
characterization
interference
properties
The major advantage of BSCCO-2212
compared to all other HTS materials is that
its critical current can be much more
effectively influenced by external magnetic
fields. This property is essential for the
magnetic field triggered MFCL concept. The
magnetic field impact is shown previously in
Figure 8-41. The right blue curve is the U-I
curve under self-field conditions. The
additional curves are measured on an MFCL
tube under a parallel constant DC field. The
magnetic impact is huge at low E-fields. The
magnetic field impact under perpendicular
field is almost the same.
Test Circuit and Plan Overview
Proof-of-Concept Test Results
Figure 8-48 shows the schematics of the test
circuit used for testing the pre-prototype
MFCL.
Testing of the pre-prototype MFCL was
conducted at KEMA PowerTest, which is
the largest short circuit test facility in the
United States. The lab is located near
Philadelphia in Chalfont, PA. KEMA
employs two generators with ratings of
1000MVA and 2250MVA that are designed
to provide short circuit test current up to
63kA. The KEMA test facility was
originally designed to test 138kV circuit
breakers and is well suited to test
requirements needed for the MFCL
development.
Test Objectives
The focus of the MFCL proof-of-concept
testing was the verification of the current
limiting performance of the device. This was
verified in a hardware test environment that
is representative of the actual power system
conditions. The data collected from the test
was used to evaluate the performance and to
improve the simulation accuracy and
capability of the MFCL analytical models.
The following performance factors were
evaluated:
•
•
•
•
•
•
Figure 8-45 shows the circuit of the current
limiting matrix in the MFCL device, which
electrically is similar to what was tested at
CAPS, but in this case has 36 current
limiting modules connected in series. The
test system consists of the short circuit
power source capable of providing a
prospective symmetrical rms current of
10kA (asymmetrical peak fault current
around 25kA), at voltages up to 8660Vrms.
The supply has the capability to vary both
voltage and system source impedance to
control the voltage and short circuit current
of the MFCL assembly. Figure 8-49 shows
the expected current limiting at the first
peak, based on the system X/R ratio of 30
for the test system at KEMA. This plot
shows the ratio of the limited current to the
prospective test current. The MFCL was
tested up to approximately 25kA prospective
peak symmetrical fault current. Figure 8-49
shows that first peak current limiting
performance ranges from approximately 85
Current limiting performance
Current sharing between HTS
elements and trigger and shunt coils
Dynamic voltage and current
limiting resistance development, R =
f(B, T, J)
Speed of superconducting-to-normal
transition under fault current
condition
Current transfer speed
Effects of trigger magnetic field
8-35
to 70 percent for 25kA prospective peak
current, depending on the input voltage.
The current limiting test procedure was
divided into three current level regions,
based on the prospective peak fault current
ranges of 15kA, 20kA and 25kA. These
prospective peak fault current values were
set by adjusting the system impedance for a
test system voltages ranging from 480Vrms
to 8660Vrms line to neutral voltages. The
lower voltages, from 480V to 2400V, were
used when testing in an open bath
environment. All tests at 4160Vrms and
8660Vrms were conducted in the MFCL
cryostat.
Figure 8-50
MFCL Cryostat in KEMA Test Cell
The test circuit in Figure 8-48 offers
flexibility in the application of the fault and
the ability to provide load current before and
after the faults. The typical switching
sequence is as follows:
1. Start short circuit fault simulation
with Backup CB, Aux CB #1, #2, and
#3 closed at time T0
2. Main Closing Switch (MS) closes the
system supply voltage at pre-defined
switching angle of the voltage signal,
at time T1 ± 0.5 ms
3. All CBs stay closed for the duration
of the short circuit fault in multiples
of 1/2 cycles
Figure 8-48
KEMA Power Test Circuit and Switching
Sequence
4. Auxiliary Vacuum CB (Aux CB #3 )
opens (clears) the short circuit fault
after a pre-set clearing time controlled
by a pre-trigger control signal at
multiples of 1/2 cycle and at current
zero crossing, at time T2
5. After pre-defined load current
duration (Post-Fault Recovery time),
all CBs open at time T3, (sequence
depends on KEMA's preferred
setting)
Figure 8-49
Expected First Peak Current Limiting as a
Function of Current and Input Voltage
8-36
prospective fault levels of 17kA to 25kA
asymmetrical first peak, and again at
8660VAC input voltage. Table 8-3
summarizes the typical level of current
limiting achieved during these tests, which
is in line with the expected performance of
the device. The cryostat system operated as
planned during the 74K sub-cooled tests,
and achieved a stable steady state subcooled cryogenic operating environment and
fast recovery to normal cryogenic operating
environment after fault. Figure 8-51 to
Figure 8-56 show all the recorded
waveforms.
6. If pre-fault load current is required,
the switching sequence uses MS
(closing switch) to close before Aux
CB #2, and Aux CB #2 with a timing
tolerance of ± 2.0ms will be used to
simulate the short circuit fault by
closing at T1 ± 2.0110ms.
The test program is broken into two main
test phases, testing in an open LN2 bath and
testing in the MFCL cryostat. Testing the
HTS matrix assembly in an open LN2 bath
provides ease of access to change parts and
reconfigure test circuits and results in faster
speed in testing. The conditions for these
tests are:
Table 8-3
Results at 4160Vrms and 8660Vrms at 74K
and 1Atm Pressure
•
Constant temperature at 77 K and at
0.1MPa (atmospheric pressure)
• Easy ability to make voltage and
current measurement - record
waveforms
• Test under variable fault current
levels (15kA to 20kA)
• Test at low system voltages up to
2.4kV
The test conditions in the cryostat are:
•
•
•
•
System
Voltage[kV
rms]
Number
of
cycles
1st Peak
Prospective
Current
[kA]
1st
Peak
Limited
Current
[kA]
1st Peak
Voltage
across
assembly
[kV]
1st Peak
Ratio of
Limited to
Prospective
Current [%]
3rd Peak
Ratio of
Limited to
Prospective
Current [%]
5th Peak
Ratio of
Limited to
Prospective
Current [%]
4.16
3.0
17.2
16.4
1.47
95.35
76.62
64.79
4.16
3.0
20.2
18.2
5.15
90.10
58.89
53.01
4.16
3.0
25.6
20.2
6.16
78.91
50.88
46.15
8.66
3.0
17.6
17.0
4.23
96.59
77.78
70.67
8.66
3.0
20.2
18.6
6.16
92.08
72.83
64.71
8.66
3.0
25.6
21.4
7.61
83.71
64.32
55.91
Variable temperature between 74K
and 77K
Variable pressure 0.1MPa and
0.3MPa
System voltage at 4160V and 8660V
Variable prospective peak fault
current levels (15kA to 25kA)
Test Results in Cryostat at 74K
Figure 8-51
Current Limiting at 4160Vrms, 17.2kA
Prospective Fault @ 74K
Figure 8-50 shows the fully assembled and
integrated cryostat system ready for test.
There were two sets of tests with the fully
integrated MFCL cryostat system, one at
77K and one at subcooled temperatures of
74K. System tests were conducted at
4160VAC and 8660VAC input voltage for
both temperatures. First, during the 74K
tests, three faults were applied at 4160VAC
input with three cycle durations and
8-37
Figure 8-52
Current Limiting at 4160Vrms, 20.2kA
Prospective Fault @ 74K
Figure 8-55
Current Limiting at 8660Vrms, 20.2kA
Prospective Fault @74K
Figure 8-56
Current Limiting at 8660Vrms, 25.6kA
Prospective Fault @74K
Figure 8-53
Current Limiting at 4160Vrms, 25.6kA
Prospective Fault @ 74K
Figure 8-57
Cryostat System Performance
Figure 8-54
Current Limiting at 8660Vrms, 17.6kA
Prospective Fault @74K
Figure 8-57 shows the temperature profile of
the LN2 during the time period over which
the 74K short circuit test program was
conducted. The measurement was made in
three regions, at the top, center and bottom
of the region in the lower portion of the
pressure vessel. The faults that were applied
over this time frame are noted along the time
line. There were notable temperature
changes at the higher energy 8660Vrms,
8-38
20kA and 25kA tests. In all cases, the bath
temperature stabilized within 15 to 20
minutes, indicating the stable operation of
the cryostat.
accuracy and
capability
Verify operation
at sub-cooled
conditions
From this profile, it can be inferred that
there are strong thermal convection effects
in the bath to stabilize the temperature.
Objective
Achieved
Comment/Status
Verify current
limiting
performance
Yes
Very good
current limiting
achieved in all
tests
Verify current
sharing between
HTS elements
and trigger and
shunt coils
Yes
Measurements
taken on select
elements in low
voltage open
bath tests
Dynamic voltage
and current
limiting
resistance
development, R
= f(B, T,J)
Yes
Measurements
taken on select
elements in low
voltage open
bath tests
Speed of
superconductingto-normal
transition under
fault current
condition
Yes
Measurements
taken on select
elements in low
voltage open
bath tests
Effects of trigger
magnetic field
Partial
Ongoing,
additional testing
forthcoming
Electromagnetic
interference
between coils
Partial
Ongoing,
additional testing
forthcoming
HTS material
properties
characterization
Partial
Additional R&D
needed to
understand some
of the element
failures during
the tests
Obtain test data
to improve
simulation
Yes
Able to improve
analytical model
so calculated
Yes
Very good
cryostat
performance
Test Summary
Table 8-4 summarizes the accomplishments
made during these tests, during which the
MFCL proof-of-concept was successfully
demonstrated. As the table shows, the
majority of the key objectives of the test
program were met. Further work,
particularly in the area of the HTS material
development, is scheduled to further
understand why some of the HTS elements
failed during the tests.
Table 8-4
Summary of Key Test Objectives
Test
result matches
actual result
Recovery Performance
The proof-of-concept tests have shown that
the MFCL provides rapid current limiting
during the fault. During the fault, the HTS
material heats up, and must cool back down
to return to a superconducting state before
the device is invisible to the system again.
The use of the parallel inductor in the MFCL
concept helps to minimize the amount of
current that the HTS material must carry
during the fault. However, even though the
parallel inductor predominantly carries the
fault current, the current in the HTS results
in heating of the material.
The time to return to a superconducting state
is typically termed the “recovery” period.
The cooldown period is proportional to the
amount of time that the material was heating
up during the fault, both in terms of the
duration and the magnitude of fault current.
The length of the recovery period will also
be a function of the conditions under which
the recovery must take place. In other
applications, the MFCL device will not be
carrying any current during the recovery,
which will result in the fastest recovery
period. In some applications, the MFCL will
have to carry nominal load current during
8-39
the recovery, which will provide some
constant level of background heat generation
in the device that will lengthen the recovery
period. In this case, the HTS elements are
not bypassing the parallel inductors, so the
MFCL will present some impedance to the
system during this recovery. This may result
in system performance concerns due to the
impact of the MFCL impedance present
during the recovery period.
The concern about the reliability of HTS
elements has been addressed by the
development of 2nd generation hightemperature superconducting wires (2G HTS
conductors) by Superpower. The evaluation
results of the use of 2G HTS wires for the
FCL applications are presented in (Xie et al.
2006). The test set-ups that were used for
performing a series of low power and high
power tests are shown in Figure 8-58 and
Figure 8-59.
As part of the future work, formal studies
will be performed to determine acceptable
recovery periods for the potential
applications of the MFCL. Preliminary
discussions with utilities have indicated that
the presence of the MFCL impedance during
the recovery may not be an issue of concern,
since the device will typically be employed
in strong systems, where the short circuit
current level is high. The primary case study
is when the MFCL is connected in series
with a generator to limit its contribution to
the fault current. This case will be studied to
determine if the presence of the MFCL
impedance after the fault may cause
instability in the operation of the generator.
In addition to determining the application
recovery
requirements,
additional
development will be performed on
minimizing the recovery period.
Figure 8-58
Lab-scale Test setup (Xie et al. 2006)
Project Update – DOE Annual peer review,
2006
The latest status of the MFCL project has
been presented in (DOE Annual peer
review, 2006). Since middle of 2005, the
program had been placed on Reduced Effort
Status due to the following reasons.
•
Concerns about the reliability of the
melt cast BSCCO-2212 elements
•
Need for a partner with high voltage
expertise
•
Escalating program cost
Figure 8-59
Assembly of elements for high voltage test
(Xie et al. 2006)
It was concluded from the results of these
tests that 2G-HTS conductors provide
superior current limiting performance, faster
response and faster recovery times in
comparison to MCP-BSSCO elements that
8-40
were used in pre-alpha prototype. Table 8-5
provides the quantitative comparison of the
performance.
Controlled LC Resonance Circuits
In all of power electronic based FCL
systems described so far semiconductor
switches had to be turned off in order to
initiate the fault current limiting sequence.
This requires that the switches carry the
continuous load current which causes
continuous operating losses. Alternatively,
in LC resonance link circuits the power
electronic switches may be off during
normal operation and only turned on in
order to limit the fault current. In the FCL
series resonance link circuit shown in Figure
8-60 the limiting inductor LL and the
compensating capacitor CC are forming a
series resonance circuit at the power
frequency (e.g. 60 Hz). Therefore, the total
impedance of LL and CC is negligible during
normal operation. If a fault occurs, the
thyristor switch bypasses the capacitor and
therefore detunes the resonance circuit
which leaves the impedance of LL to limit
the fault current.
Table 8-5
High-power current limiting performance
Parameter
2G HTS
Conductor
MCPBSCCO
Prospective
Current (KA)
90
80
Limited Current
(KA)
32
40
Current through
element (kA)
3
25
Response time
(ms)
<1
4-5
Element quality
range
Narrow
Broad
The need for the partner with high voltage
expertise was fulfilled by SEI joining the
program. They would contribute their HV
expertise in the design, development and
manufacturing of bushing and cryostat
electrical distribution system. The program
has got further technical boost by the
addition of BOC in the team. BOC would
lead the design and development of
cryogenic system including instrumentation
and overall system monitoring.
Several systems based on this principle have
been built with rated voltages up to 145 kV
and rated currents up to 1.3 kA for
demonstration purposes (CIGRE 2003).
However, using the series resonance link
solely for the purpose of fault current
limitation is not economically attractive,
especially because of the large size and
weight of the passive components.
The program is poised to restart at full pace
but the estimated cost of the project has been
revised from $12.2 M to $23.6 M. Also, the
completion date of the project has been
moved to July, 2009. The detailed schedule
of the program may be found in (DOE
Annual peer review, 2006).
Other Technologies
In addition to the conventional and emerging
technologies
(Solid-state
based,
Superconductor based) discussed so far in
this chapter, there have been additional
efforts in the fault limiter applications.
Figure 8-60
Principle of the FCL Series Resonance
8-41
Liquid Metal FCL
In the past, liquid metal (LM) fault current
limiter concepts have made use of the socalled pinch-effect to break the circuit. The
pinch-effect is caused by high current
densities in a current constriction formed by
the liquid metal (typically a non-toxic liquid
metal alloy of Gallium, Indium, and Tin
with a melting temperature around 10
degrees Celsius). The high magnetic field in
the constriction causes it to constrict further
and eventually rapture and evaporate. The
subsequent arc is finally used to build up
voltage and limit the current.
Figure 8-61
Principle of the Liquid Metal FCL a)normal
operation, b) limiting
Series Compensator
A method to control a power convertorbased series compensator (SC) has been
proposed that would allow it to be used as a
FCL (Choi, et al. 2005). This additional
functionality of the fault current limiting for
downstream faults in SC is obtained through
the magnitude and phase control of its
injection voltage. Thus, SC can be modified
to serve the dual purpose of voltage
restoration for upstream faults and current
limiting for downstream faults. A schematic
diagram of such a dual-function SC is
shown in Figure 8-62.
Most recently, ABB Corporate Research
Center in Switzerland developed a new type
of FCL with liquid metal capillaries that do
not utilize the pinch effect and do not
develop an arc. When a fault occurs, the
liquid metal is magnetically driven into the
capillary which consists of walls made of
high resistive material as shown in Figure
8-61. After the fault is cleared (typically by
a up- or downstream circuit breaker) the
liquid metal returns quickly back between
the high conductive contacts, thus resets the
FCL automatically. The ABB researchers
demonstrated successful voltage build up of
almost 100 V per capillary at currents of up
to 2.7 kA (Schoft et al, 2005). A large
number of series and parallel-connected
capillary could potentially be used to build
FCL devices for medium and high voltage
applications. However, currently, not
enough information is publicly available to
assess the potential of this technology for
HV applications appropriately.
RR
Figure 8-62
Structure of Dual-Function SC (Choi, et al.
2005)
Power circuit in the figure is required for the
normal voltage restoring functionality of the
8-42
SC. The desired mitigation function is
achieved through the voltage injection by
the Power Convertor-based System (PCS)
shown in the figure. Operation Mode
Selector in the Control Circuit makes the
selection for one of the following modes of
operation: voltage restoration, fault current
limiting and stand-by mode. The decision is
based on the measurements of the line
current and PCC voltage. Under normal
conditions, SC is in stand-by mode. In the
case of a downstream fault, PCC voltage is
decreased and the line current is increased.
This causes SC to depart from stand-by
mode to fault current limiting mode. On the
other hand, upstream fault results in the
reduction of both the PCC voltage and the
line current resulting in SC to enter the
voltage restoration mode.
The FCL functionality of the SC has been
demonstrated by simulating the test system
described in Figure 8-63 and Table 8-6
using MATLAB. A single-phase fault was
simulated in the downstream and the PCC
voltage and line current have been plotted in
the absence (Figure 8-64) and presence
(Figure 8-65) of FCL functionality in the
SC.
Distribution supply
voltage
11 kV
Injection transformer
turns-ratio
1:1
DC link voltage
15.5 kV
Filter capacitor
48 uF
Filter inductor
0.2 mH
Figure 8-65
PCC voltage and line current- With FCL
function
Table 8-6
System Parameters in Simulation Model
Values
30 ohms
Figure 8-64
PCC voltage and line current- Without FCL
function
Figure 8-63
SC connected Distribution System (Choi, et
al. 2005)
Parameter
Equivalent source
reactance
This technology has been demonstrated only
by simulations so far and the author is
unaware of any practical developments.
Reactor and Compensating capacitor based
8-43
Approach
Another FCL approach consists of a
transformer-type reactor in parallel with a
compensating capacitor (Zhang, et al. 2005).
This approach is much simpler and
economical in comparison to the
superconducting and solid-state based FCLs.
The simplified diagram of such a device is
shown in Figure 8-66.
Figure 8-66
FCL based on Reactor and Compensation
capacitor (Zhang, et al. 2005)
Figure 8-67
Line Current Waveforms - a) Without FCL
and b) With FCL (Zhang, et al. 2005)
The desired regulation of the inductive
reactance of the reactor for the purpose of
fault current limiting is achieved by a
discharge gap (DG) connected to its
secondary
windings.
Under
normal
conditions, DG does not breakdown causing
the secondary circuit of the reactor to be
open-circuited. Under such conditions, the
inductive reactance of the reactor is given by
the magnetizing reactance which is much
higher than the capacitive reactance. As a
result, reactor has almost no effect on the
current flow. During a fault, DG
breakdowns and the secondary of the reactor
is short-circuited. The resultant impedance
of the FCL serves to limit the fault current.
The current limiting effect has been
demonstrated by transient analysis using
EMTP (Figure 8-67).
This technology too has been demonstrated
only by simulations so far and the author is
unaware of any practical developments.
For the benefit of readers, a test circuit has
been modeled on EMTP platform that can
be used to demonstrate the response of
various fault current limiting technologies
like series reactor and superconducting FCL.
This application description has been
provided as an appendix to the chapter
Comparison of FCL technologies
The comparison of several of the emerging
FCL technologies that have been covered in
this chapter has been handled in EPRI
2005a. The findings of the survey are
presented here.
Losses
8-44
for such a duty with the penalties being
potentially increased size (HTS material)
and cooling power.
YBCO thin film SCFLs were found to have
less losses in comparison to BSCCO 2212
bulk material based SCFLs. Solid-state
systems currently show higher losses in
comparison to SCFCLs. With expected
advances in solid-state materials such as
silicon carbide (SiC) losses will decrease
somewhat.
Solid-state systems can be built easily with
immediate recovery capability. Some
additional cooling may be required for the
semiconductor devices to allow for another
fault limitation duty very shortly after the
first fault has been limited. If resistors or
MOV’s are used to absorb the bulk of the
limitation energy those devices must be
appropriately overrated in order to cope with
multiple, subsequent fault events, a measure
solid-state
systems
that
use
a
superconducting DC bias coil may not
require.
Losses of FCLs based on mechanical contact
systems, of course, show negligible losses.
Finally, losses of FCL systems based on DC
biased (superconducting) coils will be
similar to those of other solid-state systems.
Size
Although, information given about size of
prototype and/or test equipment is difficult
to use for comparisons it appears that
superconducting devices based on BSCCO
2212 bulk material yield a similar size per
MVA as solid-state devices. YBCO thin
film SCFCLs seem to be somewhat more
compact. However, FCLs based on
mechanical contact systems clearly are
smaller in size. This is due to the negligible
losses in mechanical contact systems which
require substantially less cooling than solidstate breakers.
FCL characteristic in the network
Any of the novel technologies present only a
negligible, (dominantly resistive) impedance
in series to the line during normal operation.
In case of a FCL actuation, SCFCLs will
result in undistorted fault currents (follow
current) except for the first half cycle where
the current will be somewhat distorted from
the sine waveform. After limitation the
SCFCL appears as a linear impedance in the
network (the quench itself is highly non
linear). Whether this impedance is
predominantly resistive or inductive depends
primarily on the shunt impedance
characteristic and therefore is a free design
parameter.
Since FCL systems based on DC biased
(superconducting) coils will also utilize
power electronic devices their size is likely
to be similar if not larger than any of the
other devices.
Solid-state FCLs however can only rely
upon phase angle control for the follow
current which will cause substantial current
distortion. This may have an impact on the
protection relay coordination.
Recovery
A
fundamental
problem
with
superconducting FCLs is the recovery back
to the low impedance state after a fault in
order to continue service immediately after
the fault has been cleared. While immediate
recovery has not been demonstrated in any
of the major demonstration projects so far it
is, in principle, possible to design an SCFCL
All the traditional means of reducing fault
current levels essentially introduce an
additional reactance into the system
permanently. Exceptions are fuse-based
devices which, if not shunted by a current
limiting reactor, appear as an open circuit
8-45
after trigger (no follow current) and
sequential tripping which does not affect the
system impedance at all.
Inductance (µH)
Case Studies
Air-core CLR in Brazil
The successful experience of using current
limiting reactors for limiting fault currents in
Brazil is documented in Amon et al, 2005.
FURNAS Experience
The traditional practice in Brazilian utilities
including FURNAS is the usage of CLR at
tertiary windings of autotransformers that
supply the auxiliary services at the
transmission
substations.
The
basic
characteristics of 15 kV CLRs installed at
FURNAS substations are shown in
Table 8-7. In December of 1998, a CLR was
installed at 362kV level by FURNAS. The
characteristic of the CLR are shown in Table
8-8.
Rated current (A)
2100
2600
Maximum voltage
drop (kV, RMS)
18.9
52
Rated power (MVAr)
40
135
Rated short circuit
current (kA, RMS)
25
10
Impulse insulation
level (kV, peak)
1300
1550
Switching insulation
level (kV, peak)
850
1180
Quality factor
300
400
Type of insulation
external
external
ELETRONORTE’s Experience
Eletronorte is a North Brazilian utility which
has been using a CLR at HV substation in
Tucurui power plant since July of 2004. The
CLR is installed as coupling device between
two switching substations as shown in
Figure 8-68. The characteristics of the CLR
are shown in Table 8-8.
Table 8-7
Characteristics of 15 kV CLRs installed at
FURNAS Substation
Figure 8-68
Electronorte CLR Configuration
Table 8-8
Characteristics of Funrnas and Electronorte
CLR
Furnas
CLR
Electronort
e CLR
Rated voltage (kV)
345
550/sqrt(3)
Rated frequency
(Hz)
60
60
Per phase
24000
53050
IEE Project in China
In December of 2005, China’s Institute of
Electrical Engineering (IEE) announced that
it has successfully demonstrated a
superconducting fault current limiter
application for the first time in a power grid
in China. The device was fabricated in
collaboration with the Technical Institute of
Physics and Chemistry and Hunan Electric
8-46
Power Company utilizing High Temperature
Superconductor (HTS) wire manufactured
by American Superconductor Corporation.
The device was installed in an electric
substation near Changsha, the capital city of
Hunan province. It has a voltage rating of
10.5 kilovolts and its normal operating
current is 400 Amperes. Since it was put
into operation in August 2005, it has
instantaneously reduced three-phase, short
circuit currents in the range of 3,500
Amperes down to 635 Amperes.
Table 8-9
Number of Buses with Excessive Fault
Currents
Year
2005
2010
2017
Number
of buses
8
29
29
The current practice is to upgrade the circuit
breakers but KEPCO is investigating the
following measures to limit the fault
currents.
• Bus-bar split
• Current limiting reactor
• SCFCL
The equivalent circuit of the 345kV grid
around the metropolitan area is shown in
Figure 8-69. In the figure, substation A is
the location where the fault current limiting
measures
have
been
investigated.
Simulations were performed using TSAT
and VSAT by Powertech. Four kinds of
faults were simulated (two on bus B3 and
two in remote areas from B3).
SCE Distribution Circuit of future
SCE has an advanced protection project
which has following objectives.
• Improve the detection and isolation
of circuit faults on the distribution
system to minimize customer
interruptions in both frequency and
duration.
• Design and test new protection
methods with and without a fault
current limiter.
• In addition, new fault sensing and
prediction techniques will be studied
and tested on the SCE Distribution
Circuit of the future
The outside parties in the project include
DOE, ORNL, EPRI, Intelligrid, CEC and
KEMA consulting. The test circuit is a 12
kV Avanti circuit located in San Bernadino,
California. The circuit would have about 23
kA fault duty and serve nearly 2000
customers. They plan to use EPRI solid state
or superconducting FCL.
Figure 8-69
Equivalent Circuit of KEPCO’s 345 kV grid
Korean Electric Power Grid
Fault current limiting measures in KEPCO’s
345 kV grid are presented in (Lee et al.
2006). The fault current levels have been
increasing steadily and the number of
substations where the fault current levels are
expected to exceed the installed interrupting
rating of 40 kA is shown in Table 8-9.
The transient stability evaluation index ( )
was computed for these faults using the
following expression:
8-47
η=
Adec − Ainc
Adec
practices. Despite its poor performance in
comparison to SFCL, CLR is being
considered as a preferred solution due to the
commercial unavailability of SFCL at 345
kV level.
Where,
Adec
=Decreasing kinetic energy
Ainc
=Increasing kinetic energy
Summary and Recommendations
The system is considered stable if the value
of is found to be greater than zero. The
summary of the findings are shown in Table
8-10.
At several locations in power system,
employing some kind of fault current
limiting measures is necessary to avoid
costly system upgrades. There are additional
benefits to system if fault current levels can
be reduced. The conventional methods that
are currently in practice are effective to an
extent but have their limitations and
drawbacks.
Table 8-10
Transient Stability Evaluation Index (unit :%)
Fault Type
FCL
CLR
Bus-split
Remote Fault I
52.02
-4.86
-4.78
Remote Fault
II
-98.37
-80.11
-80
B3 bus Fault I
60.92
61.01
58.67
B3 bus Fault II
-98.51
-86.03
-85.61
The fault current limiters that are based on
novel technologies such as solid-state and
superconducting materials are highly
effective and efficient in theory. But, these
FCLs are still in various stages of
development and not grid-ready yet. Once
ready for use, the next-generation FCLs are
expected to find widespread applications in
the transmission and distribution systems all
over the world.
Critical fault clearance time was also
computed for these faults and the results are
summarized in Table 8-11. The value of less
than 0.1 s for critical clearance time means
that the system is unstable.
Table 8-11
Critical fault Clearance Time (unit :sec)
Fault Type
FCL
CLR
Bus-split
Remote Fault I
0.422
0.043
0.043
Remote Fault
II
0.042
0.008
0.008
B3 bus Fault I
0.350
0.336
0.203
B3 bus Fault II
0.051
0.051
0.034
It is evident that FCL is the best option as
far as the transient stability and critical fault
clearance time is concerned. It was also
found that FCL provides greater power
transfer limit as compared to other practices.
Use of CLR (29 ohms) is found to be
medium favorable solution among the three
8-48
APPENDIX 8.1 EMTP Model for FCL Evaluation
A test circuit explained in Appendix 4.2 has been modified to demonstrate the concept of fault
current limiters in transmission systems (See Figure 8-70). EMTP is chosen as it is one of the
most widely used programs for performing transient studies. The circuit represents a 2-bus
transmission system and the system details are already explained in Appendix 4.2. The FCL
module is an addition to the circuit. Double-clicking on the FCL module causes a scriptbox
(Figure 8-71) to appear on the screen that may be used to select the various parameters.
Selection = 1 : No FCL
Selection = 2 : CLR in Service
Selection = 3 : Superconducting FCL
Line
a
a
b
b
kv = 230
c
c
FCL
Breaker
Selection =3
+
+
RL2
+ AC
kv = 230
2
Fault
1
+
+.2
8
R_RELAYB
+
R_LEADN
R_LeadB
+.2
R_LEADC
.2
+
1
1
+
CT_B
CT_C
2
2
CT_A
+
R_leadA
+.2
+
8
+
R_RELAYA
SRC2
RL1
R7
+
1e-6
AC +
SW2
+
-1| .183| 0
SW1
?i
+
.1|1E15|0
SRC1
8
R_RELAYC
Figure 8-70
FCL Evaluation Circuit
Parameter Description of FCL Module
•
Select : User has three options for this parameter and are explained below.
o Select = 1 ; FCL action is disabled
o Select = 2; Current Limiting Reactor (CLR) is used for fault limiting action
o Select = 3; Superconducting based FCL is enabled
•
KV_DROP: Maximum voltage drop in the CLR
•
MVAR: MVAr rating of the CLR
•
Rsc: Parallel resistance in a superconducting FCL
8-49
Figure 8-71
Script Box for FCL module
FCL Module disabled
For this selection, the simulation of the test circuit yields the following plots for the currents in
CT primaries and secondaries (Figure 8-72). It is seen from the plots that without FCL the
maximum value of the first peak after the fault is nearly 12 kA and currents in CT secondaries
are distorted due to saturation.
8-50
Figure 8-72
CT Currents –Without FCL
Current Limiting reactor in Service
Figure 8-73
CT Currents –With CLR
8-51
CT currents corresponding to the use of current limiting reactor (50 MVAR) are shown in Figure
8-73. It is found that the first peak after fault gets reduced from 12.5 kA to about 9.7 kA. That
corresponds to more than 20% drop in the maximum fault current peak. But, it is apparent from
CT secondary current waveforms that current limiting reactor in this case was unable to prevent
the saturation in the CTs.
Superconducting FCL in Service
CT currents corresponding to the use of superconducting based FCL (50 ohms parallel
resistance) are shown in Figure 8-74. It is seen that the first peak after fault gets reduced to about
4kA in this case. Also, the reduction in fault current is sufficient to prevent the saturation in CTs
as evident from the sinusoidal waveforms in CT secondary currents. It may be noted that very
simplified model has been used for Superconducting FCL as the intent is to just demonstrate the
principle of its operation.
Figure 8-74
CT Currents –With Superconducting FCL
8-52
Processed
BSCCO-2212
Superconductor under Self-Field,"
IEEE Transactions on Applied
Superconductivity, Vol. 13, No. 2,
June 2003, pp. 2028-2031.
References
1. Amon F. J., P. C. Fernandez, E. H.
Rose, A. D´Ajuz, A. Castanheira,
2005.
“Brazilian
Successful
Experience in the Usage of Current
Limiting Reactors for Short-Circuit
Limitation,”
Presented
at
the
International Conference on Power
Systems Transients (IPST’05) in
Montreal, Canada on June 19-23.
7. Chen, M.; Baumann, T.; Unternahrer,
P.; Paul, W., 1997, “Fabrication and
Characterisation of Superconducting
Rings for Fault Current Limiter
Application”, Physica C, Vol.: 282 287, pp. 2639 – 2640.
8. Chen, M.; Paul, W.; Lakner, M.;
Donzel, L.; Hoidis, M.; Unternaehrer,
P.; Weder, R.; Mendik, M., 2002.
“6.4 MVA Resistive Fault Current
Limiter Based on Bi – 2212
Superconductor”, Physica C, Vol.:
372 - 376, pp. 1657 – 1633.
2. Bock, J., F. Breuer, H. Walter, M.
Noe, et al, 2004, “Development and
Successful Testing of a MCP
BSCCO-2212 Components for a
10MVA Resistive Superconducting
Fault
Current
Limiter”,
Superconductor
Science
and
Technology No. 17, February 2004.
9. Choi,
S.
S.,
Wang,
T.X.,
Vilathgamuwa, D.M. “ A Series
Compensator with Fault Current
Limiting
Function”,
IEEE
Transactions on Power Delivery, Vol
20, No. 3, July 2005, pp 2248-2256.
3. Bock, J.; Breuer, F.; Walter, H.;
Elschner, S.; Kleimaier, M.; Kreutz,
R.; Noe, M., 2005. “CURL 10:
Development and Field-Test of a 10
kV/10 MVA Resistive Current
Limiter Based on Bulk MCP-BSCCO
2212”, IEEE Transactions on Applied
Superconductivity, Volume 15, Issue
2, June 2005 pp.1955 – 1960.
10. Das, J.C. 1997. “Limitations of FaultCurrent Limiters for Expansion of
Electrical Distribution Systems,”
IEEE Transactions on Industry
Applications, Vol. 33, No. 4,
July/August 1997, pp. 1073-1081.
4. Boenig, H. J.; Mielke, C. H.; Burley,
B. L.; Chen, H.; Waynert, J. A.;
Willis, J. O., 2002. “The Bridge-Type
Fault Current Controller - a new
FACTS Controller”, Proceedings of
the IEEE Power Engineering Society
Summer Meeting 2002, Vol.: 1, pp.
455 – 460.
11. DOE Annual Peer Review. 2006.
Presentation on Superconducting
Fault Current Limiter Project in
“Superconductivity
for
Electric
Systems”, Washington DC, July 2527, 2006.
12. Duggan, P.M. “Integration Issues for
Fault Current Limiters and other new
Technologies- a Utility Perspective”
IEEE Power Engineering Society
General Meeting. 2006
5. Boenig, H.; Paice, D., 1983. “Fault
current
limiter
using
a
superconducting
coil”,
IEEE
Transactions on Magnetics, Volume
19, Issue 3, pp.1051 – 1053.
13. CIGRE WG A3.10, 2003, “Fault
Current Limiters in Electrical
6. Cha, Y.S. 2003, "An empirical
Correlation for E(J,T) of a Melt-Cast
8-53
conductors in resistive fault current
limiters,” in: IEEE Transactions on
Applied Superconductivity, Volume:
13, Issue: 2, June 2003, pp. 2044 –
2047.
Medium and High Voltage Systems”,
CIGRE Technical Brochure 239,
December 2003.
14. EPRI. 1996. Development and
Testing of a 38 kV Current Limiting
Protector, EPRI, Palo Alto, CA, TR106992.
22. Kunde, K.; Kleimaier, M.; Klingbeil,
L., “Integration of Fast Acting
Electronic Fault Current Limiters
(EFCL) in Medium - Voltage
Systems”, 17th International
Conference on Electricity Distribution
– CIRED, Barcelona, May 2003
(SIEMENS)
15. EPRI. 2002. Technical and Economic
Evaluation of a Solid State Current
Limiter, EPRI, Palo Alto, CA,
Consolidated Edison Co. of New
York, New York, NY, Allegheny
Power, Greensburgh, PA, and ISO
New England, Holyoke, MA: 2002.
1001816.
23. Langston, J., Steurer, M., Woodruff,
S., Baldwin, T., Tang, J. “A Generic
Real-time Computer Simulation
Model for Superconducting Fault
Current Limiters and its Application
in System Protection Studies”, IEEE
Transactions on Applied
Superconductivity, Vol. 15, No.2,
June 2005, pp. 2090 – 2093.
16. EPRI. 2004a. High Temperature
Superconducting
Matrix
Fault
Current Limiter: Proof-of-Concept
Test Results, EPRI, Palo Alto, CA,
SuperPower, Inc., Schenectady, NY,
and Nexans SuperConductors, Hürth,
Germany: 2004. 1008697.
24. Lee, Soo-Hoan, Lee, Kang-Wan,
Yoon, Yong-Beum, and Hyun, OkBae. “FCL Application Issues in
Korean Electric Power Grid”, Power
Engineering Society General
Meeting, IEEE. June 2006.
17. EPRI 2004b. Progress Report on
Medium Voltage Solid-State Current
Limiter, EPRI, Palo Alto, CA, and
Consolidated Edison, 4 Irving Place,
New York, NY 10004. 1002117.
18. EPRI. 2005a. Survey of Fault Current
Limiter (FCL) Technologies. EPRI,
Palo Alto, CA: 2005. 1010760.
25. Leung, E. et al, 2000, “Design and
Development of a 15kV, 20kA HTS
Fault Current Limiter”, IEEE
Transactions
on
Applied
Superconductivity, Vol. 10, No. 1,
March 2000.
19. EPRI. 2005b. Medium Voltage SolidState Current Limiter- Progress
Report. EPRI, Palo Alto, CA: 2005.
1010610.
26. Leung, E. 2000, “Superconducting
Fault Current Limiters”, IEEE Power
Engineering Review, August, 2000.
20. Fukagawa, H.; Matsumura, T.;
Ohkuma, T.; Sugimoto; S.; Genji, T.;
Uezono, H.: “Current State and
Future Plans of Fault Current
Limiting Technology in Japan”,
CIGRE 2000 Session, Report 13-208,
Paris, 2000
27. Neumueller, H. W. 2004. “Fault
Current Limiter in Thin Film
Technology”, presented at the
SCENET
Workshop
on
Superconducting
Fault
Current
Limiters (FCL) in Siegen, Germany,
June 29 2004.
21. Kraemer, H.-P. et al. 2003.“Switching
behavior of YBCO thin film
8-54
28. Paul, W.; Chen, M.; Lakner, M.;
Rhyner, J.; Braun, D.; Lanz, W.;
Kleimaier,
M.,
2000.
“Superconducting
Fault
Current
Limiter Application, Technical and
Economical Benefits, Simulations and
Test Results”, Cigre Session, Paris
2000, pp. 13 – 201.
29. Schmitt, A, and CIGRE Working
Group A3.16, “Fault Current Limiters
Report on the activities of CIGRE
WG A3.16”, Power Engineering
Society General Meeting, IEEE,
2006.
Superconductivity, Vol. 13, No. 2,
June 2003.
35. Yazawa, T.; Ootani, Y.; Sakai, M.;
Kuriyama, T.; Nomura, S.; Ohkuma,
T.; Hobara, N.; Takahashi, Y.; Inoue,
K., 2004. “Development of a 66 kV /
750 A High - Tc Superconducting
Fault Current Limiter Magnet”, IEEE
Transactions
on
Applied
Superconductivity, Vol.: 14, No.: 2,
pp. 786 – 790.
36. Xie, Y.Y., Teklatsadik, D., Hazelton,
D.,
and
Selvamanickam.
“2nd
Generation
High-Temperature
Superconducting Wires for fault
Current
Limiter
Applications”,
Applied
Superconductivity
Conference, Seattle, WA, 2006.
30. Steurer, M., Fröhlich, K., Holaus, W.,
and Kaltenegger, K. “A Novel Hybrid
Current-Limiting Circuit Breaker for
Medium Voltage: Principle and Test
Results”, IEEE Trans. PD, Vol 18,
No. 2, April 2003, pp. 460-467
37. Zhang, X., LI, M. “Using the Fault
Current Limiter with Transformertype Reactors to Reduce Short-Circuit
Currents”, IEEE/PES Transmission
and Distribution Conference &
Exhibition: Asia and Pacific Dalian,
China, 2005.
31. Steurer, M., M. Noe, F. Breuer, 2004,
“Fault Current Limiters . R & D
Status of Two Selected Projects and
Emerging Utility Integration Issues”,
IEEE PES General Meeting, June
2004.
32. Strumpler, R.; Skindhoj, J.; GlatzReichenbach, J.; Kuhlefelt, J.H.W.;
Perdoncin, F., “Novel medium
voltage fault current limiter based on
polymer PTC resistors”, IEEE
Transactions on Power Delivery,
Volume 14, Issue 2, April 1999
Page(s):425 - 430
33. Tekletsadik, K., A. T. Rowley, et al,
1999, “Development of a 7.5MVA
Superconducting
Fault
Current
Limiter”, IEEE Transactions on
Applied Superconductivity, Vol. 9,
No. 2, June 1999.
34. Waynert, J.A., H.J. Boenig, et al,
2003, “Restoration and Testing of an
HTS Fault Current Controller” IEEE
Transactions
on
Applied
8-55
Increased power flow – This goal of utilities
is also achievable by reduction in system
impedance that can be made possible by
the use of FCLs.
9
TECHNICAL AND
ECONOMIC ANALYSIS
OF FAULT CURRENT
MANAGEMENT
SOLUTIONS
Stability— Use of fault current limiters
allows higher short-circuit capacity
during normal conditions. It results in
higher steady as well as transient
stability. System becomes stronger and
experiences lesser perturbations.
Reduced strain— Use of fault current
limiters results in lower short-circuit
capacity during fault conditions. It
results in reduced thermal and
mechanical strain on the system.
Benefits of Fault Current Limiting
The following are some of the benefits of
fault current limiters in transmission and
distribution systems:
Failures—Cable thermal failures are less
likely, and violent equipment failures are
less likely.
New Capacity - Fault current limiters could
be applied to new capacity additions
and/or “surgically” at strategic locations,
such as substation bus ties, to effectively
mitigate the fault current from multiple
generation sources. This would provide a
flexible tool that could be used to
accommodate new capacity from
generation or transmission, distributed or
aggregate generation energy storage.
Use of fault current limiters in coupling
new generation (conventional and
distributed)
provides
additional
flexibility in choosing their location.
Conductor burndowns—At the fault, the
heat from the fault current arc burns the
conductor enough to break it, dropping it
to the ground. Faster clearing and lower
magnitudes reduces the chance of
burndowns.
Damage of inline equipment—The most
common problem has been with inline
hot-line clamps. If the connection is not
good, high-current fault arcs across the
contacts can burn the connection apart.
Faster clearing and lower magnitudes
reduces the chances of such damage.
Reduced losses – The desire of utilities to
reduce the system losses could be
fulfilled if the system impedance could
be reduced without increasing the short
circuit currents (Sjöström and Politano.
2001). This is possible by use of FCL
devices in strategic locations in the
system such as generation and
transformer feeders.
Evolving faults—Ground faults are more
likely to become two- or three-phase
faults with longer, higher-magnitude
faults; current-limiting will reduce this
probability.
Underbuilt—Faults
on
underbuilt
distribution are less likely to cause faults
on the transmission circuit above due to
rising arc gases with fault-current
limiting.
Reduced Rating – In case of newly planned
networks, use of FCL can allow the
selection of lower ratings for the
equipment resulting in potential savings.
Equipment ratings—Some substations have
fault current levels near the maximum
ratings of existing switchgear; additional
9-1
short-circuit
current
requires
reconfigurations or new technology.
Fault-current limiting can solve this.
of multi-shot re-closure on generator
buses, particularly as distributed
generators are deployed at various
voltage levels and locations across utility
grids.
Shocks—Step and touch potentials are less
severe during faults.
Conductor movement—Conductors move
less during faults (this provides more
safety for workers in the vicinity of the
line and makes conductor slapping faults
less likely).
Current limiting reduces the energy at the
location of the fault. This provides safety to
workers and the public. Arc damage to life
and property occurs in several ways:
Pressure wave—The fault arc pressure wave
damages equipment and personnel.
Voltage sags—Current limiting reduces the
depth of the voltage sag to customers on
adjacent circuits.
Heat—The fault arc heat burns personnel
and can start fires.
Coordination—Fuse coordination is easier.
Fuse saving is more likely to work with
lower fault currents.
Pressure buildup in equipment—An arc in
oil causes pressure buildup that can
rupture equipment.
Superconducting Cables - A fault current
limiter added in series with a
superconducting cable can improve the
cable's performance and enable design of
smaller cable sizes, as well as eliminate
loss of superconducting cable operation
during
cryogenic
recovery
time
following an external fault. The
insulation system of a superconducting
cable is likely to have limited strength
because of the need for minimal
mechanical cross section bracing
spanning the vacuum segment. It may
not be capable of handling the magnetic
forces occurring in the worst case of a
high current fault, particularly in
transmission applications, which will
draw fault currents from higher
impedance parallel paths. It is believed
that fault current limiter will be an
essential, enabling adjunct to the regular
use of superconducting cables.
All of these effects are related to the arc
energy and all are greatly reduced with
current limiting.
Economic Analysis for Conventional
Solutions
Figure 9-1 illustrates in a nutshell several of
the practical solution options for the case in
which added fault current contributions
cause excessive duty on fault interrupting
equipment. Not all of these solutions would
have to be implemented for any particular
case. This figure illustrates the various
options and rounded-off estimates of the
associated costs. The source data for these
costs is derived from Table 9-3 and Table
9-4.
Inrush Current - The Solid State Current
Limiter has a unique capability to limit
inrush current, even for capacitive loads,
by gradually phasing in the switching
device. This may be of particular benefit
in mitigating stress on generator shafts,
while preserving the reliability benefits
9-2
Replace
Transformer
$500K+
Add Line
Reactors to
Limit Fault
Current
$80K - 100K
per feeder
must be able to achieve the necessary BIL,
which is sometimes a challenge.
An alternative would be to install the line
reactors after the breakers, taking the risk
that a rare fault in the bus or a reactor will
not yield maximum available fault current.
This could reduce the cost of line reactors in
some cases.
Replace
Breaker
(If Possible)
$50K each
Replace
Cutouts
$200 each
Reducing the fault current contribution from
the utility source will protect both feeder
circuit breakers and any fuses close to the
substation. An alternative is to replace both
with devices that can handle the increased
fault current. This assumes that there are
higher-rated breakers available. A typical
breaker change-out cost is estimated to
average approximately $50K with a range of
$25K - $80K, depending on the style of
breaker and rating. For reliability, buses are
often switched to alternate sources during
outages and for maintenance purposes.
Therefore, all breakers on any bus to which
new generation might be connected would
have to be upgraded. The substation will
typically have at least two three-phase
breakers per bus with an average of perhaps
four in the USA. In high load density areas,
particularly those served by LV networks,
there might be dozens of breakers that
would have to be changed.
Add Voltage
Regulator to
Compensate
for Weaker
Source
$70K
Figure 9-1
Some Options for Solutions to Excessive
Fault Currents
The solutions depicted either reduce the
fault current contribution from the utility
source or deal with the consequences of
other solutions.
The key element controlling the utility fault
current contribution is the utility distribution
substation transformer. Therefore, one
obvious solution would be to replace the
existing one with one of higher impedance.
This might be practical for an aging
substation that is due for replacement after
40-50 years of service. However, the cost is
quite high to consider changing a younger
transformer simply to allow more short
circuit current margin for infeed. The
replacement costs would typically total at
least $500K and can easily exceed $1M per
transformer in some cases.
Fuses are relatively inexpensive at
approximately $200 per site. If currentlimiting fuses are chosen, the cost could be
higher. The number of fuses that have to be
changed will vary considerably from one
utility to another. The area of risk is
typically 0.5 – 1.0 mile from the substation.
Some utilities will have many fuses in this
zone while others may typically have none.
One concern of distribution engineers is that
line operating personnel will have to
exercise more caution when replacing blown
fuses. If they are accustomed to carrying
only one kind of fuse in the truck, there is a
risk that they will replace a failed fuse with
the wrong kind, creating a safety hazard
While this seems like a high cost, other
options are also expensive. Line reactors
may be installed ahead of the utility breakers
to limit the current seen by the breaker
during faults. While this is a simple concept,
there are many ancillary considerations.
There frequently is insufficient space for
these reactors in the substation as built.
Therefore, there must be some rearranging
of the buswork and, perhaps, moving of the
existing breakers. The reactor installation
9-3
should the fuse explode during the re-fusing
operation when the fault is still present.
Table 9-1 and
Table 9-2 show total costs of various options
for modifying two relatively simple
distribution system design:
There can be consequences to measures to
reduce the fault current contribution from
the utility source. A significant one is that
by making the source weaker, it becomes
difficult to maintain adequate voltage at the
ends of the feeder. This will require the
addition of line voltage regulators, which
could add as much as $70K in cost per
installation. This cost is somewhat
dependent on local utility practice and the
circuit characteristics. The lower range on
the cost for a regulator installation is
approximately $25K for a smaller regulator.
However, if the feeder is designed to be
picked up from the other end in an
emergency, a large regulator is required.
1. A
single-transformer,
substation,
2. A
two-transformer,
substation.
4-feeder
6
feeder
For options that weaken the source, it is
assumed that half the feeders will require the
installation of voltage regulator banks to
maintain adequate voltage at the ends of the
feeders.
These two examples use the lower range of
cost estimates for substation transformer
replacement, but that option is still the most
expensive.
In each of these cases, the option of
changing out the circuit breakers at $50K
each is the least costly. Of course, this
assumes that it is possible to obtain the
increased fault interrupting ratings. The
margin between this option and the line
reactor option is such that even if the cost of
breaker change out is twice this estimated
cost, this option will still be the least costly.
Cost estimates typically ranged from $25K
to $80K. One special type of breaker was
$200K to replace. Thus, there can be quite a
variation depending on the type of breaker
employed.
When ageing stations are upgraded, the cost
of creating more margin for increased fault
current contribution may be absorbed in the
renovation cost with little incremental
impact, if any. For example, newer breakers
may have higher interrupting ratings.
There is another issue that may preclude
simply upgrading the feeder breakers in the
substation. If the feeders have any customers
with primary-side switchgear (e.g., those
taking service at primary voltage), the
interrupting ratings of their breakers and the
9-4
fault duty must also be considered. Some
utilities with numerous large three-phase
customers have indicated that they have
made commitments that their fault current
levels would never exceed a certain amount.
Whether that exceeds breaker ratings or not,
new generation can take the fault current
magnitudes over those promised limits.
Then the only remaining alternatives are
those that reduce the contribution from the
utility source.
Table 9-1
Costs of Fault Current Limiting Options for
Single-Transformer, 4-Feeder Substation
Option
When a greater margin cannot be
economically obtained, the main recourse is
to restrict the new generation to types that
do not contribute significantly to fault
currents on the primary feeder.
Item
No.
Cost,
ea, Total
($000) ($000)
Replace Substation Transformer
Transformer
1
$500
Voltage Reg.
2
$70
Total
$500
$140
$640
Install Line Reactors
Reactor Sets
Voltage Reg.
Total
4
2
$100
$70
$400
$140
$540
Replace Circuit Breakers
Feeder Breakers
Total
4
$50
$200
$200
Table 9-2
Costs of Fault Current Limiting Options for
Two-Transformer, 6-Feeder Substation
Option
Item
No.
Cost,
ea,
Total
($000) ($000)
Replace Substation Transformer
Transformer
2
Voltage Reg.
3
Total
$500
$70
$1,000
$210
$1,210
Install Line Reactors
Reactor Sets
Voltage Reg.
Total
6
3
$100
$70
$600
$210
$810
Replace Circuit Breakers
Feeder Breakers
Total
6
$50
$300
$300
Cost Data
Table 9-3 lists a range of cost estimates
taken from a Distributed Generation (DG)
cost survey of participating utility members
(EPRI. 2005), which focus on mitigating the
9-5
effects of increased fault current duty. In
addition, average unit cost data gleaned
from public documents supplied for general
rate cases in California and Connecticut are
included for some items as a reference to
confirm the cost values. The rate case unit
cost data may tend to be loaded more
heavily than job cost estimates depending on
accounting procedures. Also, there is
considerable variation in the types of
equipment used in different locales.
However, there was substantial overlap in
the data, giving credibility to the cost
estimates.
Item
Table 9-3
Cost Estimates Provided by Utilities for
Common Changes Required to Support
Higher Penetrations of DG
Min.
Cost
Estimate
Max.
Cost
Estimate
Unit
Cost
Range
from
Rate
Cases
Replace
Substation
Transformer
$500K
$1M
$800K
$1.6M
Line reactors,
each feeder
$80K per
feeder
$100K
per
feeder
Line recloser
replacement
$40K
$60K
Relay change
(Engineering)
$1500
$5000
Relay change
(Technician)
Add PTs in
Substation for
$30K $70K
Add PTs at
line recloser
location to
implement for
directional
overcurrent or
reclose block.
$18K
$70K
Replace
simple
overcurrent
relay with
directional
overcurrent
$8100
$10K
Add direct
transfer trip
between
substation and
DG site
$40K $55K
$100K $200K
Item
Min.
Cost
Estimate
Max.
Cost
Estimate
Change
voltage
regulator
control for
handling
reverse power
from DG
$5000
(mat’l
only)
$21.5K
(incl. 90
m-hr
labor)
Install recloser
at large DG site
as
interconnection
breaker
$5000
$13.5K
Max.
Cost
Estimate
Unit
Cost
Range
from
Rate
Cases
directional
overcurrent
relaying or
reclose block
The cost for an Is-Limiter replacement is
estimated at $2000-3000 per phase, after
each operation. In addition, switches to
prevent single phasing, and their additional
downtime must be added into the cost. (Das.
1997.)
Item
Min.
Cost
Estimate
Change
LTC/VR
$18K
9-6
$25K
$2500
$5000
Unit
Cost
Range
from
Rate
Cases
$30K $70K
Item
Min.
Cost
Estimate
Max.
Cost
Estimate
Item
Unit
Cost
Range
from
Rate
Cases
control setting
Hire outside
consultants to
analyze DG
interconnection
$2500
Replace
substation
breaker
Change fused
cutouts
Add new fused
cutouts on
laterals where
none exist
$300
(mat’l
only)
Change fuses
downline from
recloser
Assigning
technician to
supervise
installation
$20K
$50K
$200 ea
$300
$2400
(incl.
19.5 m-hr
labor)
$2400
$3200
Unit
Cost
Range
from
Rate
Cases
$6000
$5000
$7850
Table 9-4
Typical Labor Rates Used to Compute Costs
Job Classification
Hourly Rate
Engineer
$50 - $60
Technician
$40
$25K
FCL Applications
$25K
$70K
$125K
Capacitor
bank
replacement
$12K
$20K
$20K $80K
Replace a line
voltage
regulator with
3-250kVA
regulators
Replace
cutouts with 3phase recloser
or
sectionalizer
Max.
Cost
Estimate
Set or replace
pole
$25K $80K
(one
type:
$200K
)
Add a line
voltage
regulator
Replace
cutouts with 3phase switch
Min.
Cost
Estimate
When power delivery networks are
upgraded or new generation is added, fault
levels can increase beyond the capabilities
of the existing equipment, with circuit
breakers in an “overduty condition”.
Utilities are faced with extended outages and
expense to upgrade all the affected breakers.
An alternative approach is to use a FCL to
reduce the available fault current to a lower,
safer level so the existing switchgear can
still protect the grid. FCL operation is very
fast, preventing damage before the first peak
of the fault current, Figure 9-2.
$70.5K
(incl. 180
m-hr
labor)
$11.5K
$13.6K
(incl. 62
m-hr
labor)
$21K
$25K
9-7
Figure 9-2
Operation of FCL Can Reduce Fault Current
Within First Half Cycle
Fault current limiters can provide technical
and economic benefits at several locations in
a power system (Noe and Oswald. 1999).
Potential FCL locations are summarized in
Figure 9-3.
1. Generator Connection
2. Station auxillaries
Figure 9-3
Example FCL Locations in Power System
3. Network coupling
4. Busbar coupling
Generator Connection
5. Busbar coupling
Addition of a new power station serves to
increase the short circuit capacity of the
network which would put increased stress on
the system during fault conditions. Use of
FCL on the feeder connecting the station to
the network would serve to limit the short
circuit capacity during fault conditions.
Economic benefits arise out of the option
that the renewal of the older substations can
be postponed till they reach their technical
lifetime.
6. Shunting current limiting reactors
7. Transformer feeder
8. Busbar connection
9. Combination
with
superconducting devices
other
10. Coupling local generating unit
11. Closing ring circuits
An example of adding a new generation to
the system is shown in Figure 9-4. This
situation could require utility to upgrade
multiple breakers in the generating station
due to increased short circuit capacity. The
required upgrade could be avoided by the
use of a SFCL to couple the new generator.
A real case investigated in 1996 in
Hannover
(Germany)
showed
that
introduction of SCFCL in generator feeders
would result in a major delay of investment
9-8
(approx. 30 million ) for upgrading old
substations (EPRI 2005).
If FCL is considered during initial design of
new generating substation, lower impedance
could be selected for generator and/or the
step-up transformer. That could potentially
save money in investment cost as well as by
reduced losses.
Figure 9-5
Networks Coupled through FCL
Coupling of Busbars
One possible configuration in which FCL
can be used to tie two MV buses being fed
by HV grid through separate transformers is
shown in Figure 9-6 .
Figure 9-4
New Generator Connection
Coupling of Networks
Network coupling with a SFCL leads to
advantages in energy flow, voltage stability
and security of supply without increasing the
short-circuit capacity in the networks (Noe
and Oswald. 1999). The economic benefits
include lower network losses and potential
savings in equipment cost as shown in the
example here.
Figure 9-6
Example of FCL in bus-tie
The advantages of this FCL application are
as follows:
• Reduction of the short circuit current
of the system when the tie-breaker is
closed
• Improved power quality (reduced
harmonics, sags and flicker) due to
reduced source impedance
• Higher system availability
• Higher loads possible in sub-systems
• Even loading of the feeding
transformers
An example of coupling of networks
through FCL is shown in Figure 9-5
(Neumann and Bock. 2004). It serves to free
the surplus transformer capacity that is
needed in the absence of the coupling. Thus,
high savings in investment costs and lower
losses can be obtained. In case of the
transformer outage in one network, the other
network can meet the power demand. Also,
in case of a fault in any network, FCL limits
the overall short-circuit current to
admissible values.
9-9
In another example shown in Figure 9-7, use
of FCL permits tie breaker CB7 to be closed
without the need of upgrading breakers
CB3,4,8 and 9 (EPRI 2005).
Figure 9-8
Busbars Coupled through FCL and low
impedance transformers in series with FCL
(Neumann and Bock. 2004)
Coupling of Local Generation
The technical and economical advantages of
using FCL for coupling distributed
generation such as wind power stations are
similar to those that were discussed for new
conventional power stations. It provides
more flexibility in selecting the location of
local generating stations as added short
circuit capacity is not a limiting factor any
more.
Figure 9-7
Busbars Coupled through FCL
Transformer feeder
Use of FCL in existing transformer feeder
would provide benefit of higher steady-state
and transient stability. It would allow utility
to meet the increase in load demand. It can
also delay the need for system improvement
as reduced fault current can offset any
planned increase in short circuit capacity
upstream of the transformer.
In some stations that have reached the limits
of admissible short circuit capacity, use of
FCL can avoid the need of coupling the new
local generation to the high voltage grid
through costly transformers (Figure 9-9).
For a new design, use of FCL in series with
a transformer would allow the selection of
subsequent devices with lower dimensioning
as they would be subjected to lower strain
during a fault resulting in savings. It also
allows selection of transformer with lower
impedance resulting in lower purchase cost
and reduced transformer losses during
normal conditions. An example that shows
such an application in combination with the
busbar coupling is shown in Figure 9-8.
Figure 9-9
Coupling of Local Generation ( EPRI 2005)
9-10
The estimate of the transformer purchase
cost as a function of rated MVA is shown in
Figure 9-11. Further, transformer production
costs decrease with the impedance up to a
certain limit (Sjöström and Politano. 2001).
Based on the survey results by CIGRE WG
(see Figure 9-10), the majority of the fault
current limiters will be installed in Bus-ties
(52%) and incoming feeders (33%).
Figure 9-11
Transformer purchase cost with and without
on-load tap changer (EPRI 2000)
For example consider a 30MVA transformer
without taps, with load losses (LL) of 0.35%
and no-load losses (excitation losses, EL) of
0.05%. Assume for evaluation:
Figure 9-10
Preferred locations for installing FCLs(
CIGRE 2003)
Economic Analysis of Individual
Components
Previous section discussed the various
applications where use of FCL can provide
technical and economic benefits to the
system. The economics of the individual
power system components such as breakers
and power transformers that could be
potentially downsized due to FCL
applications are discussed in this section.
The total owning cost of a transformer
consists of three major components (EPRI
2000).
Purchase price
•
Capitalized cost of load losses
•
Capitalized cost of no-load losses
•the purchase price
transformer is $400,000
•
•the discount rate over 12 years is
8.2%
•
•the cost of energy is 0.04$/kWh
•
•the
transformer
is
operated
continuously (8760 hours/year)
•
•the transformer is carrying on
average a load of 60% of its rating
and the load loss is a simple square
function of the load.
for
this
The capitalization factor for 12 years at
8.2% (present worth of a uniform
investment over the period) is 7.46.
Therefore the capitalized losses for this 30
year period is:
Power transformers
•
•
•
9-11
•Capitalized cost (EL) = 30,000kW x
0.0005pu x 0.04$/kW h x 7.46 x
8760h = $39,210.
•
Economic Analysis for FCLs
•Capitalized cost (LL) = 30,000kW x
0.0035pu x 0.62 x 0.04$/kWh x 7.46
x 8760h = $98,809.
The decision of a utility to opt for a FCL
application is naturally going to be dictated
by economic aspects in addition to its
technical capabilities. The results of the
surveys that were carried out by EPRI and
CIGRE regarding the price the customers
are willing to pay for FCL devices is
presented here from CIGRE WG 13.10
report (CIGRE 2003)
Therefore, the total owning cost is estimated
to be $400,000 + $39,210 + $98,809 =
$538,019.
In this example, use of FCL can allow
selection of transformer with reduced
impedance. This will impact the total
owning cost by virtue of resultant reduction
in purchasing cost and load losses.
For existing transformer feeders, use of FCL
will result in reduced mechanical stresses on
the transformer insulation during faults. This
means increased functional life of
transformer in terms of cycles of short
circuit current that it can withstand as per
the life cycle expression (Eq. 7-8) in Chapter
7.
Figure 9-12
Worth of FCL to a Customer (price unit: price of
a conventional CB)
Circuit Breakers
It was shown in previous section that the use
of FCL can be used to avoid the otherwise
necessary upgrade of circuit breakers. The
following equation may be used to compute
the savings due to the extension of the life of
the existing circuit breaker (Salama et al.
1993).
The methodologies that have been
developed for performing economic analysis
of the novel FCL technologies are presented
in the following sections.
Solid State FCLs
S CB = K re + K in + K FCB + KVCB MVASC − new Eq. 9-1
The economic analysis of a device that
serves the dual purpose of fault current
limiting and increased transmission power
capability is presented in (Salama et al.
1993). The proposed device (FCL-TCI)
limits the fault current by using a thyristor
controlled impedance. This novel FCL
design uses back-to-back thyristors in series
with an inductor and a capacitor, Figure
9-13. The series capacitor is intended to be
part of a line compensation scheme. Under
normal conditions, the firing angle is zero.
The control circuit operates to increase the
thyristor-firing angle when a fault is
Where,
Kre
= Cost of removal of the
existing circuit breaker
Kin
= Installation cost of the new
circuit breaker
KFCB
= Fixed circuit breaker cost
KVCB = Circuit breaker variable
cost per MVASC rating
MVASC-new = Short circuit rating of the
new circuit breaker
9-12
detected. This will result in limitation of the
fault current.
C I = K FL + KVL
n2
S
n n −1
2
2
Eq. 9-3
Cost of the capacitor
CC = K FC + KVC
S
n2
n2 n2 −1
Eq. 9-4
Where
n = design constant (typical value is
3)
S = FCL rating in MVA
KFL = Fixed inductor cost
Figure 9-13
Fault Current Limiter With Thyristor
Controlled Impedance (Salama et al. 1993)
KVL = Inductor variable cost per
MVar
The economic benefits of this application
are attributed to increased life of existing
breakers and transformers and reduced
losses in the power transmission. The total
cost of operation of the FCL can be
calculated to determine whether installation
is economically justified. In other words,
“Total cost” in the following equation
should have a negative value in order to
ensure the economic feasibility of the
application:
KFC = fixed capacitor cost
KVC = capacitor variable cost per
Mvar
Cost of the thyristor + cost of the control
circuit:
CT = K FT + KVT S
Eq. 9-5
Where
“Total cost = cost of the inductor + cost of
the capacitor + cost of the thyristor
+ cost of the control circuit + cost of
energy losses in the capacitor + cost
of energy losses in the inductor +
cost of energy lost in the thyristor the saving due to the extension of the
useful life of the existing breaker the saving due to the increase in the
transmitted power.”
KFT
= Fixed thyristor cost
KVT
= Thyristor variable cost per
Mvar
Present cost of energy losses in the
capacitor, inductor and thyristors:
⎛
⎞
S n2
PCE = 8760 f v f e f u ⎜⎜ f c S +
× 10 3 ⎟⎟ Eq. 9-6
2
nQ n − 1
⎝
⎠
where Q
C FCL = C I + CC + CT + PCE − S CB − S TP Eq. 9-2
Where:
Cost of the inductor
= Capacitor MVar
fv
= Present value factor
fe
= Cost of energy loss per
fu
= LC utilization factor
kWh
9-13
fc
kW/Mvar
= Capacitor loss factor
No numerical example was offered in
(Salama et al. 1993.). But, the approach
could be extended to do economic analysis
of other topologies of solid-state fault
current limiters.
in
The savings due to the extension of the life
of the circuit breaker by installation of the
FCL are given by:
Newer designs for Solid State FCLs are
discussed in (Meyer et al. 2004.). These
utilize a number of schemes for forced
commutation using capacitors to shut off
thyristors during a fault. The simplest
example, “topology a”, is shown in
Figure 9-14. Under normal conditions, load
current flows through TMain1 and TMain2, and
both capacitors are pre-charged as shown.
When a fault occurs, the auxiliary thyristor
connected to the main thyristor, which is
conducting at the time, will be fired,
discharging its capacitor. When the next
zero crossing is reached, the main thyristors
will be turned off.
S CB = K re + K in + K FCB + KVCB MVASC − new Eq. 9-7
where Kre
= Cost of removal of the
existing circuit breaker
Kin
= Installation cost of the new
circuit breaker
KFCB
= Fixed circuit breaker cost
KVCB = Circuit breaker variable
cost per MVASC rating
MVASC-new = Short circuit rating of the
new circuit breaker
Savings due to increased transmitted power
due to the series capacitor:
S TP
⎡
⎢
⎢
1
= KVTS PSIL ⎢
⎛ 1 − m2
⎢
⎢ 1 − m 2 sin⎜
⎜ m2
⎢
⎝
⎣
⎞
⎟
⎟
⎠
More complex circuits use smaller
capacitors, IGCTs as well as thyristors,
transformers,
positive
temperature
coefficient (PTC) resistors, and other
variations. These are designated topologies
“b” “c” “d” and “IGCT circuit breaker” in
the study.
⎤
⎥
1 ⎥
⎥
−
⎛ 1 ⎞⎥
sin⎜ ⎟ ⎥
⎝ m ⎠⎥
⎦
Economic comparison shows that the
simplest “topology a” is superior both in
investment costs, Figure 9-15, and in lifecycle costs (which are composed mainly of
losses as no maintenance is required),
Figure 9-16.
Eq. 9-8
Where,
m=
1
Eq. 9-9
ω C fL Lt
KVTS
= Incremental cost
installing a new transmission system
of
PSIL
= Surge impedance loading
(SIL) of the existing transmission system
Lt
=
Inductance
transmission system
CfL
FCL-TCI
of
the
= Series capacitance of the
9-14
Superconducting FCLs
The superconducting FCL (SFCL) is a new
technology under development. As such,
installed costs are not available. Various
approaches have been used to analyze the
economic benefits of deployment of these
devices in power systems:
Present Worth Method
This method is based on the calculation of
the prospective savings by using a SFCL. In
this method a reference cost is obtained
(Currency unit/kW), that can be compared
with the conventional devices for limiting
short circuit currents. The expression for the
reference cost of an SFCL is given as:
Figure 9-14
Forced Commutation Circuit for FCL (Meyer
et al. 2004)
cp =
Cp
Sr
where ,
CP = purchase price of the SFCL
Sr = rated power of the SFCL
Figure 9-15
Comparison of Investment Costs (Meyer et
al. 2004)
CP =
N
∑ (1 + i )
PWS − (C P 0 + C Pk )
n −1 − n
q
n =1
⎡⎛ y
⎤
⎞ ⎛
z⎞
⎢⎜ (1 + i ) xm q − xm ⎟ − ⎜1 − ⎟(1 + i )N q − N ⎥ +
⎟ ⎝ m⎠
⎢⎣⎜⎝ x =0
⎥⎦
⎠
∑
N
∑f
M
(1 + i )n −1 q −n
n =1
Eq. 9-10
Figure 9-16
Total Life-Cycle Costs of Topologies (Meyer
et al. 2004)
PWS
= present worth of the savings of
using an SFCL
CP0
= cost of no-load losses
CPk
= cost of load losses
N
= average life in years of
conventional methods of limiting
short circuit currents
m
= average life in years of the SFCL
i
= inflation rate
p
9-15
= interest rate
q = 1+
analysis, it can be deduced that the power
stations and wind generator interconnections
are the locations where use of SFCL would
provide maximum economical benefit.
p
100%
y
N/m
= even numbered share of
z = N − ym
Table 9-5
Data for Economic Evaluation (Noe and
Oswald. 1999.)
= maintenance costs per year
CM
Name
Symbol
Value
Average life
N
30 years
Interest rate
p
9% - 14%
Inflation rate
i
2% - 4%
Energy
charge
ce
0.05 – 0.06
$/kWh
Demand
charge
cd
113 – 169
$/kW
ce = energy charge
Maintenance
factor
fM
3%
TN= 8760 hours
SFCL losses
PSFCL
0.1% ST
fM =
CM
CP
Sr
SFCL
(assumed)
is the rated power of the
C p 0 = (c d + ceTN ) P0
Where,
cd = demand charge
(0.75% ST
for wind
generators)
Po = No load losses
C pk = (c d ha2 + vceTN )hr2 Pk
Where,
ha = ratio of active power of subnetwork at the time of the peak power of
the whole network to the peak active
power of the subnetwork.
SFCL no
load losses
P0
2/3 PSFCL
SFCL load
losses
Pk
1/3 PSFCL
Table 9-6
Main data of investigated network (Noe and
Oswald. 1999.)
hr =
ratio of maximum power to
rated power
v = 0.17m + 0.83m 2
Where
m= load factor
The data of an example system that was
used by Noe and Oswald in their analysis, is
given in Table 9-5 and Table 9-6. The
results of the economic evaluation for
various FCL locations (Figure 9-3) for
various physical lifetimes of FCL are
summarized in Figure 9-17. As per this
9-16
Parameter
Value
Annual maximum
demand
6312 MW
Annual quantity of
energy
3301 GWH
Line length 110 kV
148 km
Line length 10 kV
1470 km
Line length 0.4 kV
CI = investment costs
4263 km
COM = operation and maintenance costs
CF = yearly cost due to faults
40.00
n = service life in years
35.00
$/kVA
30.00
25.00
10 years
20 years
30 years
20.00
15.00
i = real index
Normalised cost benefit (B*C) is then
computed as:
10.00
5.00
0.00
Generatorconnection
Power
station
auxiliaries
Network
computing
Busbar
coupling
Transformer
feeder
Block-type
thermal
power
station
connection
Wind
generator
connection
BC* =
SFCL Location
Where,
Figure 9-17
Medium Specific Purchase Price of
Superconducting FCL With Losses (Noe and
Oswald 1999)
k = number of FCL
S = rated power
Strategic and cost benefits can then be
combined into a strategic-economic benefit
by:
Strategic-Economic Benefit Method
This method combines the strategic and
economic benefits of incorporating SFCLs
into the overall evaluation (Sjöström and
Politano. 2001.). Strategic benefit (BS) is the
weighted contribution of four different
aspects that have a value between -5 and +5
depending on their impact. The weightings
for the various aspects determined by their
relative importance are as follows:
*
*
BSE
= BC* + 0.4C FCL
BS
Where,
*
C FCL
= normalized life cycle cost of FCL
and is computed as:
*
C FCL
=
1. Safety and reliability – 0.5
2. Customer satisfaction – 0.25
4. Organizational benefit – 0.1
Cost benefit (BC) is more easily quantifiable
and is computed as the difference in LCC
for the existing system and the systems with
SFCL. The LCC cost is calculated using the
“present value method”:
LCC (n ) = C I +
∑ (C
j =1
OM
⎛ 1 ⎞
+ CF )⋅ ⎜
⎟
⎝1+ i ⎠
LCC fcl
S
The comparison of strategic-economic
benefit against the normalized life cycle cost
of FCL can be done to determine the
economic feasibility of the application (See
Table 9-7).
3. Environmental impact – 0.15
n
BC
kS
Table 9-7
Economic Feasibility of FCL Application
E q. 9-11
Where,
9-17
*
BSE
C*
/ FCL
Economic feasibility
>3
Very attractive
>5/3 & <3
Attractive
>1 & <5/3
Possible profit
>1/2 & <1
Not economical
<1/2
Bad application
increased power availability was also
evaluated. The results of these case studies
are summarized in Table 9-8. As per the
analysis, regional, low voltage and industrial
systems are the most economical locations
for FCL applications.
Sjöström and Politano conducted eight case
studies, for the following locations in Swiss
power system (Figure 9-18):
Table 9-8
Summary of strategic-economic analysis
results
1. Radial feeder, 200 kV
2. Substation 280/220 kV
3. Substation 220/110 kV
4. Industrial power system
5. Power plant 1.2 and 0.2 GW
6. Inter-regional distribution system at
60 kV
7. Regional distribution system at 16 kV
8. Low voltage system at 400 V
The cost reductions that are possible by
downsizing the individual equipments due to
the use of FCL can be significant as shown
in Table 9-9.
Table 9-9
Cost Savings for Designing a Power System
Using FCLs (Sjöström and Politano 2001)
Figure 9-18
FCL locations studied in Swiss grid
(Sjöström and Politano. 2001.)
For each study, one scenario involved
design of a new system that would permit
downsizing of the equipment due to the use
of FCL. The other scenario dealt with the
existing system in which the main focus was
on the impact of FCL on the increased
equipment life as it is subjected to lower
stresses during faults. The impact of
Equipment
Cost reduction by
introduction of an
FCL
Transformers
5% - 8%
Power circuit
breakers
5% - 15%
Bus-bars
3% - 15%
Cables
0% - 3%
Overhead lines
0%
System Integration Issues
The application of fault current limiters in
the utility network will require new
integration issues to be addressed. As
9-18
• Recovery voltage.
Therefore, the protection schemes and
settings need to be adapted for FCL
applications to ensure proper network
protection selectivity.
technologies like the solid-state and
superconducting fault current limiters come
closer to commercial reality, various
industry groups are now considering these
issues. (CIGRE. 2003, Steurer, 2004) The
issues like listed below need to be
addressed::
•
•
The impact of a solid-state based FCL on the
distance
protection
scheme
(mho
characteristics) in an example distribution
system is addressed in (Henry et al, 2003).
For the purpose of the analysis, they used a
fast switching FCL that utilized GTOs to
perform the fast switching function and a
resistor as its current limiting impedance
(Figure 9-19).
Impact of FCLs on system protection
schemes
1. Relay settings
2. Selectivity
3. Protection blinding (especially in
case of directional protection)
4. Compatibility with downstream
fuses
Impact of FCLs on system reliability
1. Unintended operation of FCL.
There is a chance of undesirable
operation of FCL due to
unavoidable inrush currents
(transformer
energization,
capacitor bank switching, motor
starting etc.)
2. Failure of FCL to operate under
fault conditions. It can be
catastrophic as fault current may
exceed the interrupting ratings of
the breakers.
3. Failure mode of FCL in case of
internal fault
4. Maintenance requirements of
FCL
Development of testing standards
and test procedures for FCLs
Figure 9-19
Fast Switching FCL (Henry et al, 2003)
Protection Coordination
The test system was modeled in EMTDC to
analyze the transient response of the
distance relay with and without the presence
of FCL. If the FCL is placed behind the
relay, the reach of the relay is determined by
the impedance of the extent of the line to be
protected. When the FCL is placed in front
of the relay, its setting must be adjusted to
take into account the added impedance of
FCL during a fault.
Depending on the technique used in a FCL,
the system may get impacted during the
fault conditions in following ways (Schmitt.
2006):
• Magnitude of fault current
• Phase angle of current
• Fault duration
It was found that relay selectivity is not
affected if the reach of relay is adjusted to
take into account the FCL impedance. In
fact, the introduction of FCL was found to
be beneficial to the response of relay as it
resulted in significant reduction in
oscillatory response of relay at the
•
9-19
protection boundary. It may be attributed to
the considerable reduction in ac and dc
component of the fault current due to FCL
action.
Table 9-10
SCFCL Model Parameters (Langston et al,
2005)
In the case of superconducting FCL, the
added impedance would change during the
transient of a fault. Therefore, relay settings
in such a scenario would be more complex
when the FCL is placed in the front of it. A
generic computer simulation model of a
resistive type superconducting FCL has been
developed and explained in Langston et al,
2005.
The authors used the commercially
simulation platform RTDS to study the
impact of FCL on a Schweitzer SEL-311B
distance protection relay. The SCFCL model
comprises of a superconducting element as a
variable resistor in parallel with a fixed
resistance. The model accounts for highly
non-linear
characteristics
of
superconducting element and also includes
the thermal aspects of the transient
phenomenon during fault conditions. The
thermal model includes the heat capacity of
the material as well as a first order
approximation of the heat transfer to the
surrounding
coolant.
The
relevant
mathematical equations used to model the
SCFCL are also provided in the paper. The
model parameters that were used are given
in Table 9-10..
The test-system comprises of a thevenin
source connected to a generator through two
transmission lines (T1 and T2) as shown in
Figure 9-20 .A distance relay having a mhocharacteristic is protecting line T2. Singleline-to-ground faults are applied at various
locations along transmission line T2, and the
behavior of the relay is studied
A.) without the SCFCL,
B.) with the SCFCL placed between T2 and
the PTs, and
C.) with the SCFCL placed between B2 and
the PTs.
Figure 9-20
Test set-up with FCL placed beyond PT
(Langston et al, 2005)
It was found that relay operates correctly for
the fault on T2 in the absence of SCFCL and
also when SCFCL is located behind PT as
the relay sees only the line impedance. In
the case corresponding to the location of
SCFCL in front of PT, the relay was unable
to detect some faults inside the protection
zone as it is sees additional impedance
introduced by the FCL during a fault.
9-20
The admissible temperature rise of non
accessible parts such as contacts in vacuum
interrupters, silicon wafers in semiconductor
devices or superconducting materials will
have to be considered separately. If a fault
current limiting device has an overload
capability this shall also be verified by tests.
Testing of FCLs
This section on testing requirements of
FCLs has been taken from the CIGRE WG
13.10 report (CIGRE 2003).
Standards with rules of the testing are
presently only available for fault current
limiting reactors (IEC 60289) and for highvoltage current limiting fuses (IEC 602821). Rules for the testing of other types of
fault current limiters need to be established
in the future. In this section some basic
considerations about the tests to be carried
out are given. It is understood that for
different types of fault current limiters
different test procedures will apply.
Short-Time Withstand Current Tests
In case of self-triggered fault current limiters
(e.g. superconducting fault current limiters)
the prospective short-circuit current shall be
applied to the device. The purpose of the test
is to verify the current limiting performance
(i.e. the initiating current, the limited current
and the follow current). External triggered
fault current limiters can be divided in two
sub-groups:
• Devices which are capable of
withstanding the prospective shortcircuit current of the system (e.g.
pyrotechnic fault current limiters): These
devices shall be subjected to a peak and
short-time withstand current test with the
prospective short circuit current without
any limiting operation. The operation of
the triggering device shall be tested
separately to verify the trigger levels
required in accordance with the ratings
of the system.
• Devices which are not capable of
withstanding the prospective shortcircuit current of the system (e.g. solidstate fault current limiters): These
devices shall be subjected to a peak and
short-time withstand current test with the
prospective short circuit current with the
triggering device operative. This test will
therefore at the same time serve to verify
current limiting performance.
Dielectric Tests
Dielectric tests as described in IEC 60694
have to be performed with the fault current
limiter in closed position between phase and
ground and between the phases. The test
voltages should be chosen in accordance
with Tables 1 and 2 of IEC 60694. If a fault
current limiter (or the combination of a fault
current limiter and a series switch) can have
an open position the dielectric performance
in this position has also to be verified,
independent of the nature of the gap (e.g.
solid-state switch, mechanical switch). The
voltage imposed on an open circuit-breaker
in a grid coupling could serve as a basis for
determining the test voltages in this case.
Additionally, the long term performance
shall be investigated, especially in the case
of semiconductors.
Temperature-Rise Tests
Temperature-rise tests including the
measurement of the resistance of the main
current path have to be performed in
accordance with IEC 60694. The test current
shall be equal to the rated current of the fault
current limiter. The temperature rise of the
contacts and other parts should be within the
limits specified in Table 3 of IEC 60694.
Short-Circuit Making and Breaking Tests
These tests apply to fault current limiting
devices with current interruption. The shortcircuit current breaking tests shall be carried
out at the rated voltage of the fault current
limiter. The source impedance of the test
9-21
circuit shall be chosen so that the required
prospective short circuit current flows in the
circuit. Tests at different fault initiation
angles are to be performed in order to verify
that the fault current limiter is capable of
interrupting
both
symmetrical
and
asymmetrical currents. The transient
recovery voltage of the test circuit shall be
defined taking into account the network
condition prevailing at the location where
the fault current limiter will be installed.
When a fault current limiter can be used for
closing a circuit, short-circuit current
making tests need also to be carried out.
increased life of other equipment, reduced
losses etc. over the life time of FCL.
There are several issues that will need to be
resolved in order to ensure the smooth
integration of novel FCLs into the utility
networks. These issues include protection
coordination with the existing system,
system reliability and development of
uniform testing standards and procedures for
FCls.
References
1. Das, J.C. 1997. “Limitations of FaultCurrent Limiters for Expansion of
Electrical Distribution Systems,”
IEEE Transactions on Industry
Applications, Vol. 33, No. 4,
July/August 1997, pp. 1073-1081.
Endurance Tests
In case of fault current limiters suitable for
more than one limiting operation an
endurance test with an appropriate number
of limiting operations shall be carried out.
2. CIGRE WG A3.10, 2003, “Fault
Current Limiters in Electrical
Medium and High Voltage Systems”,
CIGRE Technical Brochure 239,
December 2003.
Electromagnetic Compatibility (EMC) Tests
Electromagnetic compatibility tests shall be
carried out in accordance IEC 60694.
Depending on the type of fault current
limiter and triggering device it may be
advisable to supplement the tests described
in IEC 60694 by additional EMC-tests.
3. Schmitt, A, and CIGRE Working
Group A3.16, “Fault Current Limiters
Report on the activities of CIGRE
WG A3.16”, Power Engineering
Society General Meeting, IEEE,
2006.
Summary and Recommendations
There are several technical and economic
advantages of fault current limiting methods.
FCLs can provide benefits at several
locations in power system but the majority
applications are likely to be at incoming
generator and transformer feeders, bus-ties
and distributed generation connections.
4. EPRI 2000. Evaluate Solid State LTC
options for Medium Power
Transformers: Project 41C3084/66586424, EPRI, Palo Alto, CA.
5. EPRI. 2005. Survey of Fault Current
Limiter (FCL) Technologies. EPRI,
Palo Alto, CA: 2005. 1010760.
It is easier to perform economic analysis of
conventional solutions such as CLRs as their
capital cost is known. On the other hand, the
purchase cost of novel technologies such as
solid-state and superconducting FCLs that
are still in development is still not known.
Any economic analysis should also factor in
additional economic benefits such as
6. Henry, S., Baldwin, T., Steurer, M.
“The effects of a Fast Switching Fault
Current Limiter on Distance
Protection,” IEEE 2003
9-22
7. Power Quality Impact of Distributed
Generation, EPRI, Palo Alto, CA:
2005. 1008507.
Emerging Utility Integration Issues”,
IEEE PES General Meeting, June
2004.
8. Meyer, C.; Kollensperger, P.; De
Doncker, R.W.; 2004. “Design of a
novel low loss fault current limiter for
medium-voltage systems,” APEC '04.
Nineteenth Annual IEEE Applied
Power Electronics Conference and
Exposition. Vol.3, pp.1825 – 1831.
9. Neumann, C., and Bock, J. “Three
phase resistive fault current limiterimpact on system design” presented at
ASC2004, Jacksonville, USA.
10. Noe, M.; Oswald, B.R.; 1999.
“Technical and economical benefits
of superconducting fault current
limiters in power systems,” IEEE
Transactions
on
Applied
Superconductivity, Volume 9, Issue 2,
Part 1, June 1999 pp.1347 – 1350.
11. Salama, M.M.A.; Temraz, H.;
Chikhani, A.Y.; Bayoumi, M.A.;
1993. “Fault-current limiter with
thyristor-controlled
impedance,”
IEEE Transactions on Power
Delivery, Volume 8, Issue 3, July
1993 pp.1518 – 1528.
12. Sjöström, M.; Politano, D.; 2001.
“Technical and economical impacts
on a power system by introducing an
HTS FCL,” IEEE Transactions on
Applied Superconductivity, Volume
11, Issue 1, Part 2, March 2001 pp.
2042 – 2045.
13. Slade, P.G., Voshall, R.E., Wu, J.L.,
Stacey, E. J., Stubler, W. F.,
Talvacchio, J. “Study of Fault Current
Liting Techniques”, EPRI-Report EL6903, 1990.
14. Steurer, M., M. Noe, F. Breuer, 2004,
“Fault Current Limiters . R & D
Status of Two Selected Projects and
9-23
expressed in terms of the crest value of
the”standard lightning impulse.”
G
GLOSSARY
Bundled Conductor. An assembly of two
or more conductors used as a single
conductor and employing spacers to
maintain a predetermined configuration. The
individual
conductors
are
called
subconductors.
AC (Alternating Current) Transmission.
The transfer of electric energy by alternating
current from its source to one or more main
receiving
stations
for
subsequent
distribution.
ACSR. Aluminium
Reinforced.
Conductor
Bus. Connection between components of
electrical substations such as switchgear,
transformers and exit lines. A bus is
generally three-phase, and is composed of
flexible or rigid conductor sections mounted
on insulators.
Steel
Air Circuit Breakers. Air circuit breakers
are used to interrupt circuits while current
flows through them. Compressed air is used
to quench the arc when the connection is
broken.
Capacitor. An electric circuit element used
to store charge temporarily, consisting in
general of two metallic plates separated and
insulated from each other by a dielectric. It
comes in a huge variety of sizes and types
for use in regulating power as well as for
conditioning, smoothing and isolating
signals.
Alternating Current. An electric current
that reverses direction at regular recurring
interval of times.
•
Axial Stress. Axial (or Normal) Stress,
often symbolized by the Greek letter sigma,
is defined as the force perpendicular to the
cross sectional area of the member divided
by the cross sectional area.
•
large investments in testing facilities to
qualify
Italian
electromechanical
industry
skills on power system analysis for the
development of the national grid.
Ferromagnetic Hysteresis. When a
ferromagnetic material is magnetized in one
direction, it will not relax back to zero
magnetization
when
the
imposed
magnetizing field is removed. It must be
driven back to zero by a field in the opposite
direction. If an alternating magnetic field is
applied to the material, its magnetization
will trace out a loop called a hysteresis loop.
The lack of retraceability of the
magnetization curve is the property called
Bending stress. A compressive and/or
tensile stress resulting from the application
of a nonaxial force to a structural member.
BIL. Basic lightning impulse insulation
level is the electrical strength of insulation
G-1
hysteresis and it is related to the existence of
magnetic domains in the material.
Fuses. Safety device used to protect an
electric circuit against an excessive current.
A fuse consists essentially of a strip of lowmelting alloy enclosed in a suitable housing.
It is connected in series with the circuit it is
to protect. Because of its electrical
resistance, the alloy strip in the fuse is
heated by an electric current; if the current
exceeds the safe value for which the fuse
was designed, the strip melts, opening the
circuit and stopping the current. The fuse
housing is designed to resist the pressure
generated if the overcurrent vaporizes the
alloy strip, provided the voltage across the
fuse does not exceed its rating. Some fuses,
called slow-blow fuses, are designed to
carry a small overload for a short time
without opening the circuit, while others are
designed to open very rapidly if the rated
current is exceeded. The choice of one type
or the other depends on the ruggedness of
the equipment to be protected and whether
large pulses of current often occur in the
circuit; a slow-blow fuse is usually used to
protect motors, and a fast-blow fuse to
protect electronic equipment. A circuit can
also be protected by a circuit breaker.
Finite Element Analysis. A mathematical
technique for analyzing stress, which breaks
down a physical structure into substructures
called "finite elements." The finite elements
and their interrelationships are converted
into
equation
form
and
solved
mathematically.
Flashover. Disruptive discharge through air
directly or around or over the surface of
solid or liquid insulation between electrodes
of different potential or polarity, produced
by the application of voltage wherein the
breakdown path becomes sufficiently
ionized to maintain an electric arc.
Fixed Bus Bar End. The end of a rigid bus
bar that is not free to rotate.
Fiberglass Reinforced Plastic (FRP). It is
a composite made from fiberglass
reinforcement in a plastic (polymer) matrix.
A construction analogy would be the steel
reinforcing bars in a concrete matrix for
highways.
Fusing Current. The amount of current that
a conductor can conduct without burning.
By reinforcing the plastic matrix, a wide
variety of physical strengths and properties
can be designed into the FRP composite.
Additionally, the type and configuration of
the reinforcement can be selected, along
with the type of plastic and additives within
the matrix. These variations allow an
incredible range of strength and physical
properties to be obtained. FRP composites
can be developed specifically for the
performance required versus traditional
materials: wood, metal, ceramics, etc.
Gas Insulated Substation (GIS). For
applications where space requirements are a
problem, conventional bus arrangements are
replaced with gas-insulated substations. The
gas is generally SF6. Each conductor is
placed in the center of an enclosure filled
with SF6, which exhibits excellent dielectric
strength and, therefore, allows small
distances between the energized conductor
and the enclosure, which is at ground
potential.
G-2
electromagnetics,
electroacoustics,
multimedia, telecommunication, and energy
production and distribution, as well as
associated general disciplines such as
terminology and symbols, electromagnetic
compatibility,
measurement
and
performance, dependability, design and
development, safety and the environment.
Grading Ring. An electrode of toroidal or
similar in shape that is placed at the ends of
conductors, insulator strings, bushings, etc.
in order to grade the electric field and lower
the surface gradient on the metallic
components and on the di-electric surfaces
of insulators. They are often applied on
polymer insulators, and may have
“horseshoe” shapes to facilitate installation.
. IG = D f × I g
Heat Sink. A material that absorbs heat.
Typically made of aluminum, heat sinks are
widely used in amplifiers and other
electronic devices that build up heat.
Modulus of Elasticity. A measure of the
resistance of material to deformation. It is
the ratio of normal stress corresponding
strain for tensile or compressive stresses
below the proportional limit of the material.
High Temperature Superconductors
(HTS).
The
term
High-temperature
superconductor was initially employed to
designate the new family of cuprateperovskite ceramic materials discovered by
J.G. Bednorz and K.A. Müller in 1986.
These materials are characterized by
presenting superconductivity at a higher
temperature
than
conventional
superconductors
(which
require
temperatures a few degrees above absolute
zero (• 273.15 °C or • 459.67 °F)), and by
other unconventional features. So-called
high-temperature
superconductors
are
generally considered to be those that
demonstrate superconductivity at or above
the temperature of liquid nitrogen, or • 196
°C (77 K).
Moment of Inertia. The property of an
object associated with its resistance to
rotation. It depends on the objects mass and
the distribution of mass with respect to the
axis of rotation.
National Electrical Code (NEC). Published
volume of rules whose purpose is “the
practical safeguarding of persons and
property from hazards arising from the use
of electricity.”
NCI. Non-Ceramic Insulator.
ORNL. Oak Ridge National Laboratory
(ORNL) is a multi-program science and
technology national laboratory managed for
the United States Department of Energy by
UT-Battelle, LLC. ORNL is located in Oak
Ridge, Tennessee, near Knoxville. Scientists
and engineers at ORNL conduct basic and
applied research and development to create
scientific knowledge and technological
solutions that build the nation's expertise in
IEC. The International Electrotechnical
Commission (IEC) is the leading global
organization that prepares and publishes
international standards for all electrical,
electronic and related technologies. These
serve as a basis for national standardization
and
as
references
when
drafting
international tenders and contracts. The IEC
charter embraces all electrotechnologies
including electronics, magnetics and
G-3
key areas of science; increase the
availability of clean, abundant energy;
restore and protect the natural environment;
and contribute to national security. ORNL
also performs other work for the Department
of Energy, including isotope production,
information management, and technical
program management, and provides research
and
technical
assistance
to
other
organizations.
Polymer. It is a term used to describe a very
large molecule consisting of structural units
and repeating units connected by covalent
chemical bonds. The term is derived from
the Greek words: polys meaning many, and
meros meaning parts. The key feature that
distinguishes polymers from other molecules
is the repetition of many identical, similar,
or complementary molecular subunits in
these chains. These subunits, the monomers,
are small molecules of low to moderate
molecular weight, and are linked to each
other during a chemical reaction called
polymerization.
Partial
Discharge.
In
electrical
engineering, a partial discharge (PD) is a
localised dielectric breakdown of a small
portion of a solid or liquid electrical
insulation system under high voltage stress.
While a corona discharge is usually revealed
by a relatively steady glow or brush
discharge in air, partial discharges within an
insulation system may or may not exhibit
visible discharges, and discharge events tend
to be more sporadic in nature than corona
discharges.
Protection Relays. These devices will sense
the fault and initiate a trip, or disconnection,
order. A protection relay is a complex
electromechanical apparatus, often with
more than one coil, designed to calculate
operating conditions on an electrical circuit
and trip circuit breakers when a fault was
found. Unlike switching type relays with
fixed and usually ill-defined operating
voltage thresholds and operating times,
protection relays had well-established,
selectable, time/current (or other operating
parameter) curves. Such relays were very
elaborate, using arrays of induction disks,
shaded-pole magnets, operating and restraint
coils, solenoid-type operators, telephonerelay style contacts, and phase-shifting
networks to allow the relay to respond to
such conditions as over-current, overvoltage, reverse power flow, over- and
under- frequency, and even distance relays
that would trip for faults up to a certain
distance away from a substation but not
beyond
that
point.
An
important
transmission line or generator unit would
have had cubicles dedicated to protection,
with a score of individual electromechanical
devices. Related terms include:
Present Value Method. Present value is the
current worth of future sums of money. The
process of calculating present value is
actually the opposite of finding the
compounded future value. The present value
method, also called the present worth
method, is widely used in corporate finance
to evaluate a proposed capital investment
project or to measure the expected return.
Pinch force. Maximum tensile force in a
bundled flexible conductor due to the
attraction of the sub-conductors in the
bundle.
Pinned Bus Bar End. The end of a rigid
bus bar that is free to rotate.
G-4
replaced thermionic devices (vacuum tubes)
in most applications. They use electronic
conduction in the solid state as opposed to
the gaseous state or thermionic emission in a
high vacuum. Semiconductor devices are
manufactured as single discrete devices or
integrated circuits (ICs), which consist of a
number—from a few devices to millions—
of devices manufactured onto a single
semiconductor substrate.
Differential Relay. A relay with multiple
windings that functions when the voltage,
current, or power difference between the
windings reaches a predetermined value.
Electromechanical Relay. An electric relay
in which the designed response is developed
by the relative movement of mechanical
elements under the action of a current in the
input circuits. (IEC).
Series Compensator (SC). It is a power
electronics based device that acts to mitigate
voltage perturbations due to faults in power
system.
Numerical Relay. Numerical relays emulate
their electromechanical ancestors with great
precision and convenience in application. By
combining several functions in one case,
numerical relays also save capital cost and
maintenance cost over electromechanical
relays.
SF6 Circuit Breakers. SF6 Circuit
Breakers operate to switch electric circuits
and equipment in and out of the system.
These circuit breakers are filled with
compressed sulfur-hexafluoride gas which
acts to open and close the switch contacts.
The gas also interrupts the current flow
when the contacts are open.
Solid-State Relay. A relay whose functions
are achieved by means of electronic
components and without the use of moving
parts.
Shell-form Transformer. A transformer in
which all the windings are on the center of
three legs.
Resonance. The state of a system in which
an abnormally large vibration is produced in
response to an external stimulus, occurring
when the frequency of the stimulus is the
same, or nearly the same, as the natural
vibration frequency of the system.
Short Circuit Tensile Force. Maximum
tensile force in a flexible main conductor
due to swing out reached during a short
circuit.
SCADA System. Supervisory Controls and
Data Acquisition System.
Single-Phase. Producing, carrying,
powered by a single alternating voltage.
Semiconductor Devices.
Electronic
components that exploit the electronic
properties of semiconductor materials,
principally silicon, germanium, and gallium
arsenide. Semiconductor devices have
or
Skin Effect. The skin effect is the tendency
of an alternating electric current to distribute
itself within a conductor so that the current
density near the surface of the conductor is
G-5
point of installation, maintains the clearance
between sub-conductors.
greater than that at its core. That is, the
electric current tends to flow at the “skin” of
the conductor. The highr the frequency, the
more the skin effect and the greater the
resistance. Stranded wire produces less skin
effect than solid, because there is more
surface area. The skin effect enables
copperclad steel wire to be used. The steel
adds cable strength, and the current flows
mostly through the better-conducting
copper.
Span. Distance between contiguous power
line support structures
Spring Constant. Hooke's Law states that
for small forces and extensions, the force on
a spring is proportional to its extension. The
constant of proportionality k is called the
spring constant and is a property of the
material and shape of the spring.
Slack Bus. A bus made of flexible
conductors, which hangs from post
insulators, such that
Error! Objects cannot be created
from editing field codes.
SFCL. Superconducting
Limiter
fault
Current
where
l is the distance between supports (m)
SSB. Solid State Breaker
lc is the length of bus conductor (m)
SSCL. Solid State Current Limiter
SML. Specified mechanical load.
Snubber. Snubber consists of just a small
capacitor in series with a small resistor. This
combination can be used to suppress the
rapid rise in voltage across a thyristor,
preventing the erroneous turn-on of the
thyristor; it does this by limiting the rate of
rise in voltage (dV/dT) across the thyristor
to a value which will not trigger it.
Step Voltage. If a person is standing on the
surface, and the flow of ground current
causes a dangerous voltage drop to occur
between their feet, they are exposed to a step
voltage. It may be calculated as the
difference in surface potential experienced
by a person bridging a distance of 1 m with
the feet without contacting any grounded
object.
Solid-State. Pertaining to circuits where
signals pass through solid semiconductor
material such as transistors and diodes as
opposed to vacuum tubes where signals pass
through a vacuum.
Stiffener. A special spacer intended to
reduce the mechanical stress of rigid
conductors.
Error! Objects cannot be created from
editing field codes.
Spacer. A mechanical element between subconductors, rigid or flexible, which, at the
G-6
electricity consumers to the main
transmission network (unless they
use large amounts of energy); so the
distribution station reduces voltage
to a value suitable for connection to
local loads.
Strain-bus Structure. A bus structure
comprised of flexible conductors supported
by strain insulators, such that
Error! Objects cannot be created
from editing field codes.
where
l is the distance between supports (m)
Superconductivity. A property of some
materials in which their electrical resistance
drops to zero, and they acquire the ability to
carry electric current with no loss of energy
whatsoever. Formerly, materials developed
superconductivity only at temperatures near
absolute zero, but new materials have been
found that remain superconductive at
temperatures above those of liquid nitrogen.
lc is the length of bus conductor (m)
li is the length of one insulator chain (m)
Sub-conductor. A single conductor which
carries a certain part of the total current in
main phase and is a part of the main
conductor.
Surface Material. A material installed over
the soil consisting of, but not limited to,
rock or crushed stone, asphalt, or man-made
materials. The surfacing material, depending
on the resistivity of the material, may
significantly impact the body current for
touch and step voltages involving the
person’s feet.
Substation. A substation is a subsidiary
station of an electricity generation,
transmission and distribution system where
voltage is transformed from high to low or
the reverse using transformers. Related
terms include:
Transmission
Substation.
A
transmission substation is one whose
main purpose is to connect together
various transmission lines. The
simplest case is where all
transmission lines have the same
voltage. In such cases, the substation
contains high-voltage switches that
allow lines to be connected together
or isolated for maintenance.
Switchgear. The term switchgear, refers to
the combination of electrical disconnects
and/or circuit breakers meant to isolate
equipment in or near an electrical substation.
For transmission levels of voltage (high
voltages over 66 kV), often switchgear will
be mounted outdoors and insulated by air,
though this requires a large amount of space.
A compact, though more costly form of
switchgear is "gas insulated switchgear"
(GIS), where the conductors and circuit
breakers are insulated by sulfur hexafluoride
gas.
Distribution
Substation.
A
distribution substation is one whose
main purpose is to transfer power
from the transmission system to the
distribution system of some area. It
is uneconomical to directly connect
Switching Surge. Voltage surge resulting
from a switching operation.
G-7
magnitudes of the components within each
of the three phase sequences (positive,
negative, zero) are equal, but they may or
may not be equal to each other in
magnitude. The rotating vectors of each of
the three sets of components may be shifted
by some electrical angle from a common
reference point in the rotating vector
diagrams. A physical interpretation of
positive sequence components, for example,
would be the currents that would occur on a
balanced power system; the sum of their
instantaneous values is zero. Negative
sequence components would exist (along
with positive sequence components) in an
unbalanced system in which the phase
current magnitudes are numerically unequal
but still sum to zero. Zero sequence currents
exist when there is a net current (i.e., the
instantaneous values of the phase currents
do not sum to zero. The zero sequence phase
currents would flow in the neutral or ground
paths.
Symmetrical Components. Mathematical
technique developed by C. L. Fortescue
(published in 1918) that can be applied to a
variety of engineering problems. It is widely
used for solving power engineering
problems involving unsymmetrical (or
unbalanced) power systems. In three-phase
power systems, it is applied to current,
voltage, and impedance problems. The
method is used to transform an unbalanced
three-phase system into three sets of
balanced three-phase phasors. The method
of symmetrical components is one form of a
general matrix transformation. In this
reference book, the method is applied
usually to circuit or load condition issues
that could affect EMF levels and attenuation
characteristics. In general, an unbalanced
system can be resolved into balanced,
symmetrical, three-phase or single-phase
vector
systems
(called
symmetrical
components), and used to obtain solutions to
the original unbalanced, nonsymmetrical
problem. The unbalanced system is uniquely
resolved into three sets of balanced phasesequence vectors called: positive sequence,
negative sequence, and zero sequence
components. In the method of symmetrical
components, all of the resolved phasesequence vectors rotate in the positive
(counterclockwise) direction. The three
vector components of the positive sequence
components are equal to each other in
magnitude, are 120 electrical degrees apart
in phase, and achieve maximum values in
the positive phase rotation sequence of A, B,
C. The three negative sequence vector
components are also equal to each other in
magnitude, are 120 electrical degrees apart,
and rotate counterclockwise, but in the
negative phase sequence of A, C, B. The
zero sequence vector components are
likewise equal to each other in magnitude,
but all three components are in phase (i.e.,
there is a zero degree angle or zero sequence
between the three components). The
Tensile Force. A stretching force pulling at
both ends of a component or structure along
its length.
Thyristor. It is semiconductor switch used
chiefly in power-control applications. Also
called a silicon-controlled rectifier (SCR), it
is a variation of the transistor. A thyristor is
capable of producing large direct currents by
rectification of alternating currents and can
be automatically triggered “off” for
specified periods of time. Thyristors are
used in variable-speed electric motors,
power
supplies
for
electrochemical
processes, lighting and heating control, and
controllers for electric utility power systems.
Time-to-saturation. The time during which
the secondary current is a faithful replica of
the primary current in a current transformer.
G-8
Transmission Line. A transmission line is
the material medium or structure that forms
all or part of a path from one place to
another for directing the transmission of
energy, such as electromagnetic waves or
acoustic waves, as well as electric power
transmission. Components of transmission
lines include wires, coaxial cables, dielectric
slabs, optical fibres, electric power lines,
and waveguides.
Touch Voltage. The highest voltage
potential difference between a conductive
structure and a point on the earth’ssurface
separated by a distance equal to the normal
maximum horizontal reach, approximately 1
m. This voltage potential, applied between
hand and foot, can cause a body current in
excess of safe levels defined by fibrillation
current.
Transformer. A transformer is an electrical
device that transfers energy from one circuit
to another by magnetic coupling with no
moving parts. A transformer comprises two
or more coupled windings, or a single
tapped winding and, in most cases, a
magnetic core to concentrate magnetic flux.
An alternating current in one winding
creates a time-varying magnetic flux in the
core, which induces a voltage in the other
windings. Transformers are used to convert
between high and low voltages, to change
impedance, and to provide electrical
isolation between circuits.
Turns ratio. The ratio of the secondary
winding turns to the primary winding turns.
Underbuild. Term used in transmission-line
engineering to describe lower voltage
circuits placed below and on the same
structure as the phase conductors of a
transmission line.
Varistor. A varistor is an electronic
component with a significant non-ohmic
current-voltage characteristic. The name is a
portmanteau of variable resistor. Varistors
are often used to protect circuits against
excessive transient voltages by incorporating
them into the circuit in such a way that,
when triggered, they will shunt the current
created by the high voltage away from the
sensitive components.
Transformer Short-Circuit Impedance
For Category I and Category II transformers,
the transformer impedance, expressed in
percent on the transformer’s rated voltage
and rated base kilovoltamperes. For
Category III and Category IV transformers,
the sum of transformer impedance and
system short-circuit impedance at the
transformer location, expressed in percent
on the transformer’s rated voltage and rated
base kilovoltamperes
Voltage Regulator Device or circuit that
maintains constant output voltage (within
certain limits) in spite of changing line
voltage and/or load current.
Voltage Sag. A decrease to between 0.1 and
0.9 pu in rms voltage at the power frequency
for durations of 0.5 cycle to 1 min.
Transient Recovery Voltage (TRV). The
transient voltage that occurs across an
opening contact- for example, in a circuit
breaker.
G-9
Zero Crossing. In alternating current, the
zero crossing is the instantaneous point at
which there is no voltage present. In a sine
wave or other simple waveform, this
normally occurs twice during each cycle.
Weibull Distribution. The Weibull
distribution is most commonly used in life
data analysis, though it has found other
applications as well. The Weibull
distribution is often used in place of the
normal distribution due to the fact that a
Weibull variate can be generated through
inversion, while normal variates are
typically generated using the more
complicated Box-Muller method, which
requires two uniform random variates.
Zero Sequence. A balanced set of voltages
or currents in a symmetrical component
analysis corresponding to the average of the
phase voltages or currents involving a return
path outside the phase conductors.
WINIGS. The program WinIGS performs
analysis and design of a grounding system
or multiple grounding systems that are an
integral part of an electric power system.
Specifically, it allows the user to model any
power system together with its grounding
structures, it analyzes the performance of the
system under steady state, normal, and fault
conditions, and evaluates its performance
against industry-standard criteria. The user
may select either the IEEE Std. 80 criteria or
the IEC-479-1 criteria, both of which have
been integrated into the program.
Zero
Sequence
Components.
Symmetrical Components.
X/R ratio. Ratio of the system reactance to
resistance. It is indicative of the rate of
decay of any dc offset. A large X/R ratio
corresponds to a large time constant and a
slow rate of decay.
Young Modulus. Within the limits of
elasticity, the ratio of the linear stress to the
linear strain is termed the modulus of
elasticity or Young's Modulus and may be
written Young's Modulus, or E =
(Stress/Strain) It is this property that
determines how much a bar will sag under
its own weight or under a loading when used
as a beam within its limit of proportionality.
G-10
See
Export Control Restrictions
The Electric Power Research Institute (EPRI)
Access to and use of EPRI Intellectual Property is
granted with the specific understanding and
requirement that responsibility for ensuring full
compliance with all applicable U.S. and foreign export
laws and regulations is being undertaken by you and
your company. This includes an obligation to ensure
that any individual receiving access hereunder who is
not a U.S. citizen or permanent U.S. resident is
permitted access under applicable U.S. and foreign
export laws and regulations. In the event you are
uncertain whether you or your company may lawfully
obtain access to this EPRI Intellectual Property, you
acknowledge that it is your obligation to consult with
your company’s legal counsel to determine whether
this access is lawful. Although EPRI may make
available on a case-by-case basis an informal
assessment of the applicable U.S. export classification
for specific EPRI Intellectual Property, you and your
company acknowledge that this assessment is solely
for informational purposes and not for reliance
purposes. You and your company acknowledge that it
is still the obligation of you and your company to make
your own assessment of the applicable U.S. export
classification and ensure compliance accordingly. You
and your company understand and acknowledge your
obligations to make a prompt report to EPRI and the
appropriate authorities regarding any access to or use
of EPRI Intellectual Property hereunder that may be in
violation of applicable U.S. or foreign export laws or
regulations.
The Electric Power Research Institute (EPRI), with
major locations in Palo Alto, California, and Charlotte,
North Carolina, was established in 1973 as an
independent, nonprofit center for public interest energy
and environmental research. EPRI brings together
members, participants, the Institute’s scientists and
engineers, and other leading experts to work
collaboratively on solutions to the challenges of electric
power. These solutions span nearly every area of
electricity generation, delivery, and use, including
health, safety, and environment. EPRI’s members
represent over 90% of the electricity generated in the
United States. International participation represents
nearly 15% of EPRI’s total research, development, and
demonstration program.
Together…Shaping the Future of Electricity
© 2006 Electric Power Research Institute (EPRI), Inc. All rights
reserved. Electric Power Research Institute and EPRI are registered
service marks of the Electric Power Research Institute, Inc.
Printed on recycled paper in the United States of America
ELECTRIC POWER RESEARCH INSTITUTE
3420 Hillview Avenue, Palo Alto, California 94304-1338 • PO Box 10412, Palo Alto, California 94303-0813 • USA
800.313.3774 • 650.855.2121 • askepri@epri.com • www.epri.com
1012419