High Voltage Circuit Breakers

7/15/2019 High Voltage Circuit Breakers

http://slidepdf.com/reader/full/high-voltage-circuit-breakers-5632800234839 1/381



Library of Congress Cataloging-in-Publication Data

Garzon,R D. p u b e n D.)

High voltage circuit breakersdesign and applications Ruben Garzon.

Includes index.ISBN 0-8247-9821-X (hardcover paper)1. Electric circuit-breakers. 2. Electric power distribution-

-High tension-Equipment and supplies. I. Title.11.

Series.TK2842.G279966 2 1 . 3 1 ' 7 4 ~ 2 1

p. c m .- Electrical engineering and electronics v. 100)

96-37206C P

The publisher offersdiscounts on book when ordered in bulk quantities.

For more information, write to Special SalesProfessional Marketing at theaddress below.

This book is printed on acid-& paper.

Copyright 1997by MARCEL DEKKER, INC. All Rights Reserved.

Neither book nor any part may be reproduced or transmitted in any formor by any means, electronic or mechanical, including photocopying, micro-filming,and recording, or by any information storage and retrieval system,without permissionn writing from the publisher.

MARCEL DEKRER, INC.

270 Madison Avenue, New York,New York 10016

Current printing last digit):1 0 9 8 7 6 5 4 3 2 1

PRINTED NTRE JNITJ3D STATES OF AMERICA



To my wife,Maggi, and to

Gigi,Mitzi andNatalie

m e ourpillars of my life



This Page Intentionally Left Blank



PREFACE

Ever since the time when electrical energy was beginningo be utilized, therehas been a need for suitable switching devices capablef initiating and inter-rupting the flow the electric current. The early designs of such switchingdevices were relatively crude andhe principles of their operation relied onlyon empirical knowledge. Circuit breakers were developed on the basis of *a"cut and try" approach, but as the electrical system capacity continued to de-velop and grow, a more scientifk approach was needed to achieve optimized

designs of circuit breakers that would offer higher performance capabilitiesand greater reliability.The transition of current interruption from being an empirical to an

applied science began n the 1920's. It was only then that worldwide researchstarted to unravel the subtleties of the electric arc and its signifcance on thecurrent interruption process. Since those early research days, there has been agreat deal of literature on he subject of current interruption. There havebeen numerous technical articles on specific applications of circuit breakers,

but most of these publicationsare highly theoretical. What has been missingare publications geared specifically tohe needs the practicing engineer-asimple source of reference that provides simple answers to the most oftenasked questions: Where does come from? What does t mean? What can Ido with it? How can I uset? How can specify the right kindof equipment?

Circuit breakers are truly unique devices. They are a purely mechanicalapparatus connected to the electrical system. They must systematically inter-act with the system, providing a suitable ath for the flow the electric

rent; furthermore, they must provide protection and control ofhe electric cir-cuit by either initiating or stopping the current flow. Combining these tasks

into one device requires a close interaction f two engineering disciplines. Agoodunderstanding of mechanicalandelectricalengineeringprinciples is

paramount for the proper design and applicationf any circuit breaker.The purposeof book is to bridge the gap between theory and practice,

and to do without losing sight ofhe physics of he interruption phenom ena.The approach is to describe the most common application and design require-

ments and their solutions based on experience and present established prac-tices. A strictly mathematical approach s avoided; however, the fundamentalsof the processes are detailed and explained from a qualitative pointf view.

V



Beginning with a simplified qualitative, rather than quantitative, descrip-tion of the electric arc and ts behavior during the time when current is beinginterrupted, we will then proceed to describe the response of the electric sys-

tem and he inevitable interaction f current and voltage duringhe critical ini-tial microseconds following he interruption of the current. We will show thespecific behavior of different types of circuit breakers under different condi-tions.

After explaining what a circuit breaker must do, we will proceed to de-scribe the most significant design parameters of such device. Particular em-phasis will be placed in describing the contacts, their limitations in terms ofcontinuous current requirements and possible overload conditions, and theirbehavior as the result of the electromagnetic forces that are present duringshort circuit conditions and high inrushcurrentperiods.Typicaloperatingmechanisms willbe described and he terminology and requirements or thesemechanisms will e presented.

the years performance standards have been developed not onlyn theUnited Statesbut in other partsof the world. Today, with he world tending obecome a single market, it is necessary to understand the basic differencesamong these standards. Such an understanding will benefit anyonewho is in-

volved in the evaluation of circuit breakers designed and tested according tothese different standards.

The two most widely and commonly recognized standards documents todare issued by the American National Standards Institute and the Inter-national Electrotechnical Comm ission (IEC). The standards set fort by thesetwo organizations will be examined and their differences will be explained.By realizing thathe principles upon which they are basedre mainly localizedoperating practices, it is hoped that the meaningof each of the required capa-bilities will be thoroughlyunderstood.Thisunderstandingwillgivemoreflexibility to the application engineer for choosing the proper equipment forany specific application and tohe design engineer or selecting he appropriate

parameters upon which to base the design of a circuit breaker, which mustmeet the requirements of all of the most significant applicable standardsf it isto be considered of world-class.

type of book is long overdue. For those ofs who are involvedin thedesign of these devices, it has been a long road f learning. Many times, nothaving a concise source f readily available collection f design tips and gen-

design information, we have hado learn the subtleties of these designs byexperience. For those whose concern is the application and selection of the

devices, there is a need for some guidance that is independent of commercialinterests. There have been a numberof publications on he subject, but most,if not all of them, use a textbook approach to treat the subject with a strictmathematical derivationof formulae. The material presented heres limited to



what is believed to be the bare essentials, the fundamentals of the fundamen-tals, and the basic answers tohe most common questions n the subject.

The book Circuit Interruption: Theory and Techniques,edited by ThomasBrowne, Jr. (Marcel Dekker, 1984) partially meetshe aims of the new publi-cation, but earlier works have become obsolete since somef the new designconcepts of interrupter designs and revisions to he governing standards werenot thoroughly covered. None of these previous publications covers specific

design details, application, interpretationf standards, and equipment selectionand specification.

There are agreat number of practicing electrical engineersn the electricalindustry, whether in manufacturing, industrial plants, construction, or publicutilities, who will welcome book as an invaluable tool tobe used in theirday-to-day activities.

The most important contributorso book are those pioneer researcherswho laid the foundations for the development of the circuit breaker technologyIm specially indebted to Lome M cConnell, from whomearn the trade andwho encouraged me n my early years. I am also indebted o all those who ac-tively helped me with their timely comments, and especiallyo the Square “ DCo. for its support on project. Most of all, Im especially grateful to mywife, Maggi, or her support and patience duringhe preparation of book.

Ruben D. Garzon



Preface1. The Fundamentalsof Electric Arcs

1O Introduction1.1 Basic TheoryofElectricalDischarges

1.1.1 Non-Self-sustaining Discharges1.1.2 Self-sustaining Discharges

1.2.1 High Pressure Arcs1.2.2 Low Pressure (Vacuum) Arcs

1.3 The Alternating CurrentArc1.4 The Current Interruption Process

1.2 The Electric Arc

1.4.1 Interruption of Direct Current1.4.2 Interruption of Alternating Current

1.5 Review of Main Theories of ac Interruption1.5.1 Slepian’s Theory

1.5.2 Prince’s Theory1.5.3 Cassie’s Theory1.5.4 Mayr’s Theory1.5.5 Browne’s Combined Theory1.5.6 Modem Theories

References

2. Short Circuit Currents

2.0 Introduction2.1 Characteristics of he Short Circuit Current

2.1.1 Transient Direct Current Component2.1.2 The Volt-Time Area Concept2.1.3 Transient Alternating Current Components2.1.4 of Three Phase Short Circuit Currents

2.1.5 Measuring Asymmetrical Currents

2.2.1 The per Unit Method2.2 Calculation of Short Circuit Currents

V

1 2

2

4

5

5

8

10

11 11 14

19

20

21

22

23

24

25

26

29

29

29

30

32

35

36

39 42

43

ix



2.2.2 The MVA Method

2.3.1 Introduction to Components

2.4.1 Directionof the Forces Between Current Carrying

2.4.2 CalculationofElectrodynamic Forces Between Conductors

2.4.3 Forces on Conductors Produced by Three Phase Currents

2.3 Unbalanced Faults

2.4 Forces Produced byhort Circuit Currents

Conductors

References

3. Transient Recovery Voltage

3.0 ntroduction3.1 Transient Recovery Voltage: General Considerations

3.1.1 Basic Assumptions or TRV Calculations

3.1.2 Current Injection Technique3.1.3 Traveling Waves andhe Lattice Diagram3.2 Calculationof Transient Recovery Voltages

3.2.1 Single Frequency Recovery Voltage3.2.2 General Caseof Double Frequency Recovery Voltage3.2.3 Particular Caseof Double Frequency Recovery Voltage3.2.4 Short Line Fault Recovery Voltage3.2.5 Initial Transient Recovery Voltagel”RV)

References

4. Switching Overvoltages

4.0 Introduction4.1 Contacts Closing

4.1.1 Closingof a Line4.1.2 Reclosingof a Line

4.2.1 Interruptiono f Sma l l Capacitive Currents4.2.2 Interruption of Inductive Load Currents4.2.3 Current Chopping4.2.4 Virtual Current Chopping4.2.5 Controlling Overvoltages

4.2 Contact Opening

References

5. Types of Circuit Breakers

5.0 Introduction5.1 Circuit Breaker Classifications

51 53 54 57

58 65

72 75

77

77

78 80

80 81 85 85 90 99

104 107

109

111

111

112 116 116

117 118

120 124 126 126

129

129 130



5.1.1 Circuit Breakers Typesy Voltage Class5.1.2 Circuit Breaker Types by Installation5.1.3 Circuit Breaker Typesy External Design5.1.4 Circuit Breaker Types by Interrupting Mediums

5.2.1 Arc Chute Type Circuit Breakers5.2.2 Air Magnetic Circuit Breakers: Typical Applications

5.3.1 Blast Direction and Nozzle Types5.3.2 Series Connectionof nterrupters5.3.3 Basic Interrupter Arrangements5.3.4 Parameters Influencing Air Blast Circuit Breaker

5.2 Air Magnetic Circuit Breakers

5.3 Air Blast Circuit Breakers

Performance5.4 Oil Circuit Breakers

5.4.1 Propertiesof nsulating Oil

5.4.2 Current Interruption in Oil5.4.3 TypesofOil Circuit Breakers5.4.4 Bulk Oil Circuit Breakers5.4.5 MinimumOil Circuit Breakers

5.5.1 Propertiesof SF6

5.5.2 Arc Decomposed By-products

5.5.4 Current Interruption n SF65.5.5 T w o ressure SF6 Circuit Breakers5.5.6 Single Pressure F6 Circuit Breakers5.5.7 Pressure Increase f SF6 Produced y anElectric Arc5.5.8 Parameters Influencing SF6 Circuit Breakererfomance5.5.9 SF6-NitrogenGasMixture

5.6.1 Current InterruptioninVacuum Circuit Breakers

5.6.2 Vacuum Interrupter Construction5.6.3 Vacuum Interrupter Contact Materials5.6.4 Interrupting CapabilityofVacuum Interrupters

5.5 Sulfurhexafluoride

5.5.3 SF6 Environmental Considerations

5.6 Vacuum CircuitBreakers

References

6. MechanicalDesign of CircuitBreakers

6.0 Inmduction

6.1 ContactTheory6.1.1 Contact Resistance6.1.2 Insulating Film Coatings n Contacts

130 130 130 132 134 135 141 141 142 145 146

147 149 150

150 152 156 160 161 162 164 165

167 169 171 175 178 181 182 182

188 189 192 192

195

195

195 196 198



Contact Fretting' Temperature at the Point ofContact

Short Time HeatingfCopperElectromagnetic Forces n ContactsContact Erosion

Mechanical Operating CharacteristicsCircuit Breaker Opening Requirements

Closing Speed Requirements

Cam versus LinkageWeld Break and Contact BounceSpring MechanismsPneumatic MechanismsHydraulic Mechanisms

Operating Mechanisms

References

7. A Comparison of High Voltage Circuit Breaker Standards

IntroductionRecognized Standards Organizations

ANSVIEEENEMAIEC

Circuit Breaker Ratings

Normal Operating ConditionsRated Power Frequency

Maximum Operating VoltageRated Range FactorRated Dielectric StrengthRated Transient Recovery Voltage

Rated Continuous CurrentRated Short Circuit CurrentAsymmetrical CurrentsClose and Latch, r Peak Closing CurrentShort Time CurrentRated Operating Duty CycleService Capability

Additional Switching Duties

Capacitance SwitchingLine ChargingCable Charging

Voltage Related Ratings

Current Related Ratings

225

1



7.4.4 Reignitions, Restrikes, and Overvoltages

References

8. Short Circuit Testing

8.0 Introduction8.1 Test Methodology

8.1.1 Direct Tests8.1.2 Indirect Tests8.1.3 Single Phase Tests8.1.4 Unit Tests8.1.5 TwoPart Tests8. l 6 Synthetic Tests

8.2.1 Measured Parameters and Test Set-Up

8.2.2 Test Sequences

8.2 Test Measurements and Procedures

References

9. Practical CircuitBreakerApplications

9.0 Introduction9.1 Overload Currents and Temperature Rise

9.1.1 Effects ofSolarRadiation9.1.2 Continuous Overload Capability9.1.3 Short Time Overloads9.1.4 Maximum Continuous Current t High Altitude

Applications9.2 Interruptionof from HighX / R Circuits9.3 Applications t Higher andLower Frequencies9.4 Capacitance Switching Applications

9.4.1 Isolated Cable9.4.2 Back o Back Cables

9.4.3 Isolated Shunt Capacitor Bank9.4.4 Back to Back Capacitor Banks9.4.5 General Application Guidelines

9.5.1 Ferroresonance9.5 Reactor Current Switching, HighRV Applications

9.6 High Altitude Dielectric Considerations9.7 Reclosing Duty Derating Factors9.8 Choosing Between Vacuumr SF6

References

254 254

257

257 259

263 264 264 267267268 274 274

277 290

293

293 293 294 295

297

299 299 304 305 306 306

307 308 3 10311 3 133 143 163 18

3 19



10. Synchronous Switchingand Condition Monitoring

IntroductionSynchronous Switching

Mechanical Performance ConsiderationsContactGap Voltage WithstandSynchronous Capacitance Switching

Synchronous Reactor SwitchingSynchronous Transformer SwitchingSynchronous Short Circuit Current Switching

Condition MonitoringfCircuit BreakersChoice ofMonitored ParametersMonitored Signals Management

References

Appendix: Conversion Tables

321

351

Index 357



1

THE FUNDAMENTALS OF ELECTRIC ARCS

1O IntroductionFrom the time when the existence of the electric current flow wasirst estab-lished and even before the basic thermal, mechanical, and chemical effectsproduced by such current were determined,t had became clear hat there wasa need for inventing a device capablef initiating and of stopping the flow ofthe current.

Fundamentally, there are two ways in which the flow of current bestopped, one s to reduce he driving potentialo and the other is to physi-cally separateor create an open gap betweenhe conductor that is carrying thecurrent. Historically, t has been the later method the one most commonly usedto achieve current interruption.

Oersted, Ampere and Faraday are among the first known users of circuitbreakers and according to recorded history those early circuit breakers areknown to have been a mercury switch which simply consistedf a set of con-ducting rods hat were immersed n a pool of mercury.

Later in the evolutionary history of the current switching technology, themercury switch, or circuit breaker was replaced by the knife switch design,which even now is still widely used for some basic low voltage, low powerapplications. Today, under the present state of the art in the current interrup-tion technology, the interruption process beginst the very instant when a airof electric contacts separate. It continues as the contacts recede from eachother and as the newly created gap is bridged by a plasma. The interruptionprocess is then finally completed whenhe conducting plasma s deprived of its

conductivity.By recognizing hat the conducting plasma s nothing more han the core of

an electric arc, t becomes quite evident hat inherently the electric arc consti-tutes a basic, indispensable, and active elementn the process of current inter-ruption. Based on simple knowledge, it follows that the process of extin-guishing the electric arc constitutes the foundation upon which current inter-ruption is predicated. It is rather obvious then that a reasonable knowledgeofthe fundamentals of arc theory is essential to the proper understanding of theinterrupting process. It s intended that the following basic review describingthe phenomena of electric discharges, will serve to establishhe foundation forthe work that s to be presented later dealing with current interruption.

1



1.1 Basic Theory of Electrical Discharges.

The principles that govern the conduction of electricity through, either a gas,or a metal vapor,are based on the fact that such vapors always contain positand negative charge and that all types of discharges always involvethe very fundamental processesf production, movement, and final absorptionof the charge carriers as the means of conveying the electric current between

the electrodes.For the sake of convenience and in order to facilitate the reviewf the gas

discharge phenomena, the subject will e divided into the following three verybroad categories:

a) The non-self-sustaining dischargeb) The self-sustaining discharge andc) The electric arc

Non-Self-Sustaining DischargesWhen a voltage is applied across two electrodes, the charge are actedupon by a force that is proportional to the electric field strength, force es-tablishes a motion f the ions towards the cathode andf the electrons towardsthe anode. When the moving charges strike the electrodes they give up theircharges; thus producing an electric current through the gaseous medium. Acontinuous flow of current can take place only if the carriers whose chargesare absorbed by the electrodes are continuously replaced. The replacement ofthe charge carriers canbe made by a number of ionization processes suchs,photoelectric, or thermionic emissions.

Initially, the discharge current s very small; however as the voltage is in-

creased it is observed that the current increasesn direct proportion o the volt-age applied across the electrodes, until a level is reached where the chargecaniers are taken by the electrodest the same rate as they are produced. Once

equilibrium state is attained the current reaches afirst recognizable stablelimit that is idenW1ed as the saturation current limit. The value of thetion current is dependent upon the intensity ofhe ionization, it is also propor-tional to the volume f gas filling the space between the electrodes and to thegas pressure.

At the saturationlimit the current remains constant despite increasesf thesupply voltage to levels that are several times the level originally required toreach the saturation current imit. Because the saturation currents entirely de-pendent on the presence of charge carriers that are supplied by external ioniz-

ing agents type of discharge is called a non-self-sustaining discharge.Since the charge carriers are acted upon not only by the force exerted bythe electric field, but y electro-static forces that are dueo the opposite polar-ity of the electrodes, the originallyuniform distribution of the charge



be altered by the application of a voltage across he electrodes. It can beobserved that an increase in the electrode’s potential produces an increasedconcentration of electrons near the anode and of positive molecular ions nearthe cathode; thus creating what is known as space charges at the electrodeboundaries.

The space charges leado an increase in the electric fieldat the electrodes,which will resultn a decrease ofhe field in the space between the electrodes.

The drops of potential at the electrodes are known as the anode fall of poten-tial, or anode drop, and s the cathode fall of potential, or cathode drop.

It was mentioned previously, that whenever the current reaches its satura-tion value he voltage applied acrosshe electrodes and hencehe electric field,may be substantially increased without causing any noticeable increasen thedischarge current. However s the electric field strength increases, does thevelocity of the charge and since an increase in velocity represents anincrease in kinetic energy it is logical to expect that when these acceleratedcharges collide with neutral particles new electrons wille expelled from theseparticles; thus, creatinghe conditionknown as shock ionization.

In the event that the kinetic energy s not sufficient or N l y ionizing a par-ticle it is possible that t will be enough to re-arrange the original grouping ofthe electrons by moving them from their normal orbits to orbits situated at agreater distance from the atom nucleus. This state is described as the excitedcondition of the atoms; once conditions is reached a smaller amountenergy willbe required in order to expel the shifted electron from excitedatom and o produce complete ionization. Its apparent then, hat with a lesserenergy level of the ionizing agent, successive impactsan initiate the processof shock ionization.

The current in the region of the non-self-sustaining discharge ceases as

soon as the external source is removed; However when the voltage reaches acertain critical evel the current increases very rapidly and a spark resultsn theestablishing of a self-sustained dischargen the form of either a glow dischargeor the electric arc.

In many cases, such as for example between parallel plane electrodes, thetransition froma non-self sustaining to a self-sustaining discharge leads o animm ediate complete puncture or flashover which, provided that the voltagesource is sufficiently high, will result in a continuously burningarc being es-tablished. In the event that a capacitor is discharged across the electrodes, theresulting discharge takeshe form of a momentary spark.

In other cases, where the electric field strength decreases rapidly as thedistance between the electrodes increases, the discharge takes the form of a

partial flashover. In case the dielectric strength the gas space is ex-ceeded only nearhe electrodes andas a result a luminous dischargenown as

appears around he electrodes.



300

0. a

1K

Current (Amperes)

FigureSchematic representation of he voltage-current relationship of a self-sustaining electrical discharge.

1.1.2 Self Sustaining Discharges

The transition from a non-self-sustaining discharge to a self-sustained dis-charge is characterized by an increase in the current passing though the gas,

whereas the voltage across the electrodes remains lmost constant. When theelectrode potential s increased to the point where ionization occurs freely, thepositive ions produced in the gas may strike the cathode with a force that is

sufficient to eject the num ber of electrons necessary for maintaining the dis-charge. Under these circumstances no external means f excitation are neededand the discharges said tobe self-sustaining.

During theinitial stages of the self-sustaining discharge he current densityis only in the order of a few micro-amperes per square centimeter, the dis-

charge has not yet become luminous and consequently it is called a dark dis-charge; However as the current continues to increase a luminous glow appearsacross the gas region between the electrodes, and,s illustrated in figure 1.1,

the stage known as the "glow discharge" takes place and the luminous glow



then becomes visible. The colorsf the glow differ between he various glow-ing regions and varyn accordance with he surrounding gas.In air, for exam-ple, the negative or cathode glow exhibits a very light bluish color and thepositive columns salmon pink. The glow discharge characteristic have proveto be very important or applications dealing with illumination.

The regioncalled the normal glow regions that in which the current is lowand the cathode is not completely covered by the cathode glow, the cathode

current densityat time is constant and s independent of the dischargerent. When the current is increased so that the cathode is completely coveredby the negative glow,he current density has increased ando has the cathodevoltage drop and region is called the abnormal glow region.

As the current increases in the abnormal glow region the cathode dropspace decreases in thickness. This leads to a condition where the energy im-

parted to the positive ions is increased and the num ber of ionizing collisionsencountered by an ion in the cathode drop space is decreased. The increased

energy of the incoming positive ions increasehe cathode temperature whichnturn eads to a conditionof thermionic emission, that subsequently resultsn anincrease in current that is accompanied by a rapid collapse in the dischargevoltage. During transitional period, he physical characteristicsof the dis-charge change from those of a glow discharge to those of a fully developedarc.

1.2 The Electric Arc

The electricarc is a self-sustained electrical dischargehat exhiiits a low volt-age drop, that is capable of sustaining large currents, andhat it behaves like anon-linear resistor.Though he most commonly observedarc discharge occursa m s s air at atmospheric conditions, he arc discharge is also observed at highand low pressures, in a vacuum environment, and in a variety of gases andmetal vapors. The gases and vapors, that serves conductors for the arc, origi-

. nate, partly from the electrodes, and partly from the surrounding environmentand from their reaction products. The description of the electric arc will bearbitrarily divided into two separately identitiable typesof arcs. This is doneo d y as an attempt to provide with a simpler way to rela te to future subjectsdealing with specific interrupting technologies. Theirst arc type willbe iden-tified as the high pressure arc and the second type, which is an electric arcburning in a vacuum environment, wille identified as a low pressure arc.

HighPressureArcs

High pressure arcs are considered to be those arcs that exist at, or above at-

mospheric pressures. The high pressurerc appears as a bright column charac-terized by having a small highly visible, brightly burning core consisting ofionized gases that conveyhe electric current. The core f the arc always exist



at a very high temperature and therefore the gasesre largely dissociated. Thetemperature of he arc core under conditions f natural air cooling reach tem-peratures of about K while when subjected to forced cooling, tempera-tures in excess of K have been observed. The higher temperatureshathave been observed when the arcs being cooled at firsts appears to be a con-tradiction. One wouldthink hat under forced cooling conditionshe tempera-ture should be lower, however the higher temperature is the result of a re-

duction in the arc diameter which producesan increase in the current densityof the plasma and consequently leads tohe observed temperature increase.

By comparing the cathode regionof the arc with the cathode region of theglow discharge, it is seen that the cathode of the glow discharge has a fall ofpotential in the range of 100 to 400 volts, it has a low current density, thethermal effects do not contribute to the characteristics of the cathode and thelight emitted from he region near he cathode has the spectrum of the gas sur-

rounding the discharge. In contrast the cathode f the arc has a fall of potentialof only about 10volts, a very high current density, and the light thats emittedby the arc has the spectrum of the vapor of the cathode material.

Some of the most notable arc characteristics, that have a favorable influ-ence during the interrupting process, arehe fact that the arc can be easily in-fluenced and diverted by he action of a magnetic field or by the action of ahigh pressure fluid flow and that he arc behaves as a non-linear ohmic resis-tance. If the arc behaves like a resistors it follows then that the energy ab-

sorbed in the arc is equal to the productf the arc voltage drop andhe currentflowing through he arc.Under constant current conditionshe steady state arc is in hermal equilib-

rium, which means that he power losses from he arc column are balanced bythe power input into the arc. However dueo the energy storage capabilityofthe arc there s a time ag between the instantaneous power loss andhe steadystate losses and therefore t any given instant the power inputo the arc, plusthe power stored in the arc is equal to the power loss from the arc. This time

lag condition, as it will be seen later in chapter, is extremely significantduring the time of interruption, near current zero.As a result of the local thermal equilibrium it is possible to treat the con-

ducting column of he arc as a hot gas which satisfieshe equations of conser-vation of mass, momentum and energy. To which, ll the thermodynamic lawsandMaxwell'selectromagneticequationsapply.This mplies that the gascomposition, its thermal, and its electrical conductivity are factors which areessentially temperature determined.

The voltage drop acrossn arc canbe divided into three distinct regions,sillustrated in figure 1.2. For short arcs the voltage drop appearing in a rela-tively region immediately n front of the cathode, represents a ratherargepercentage of the total arc voltage. voltage drop across the region nearhe



ANODE CATHODE

ARCLENGTH -Figure 1.2 Voltage distribution of an arc column, V, represents the anode voltage;V,, represents the cathode voltage, nd V,, represents the positive column voltage.

cathode is typically between 10 to 25 volts and is primarily a function of thecathode material. In the opposite electrode, the anode drop, is generally be-tween 5 to volts. The voltage dropacross the positive columnof the arc ischaracterized by a niform longitudinal voltage gradient, whose magnitude,nthe case of an arc m u n d e d by a gaseous environment, depends primarily n

the type of gas, the gas pressure, the magnitude of the arc current, and thelength of the column itself. Voltage valuesor the positive column gradient n

the range of only a few volts per centimeter to hundred volts per cm -timeter have been observed.

The firstextensive study of the electric rc voltage relationships, for mod-erate levels of current and voltage, was made by Hertha Ayrton [l],who de-veloped a formula defining the arc voltage on the basis of empirical experi-mental results. The relationship still is considered to be valid, and is stillwidely used, although within a limited rangef current and voltage.

The classicalAyrton equation is given as:

C + D deo = A + B d + -

i



where:e, = arc voltage

i = arc currentA=19, B=11.4, C=21.4 a n d B 3

d = arc length

The valuesof these constantsA, B, C, and D, are empirical values,for cop-

per electrodes in air.The current density t the cathode is practically independent ofhe arcrent, but it is strongly dependent upon the electrode material. In refractorymaterials such as carbon, tungsten, or molybdenum, that have a high boilingpoint, the cathode spot is observed to be relatively fixed, the cathoqe operatesby $hem ionic emission and its current density is in the order of 10 amps percm . The "cold cathode arc" is characteristic of low boiling point materialssuch as copper and mercury. The cathode spot in these materials is highly

mobile, it operpes some form04 field emission and ts current density is inthe order of 10 to 10 amps per . n those materialsthat have a low boilingpoint a considerable amount of material is melted away from the electrodes;while the material losses of refractory materials is only due to vaporization.Under identical arcing conditionshe refractory material losses are considem-bly less than the losses of low boiling point materials, and consequentlyconstitutes an important factor that mustbe kept on mind when selecting ma-terials for circuit breaker contacts.

1.2.2 Low Pressure (Vacuum) Arcs

The low pressureor vacuum arc, ike those arcs hat occur at,or above atmos-pheric pressure, share most f the same basic characteristics ust described forthe electric arc, but the most significant differences are: (a) An average arcvoltage of only about 0 volt which s significantly lower the arc voltagesobserved in high pressure arcs.@) The positive column of the vacuum arc issolely influenced by the electrode material because the positive column is

composed of metal vapors hat have been boiledff from the electrodes; whilein the high pressure arche positive column is made up of ionized gasesromthe arc's surrounding ambient. (c) Finally, and, perhaps the most significantand fundamental difference whichs the unique characteristic f a vacuum arcthat allows the arc to exist n either a diffuse mode orn a coalescent or con-stricted mode.

The diffuse mode is characterized by a multitude of fast moving cathodespots and by what looksike a multiple numberf arcs in parallel. It should e

pointed out hat is the only time when arcsn parallel exist without heneed of balancing, or establishing inductances. The magnitude of the currentbeing carriedby each of the cathode spots s a functionof the contact material,and in most casess only approximately amperes. Higher current densities



I I

a=DifhsePn=

I -Figure 1.3 Outline of an arc in vacuum illustrating the characteristics of the arc in a

diffuseand ina constricted mode.

are observed on refractory materials such s tungsten or graphite, while lowercorrespond to materials that have a low boiling point suchs copper,

When the current is increased beyond a certain limit, that depends on the

contact material, one ofhe roots of the arc gets concentrated into a single spotat the anode while the cathode spots split to form a closely knit group ofhighly mobile spots as shown in figure If the cathode spots are not influ-enced by external magnetic fields, they move randomly aroundhe entire con-tact surface at very high speeds.

When the current is increased even further, a single spot appears at theelectrodes. The emergenceof a single anode spot s attributed to the fact thatlarge currents greatly increase he collision energy of he electrons and conse-

quently when they collide with he anode, metal atoms are released thus pro-ducing a gross melting condition of the anode. There is a current threshold atwhich the transition from a diffuse arco a constricted arc mode takes place,t h i s threshold level is primarily dependent upon the electrode size and theelectrode material. With today’s typical, commercially available vacuum interupters, diffused arcs generally occurat m e n t values below 15 kilo-amperesand therefore n some ac circuit breaker applicationst is possible for the arc tochange from a iffuse mode to a constricted mode s the current approaches ts

peak and then backo a diffuse mode s the current approaches ts natural zerocrossing. It follows then, that he longer the time prior to currentzero that aninterrupter is in the diffuse mode, he greater it is its intemqting capability.



1.3 The Alternating Current Arc

The choice of sinusoidal alternating currents as the standard for the powersystems is indeed a convenient and fortuitous onen more ways than one. As itwas described earlier, in the case of an stable arc, as the arc current increasesthe arc resistance decreases due to the increasen temperature which enhancesthe ionizing process, and when the current decreases the ionization level also

decreases while the rc resistance increases; thus, there is a collapsef the arcshortly before the alternating current reaches ts normal zero value at the endof each half cycle. The arc will reignite again when the current flows in theopposite direction, during the subsequent half cycle, provided that the condi-tions across the electrodes re still propitious for the existence of the arc. Thetransition time between the two half cycles is greatly influenced by the me-dium on which the arc is being produced and by the characteristics of the ex-ternal circuit.

The arc current as it approaches zero is slightly distorted from a true sinewave form due to the influence of the arc voltage, and thereforehe arc is ex-tinguished just prior to the nominal current crossing. The current zerotransition is accompanied by a sharp increase in the arc voltage, and the peakof voltage is defined as the peak of extinction voltage. When he peak ofthe extinction voltage reaches a value equal to the instantaneous value f thevoltage applied to the rc by the circuit, the arc current can notbe maintainedany longer and thereafter, the current in the opposite direction cannot be re-

established immediately; thus,t every current zero theres a f in te time periodwhere there not be any current flow. This is the time period generally re-ferred as the “current zero .pause”. During the zero current period, the dis-

charge path i s partially de-ionized on accountof the heat losses and thereforethe electric field needed to re-establishhe arc after the reversalf the currentbecomes greater than the fieldequired to maintain the arc. This means that threquired reignition voltage is higher than the voltage which is necessary tosustain the arc, and therefore the current will remaint its zero valueuntil the

reignition voltage level is reached. If the arc is re-established the current in-creases and the voltage falls reachingts minimum value, which is practicallyconstant during most of the half cycle, andn the region of maximum current.

The sequence that have just been described will continue to repeat itselfduring each oneof the subsequent half cycles, provided that the electrodesresymmetrical; however n most cases there willbe some deviations n the arc’sbehavior which arise from differencesn the electrode materials, cooling prop-erties, gas am bient, etc. This asymmetric condition is specially accentuatedwhen the electrodes re each made of different materials.

The time window that follows a current zero and during which arc re-ignition can occur depends upon the speed at which the driving voltage in-



creases at the initiation of each half cycle, and on the rate at which de-ionization takes place in the gap space; in other words, the reignition processrepresents the relationship between the rate f recovery of the supply voltageand the rate of de-ionization, or dielectric recovery of the space a m s s theelectrode gap.

1.4 The Current Intemption Process

In the preceding paragraphs he electric arcs were assumed to e either static,as in the case of direct current arcs, oruasi static, as in the case of alternatingcurrent arcs. What this means is that we had assumed that the arcs were asustained discharge, -and that they were burning continuously. However, whatwe are interested in, is not with the continuously burning arc; but rather withthose electric arcs which are in the processof being extinguished because, as

we have learned, current interruption is synonymous with arc extinction, andsince we further know that the interrupting processs influenced by the charac-teristics of the system and byhe arc’s capacityfor heat storage we can expectthat the actual interruption process, from the timef the initial creationof thearc until ts extinction, will depend primarily on whether the current changesnthe circuit are forced by the arc discharge, or whether those changes are con-trolled by he properties of the power supply. Theirst alternative is what it

be observed during direct current interruption where the current to be inter-rupted must be forced to zero. The second case s that of the alternatingrent interruption where aurrent zero occws naturally twice during each cycle.

1.4.1 Interruption of Direct Current

Although the subject relating to direct current interruption does not enter intothe later topics of book, a brief discussion explaining the basics of directcurrent interruption s given for reference and general information purposes.

The interruption of direct current sources differ in several respects fromthe phenomena involving the interruption of alternating currents. The mostsignificant difference s the obvious fact hat in direct current circuits there re

no natural current zeroes and consequently a current zero must be forced insome fashion n order to achieve a successful current interruption. The forcingof a current zero is done either by increasinghe arc voltage to a level thatsequal, or higher than the system voltage, ory injecting into the circuit a volt-age that has an opposite polarity to thatof the arc voltage, which n reality isthe equivalent of forcing a reverse current flowing into the source.

Generally the methods usedor increasing thearc voltage consistof simplyelongating thearc column, of constricting the arc by increasing the pressure o

the arc’s surroundingsso as to decfease the arc diameter and to increase thercvoltage, orof introducing a numberf metallic plates, alonghe of the arc,in such a way o that a seriesf short arcsare developed.



2

Figure 1.4 Relationships arc and system voltages during the interruptionof directcurrent.

With latter approachit is to see that s a minimum the of thecathode and anode drops, which amounts tot least 30 volts, be expectedfor each arc. Since thearcs are in series their voltage is additive, and thefinalvalue of the arc voltage s simply a function of the numberf pairs of in-

terposing plates. The second method, thats the driving of a reversed current,is usually doneby discharging a capacitor m s s the arc. The former methodscommonly used on low oltage applications, while the later methods used forhigh voltage systems.

For a better understanding of the role played byhe arc voltage during theinterruption of a direct current,et us consider a direct current circuit that hasvoltage E, a resistanceR , and an electric arc,all in series with each other. Thecurrent in the circuit, and therefore the current in the arc will adjust itself in

accordance with the values of the source voltage E, the series resistance R,and the arc characteristics. By efening to figure 1.4, where the characteristicsof the arc voltage are shown as a function of the current for two different arc

lengths, it is seen that he arc voltage e, smaller than the supply voltagebyan amount to iR, so that e, = E - iR. 'If the straight line that representsvoltage is plotted as shown in the figure, it is seen that line intercepts

with the curve representinghe arc characteristics for length 1 at the points in-



dicated as 1and 2. Only at the intersection of these points, s dictated by theirrespective currents, it ispossible to havean stable arc.If for example therent Corresponding to point increases it can be seen that he arc voltage is toolow, and if the current decreaseshe corresponding arc voltages too high, andtherefore the current always will to revert to its stable point. In order toobtain a stable conditiont point l it will be necessary for the circuit to havevery high series resistance and a high supply voltage. We already know that

the net result of lengthening the arc is an increase in the arc resistance and areduction in current, provided that the supply voltage remains constant. How-ever, and as t isgenerally the case in practical applications, some resistancesalways included in the circuit, and the reductionof the current will produce acorresponding reduction in the voltage across the series resistance. The elec-trode voltage is thus increased until eventually,hen the arc is extinguished, itbecomes equal to the system voltage. These conditions re shown graphicallyin figure 1.4 for the limiting case where therc characteristics of the longerrc

represented by the curve labeled length have reached a position where it nolonger intercepts the circuit characteristics that are representedy the line E -iR , and therefore the condition where the arc can not longer exist has beenreached.

When inductance s added to he circuit that we haveust described we findthat the fundamental equationfor th is inductive circuit we represented as

follows:

L”= ( E - i R ) - e , = Aei

dt

This equation indicates that the inductive voltage produced during inter-ruption is equal to the source voltage reduced by the voltage drop acrossheinherent resistance of the circuit, and by the arc voltage. For the arc to be ex-tinguished the currenti must continually decrease which n suggests thatthe derivativeof the current (dddt) m uste negative.

As it is seen from the equation, the arcing conditionsat the time of inter-ruption are significantly changed, in relation to the magnitude of the induc-tance L. Since the inductance opposes the current changes, the falling of thecurrent results in n induced e.m.f. which acts additively to the source voltage.The relationship between the source voltage and the arc voltage still holdsorthe inductive circuit, therefore it is necessary to develop higher arc voltageswhich require that the rupturing arc length be increased to provide the addi-

tional voltage.It is also important to remember that when interrupting a direct currentir-

cuit, the interrupting device must be able to dissipate the total energy that is

stored in the circuit inductance.



1.4.2 Interruption of Alternating Current

As it has been hown in theprevious ection, in orderocurrent arc it is required to create, orin some way, to force a current zero.han alternating current circuit the instantaneous value of the current passesthrough zero twice during each cycle and therefore the zero current conditionis already self-Milled; consequently, to interrupt an alternating current it isonly Mlcient to prevent the reignitionof the arc after the currenthas passedthrough zero. It s for reason that de-ionizationof the arc gap close to thetime of a natural current zero is of the utmost importance. While any reductof the ionization of the arc gap close to the pointf a current peak it is some-what beneficial, action does not significantly aid n the interrupting proc-ess. However, because of thermodynamic constrains that existn some typesofinterrupting devices, it is advisable that all appropriate measures to enhanceinterruption be taken well n advance of the next natural current zerot whichinterruption is expected to take place. Successful current interruption dependson whether the dielectric withstand capabilityf the arc gap s greater than theincreasing voltage that is being impressed across the gap by the circuit in anattempt re-establish the flowf current. The dielectric strengthf the arc gap isprimarily a function of the interrupting device; while the voltage appearingacross the gaps a functionof the circuit constants.

At very low frequencies, in the range of a few cycles but well below thevalue commonly used in general for power frequencies, the rate of change of

the current passing through zeros very small, and n spite of the heat capacityof the arc column, the temperature and the diameterf the arc have mflicienttime to adjust tohe instantaneous valuesof the current and therefore when thecurrent to a sufficiently small value, (less than a few amperes, dependiupon the contact gap), the alternating current arc will self extinguish, unlessthe voltageat the gap contacts t the timeof interruption is sufficiently high toproduce a glow discharge.

Normal power application frequencies, which generally in the range of

16 to 60 Hertz, re not M ~ ci en t lyow to ensure that the rc will go out onits own. Experience has shown that an alternating current arc, supported by a50Hz. system of 30 kilovolts, that is burning across apair of contacts in openairand up to1meter in length, can not be extinguished. Special measures needto be takenf the effective current exceeds about 10 amperes.his is due to hefact that at these frequencies when the current reachests peak value the elec-tric conductivity of the arc is relatively high and since he current zero periodis very short the conductivity of the arc, if the current is relatively large, can

not be reduced enough to prevent re-ignition. However, since the current oscil-lates between a positive and a negative value t h m is a ten-dency to extinguish thearc at the current zero crossing dueo the thermal ag



e, \

Figure 1.5 Typical variations of current and voltage showing the peak of extinctionvoltage e, and peak of e-ignition voltage q.

previously m entioned. The time lag between temperature and current is com-monly referred tos the “arc hysteresis”.

When the alternating current crosseshrough its zero, the arc voltage takes

a sudden jump to a value equal o the of the instantaneous peak valueofthe extinguishing voltagefrom he previous current loop, plus the peak valueof the reignition voltageof the next current loop, which s associated with thereversal o f the current. In the event that the arc is reignited, and immediatelyafter the reignition has taken place, therc voltage becomes relatively constantand ofa significantly lower magnitude, s illustrated in figure 1.5, In order fora reignition o occur, the applied voltage must exceedhe value of the total re-ignition voltage, (G). One practical application, derived from observing the

characteristicsof the extinction voltage s that during testing f an interruptingdevice, it provides a good indication of the behavior the device under test.Agood, large and sharp peakf voltage indicates thathe interrupter is perform-ing adequately, but if the peak of the voltage begins o shown a smooth roundtop and the voltage magnitude begins to drop, it is a good indication that heinterrupter is approaching its maximum interrupting limit. If reignition doesnot happen, he flow of the current will cease and therefore interruption willeaccomplished. From what has been described,t is rather obvious thathe most

favorable conditions for intem qtion are those n which the applied voltage isat its lowest when he current reaches he zero value, howeverhis ideal condi-tion be realized only with a purely resistive circuit.



1.4.2.1 Interruption of Resistive Circuits

In an alternating current circuit containing only resistance, r having a negli-gible amountof inductance, the current is practically in phase with he voltage,and during steady state operating conditions, both he current and the voltagereach their zero value simultaneously. But when aair of contacts separate andan arc is developed between the contacts, the phase relationship, in theory itstill exists, but in practice, as explained before the current will reach the zerovalue slightly aheadof the voltage. As the current(I)asses through zero theinstantaneous value of the peak of extinction voltage, shownas e. in figure

is equal to the instantaneous value of the applied voltage (E). From this

point on no new charges are produced in the gas space between the contactsand those charges still presentn the gas space are being neutralizedy the de-ionized processes thatare taking place. The gas space and he electrodes con-tinue to increasingly cool down and therefore the minimum voltage required

for the arc to re-ignite is increasing with time, he general ideaof this increaseis shown approximately in figure in the curve markedasIf he applied voltageE, shown as curve 2, rises at a higher ate the re-

ignition voltage , so that the corresponding curves intersect,he arc willbe re-established andwill continue to burn for an additional half cycle,t the end ofwhich the process will be repeated. It will e assumed that during time thegap length had increased and thereforehe arc voltage and he peak of the ex-tinction voltage be assumed to be larger. The increase in the gap length

will also provide an additional withstand capability andf the supply voltage sless than the re-ignition voltage then a successful interruptionf the current inthe circuit will e achieved.

1.4.2.2 Interruptionof Inductive Circuits

Generally, in an inductive circuithe resistance is rather small n relation to heinductance and therefore there is a large phase angle difference between thevoltage and the current. The current zero no longer occurs at the point where

the voltage is approaching zero but instead when it is close to its maximumvalue. This implies that the conditions favor he re-striking of the arc immedi-ately after the current reversal point.

It is important to note that in actual practice all inductive circuits have acertain small amountf self capacitance, such s that found between andcoils in transformersand in the self-effective capacitanceof the device itself nrelation to ground. Although the effective capacitance, under normal condi-tions, can be assumed to be very small, it plays an important role during the

interrupting process. The capacitanceo ground appears as a parallel elementto the arc,and herefore at the instant of currentzero the capacitance ischarged to a voltage equal to he maximum value of the supply voltage, plusthe value of the peak of the extinction voltage.



Figure 1.6 Interruption of a purely resistive circuit showing the current, voltages andrecovery characteristics, for the electrode space (curve 1) and for the system voltage(curve 2).

When the arc isextinguished, the eleclro-magnetic energy stored n the induc-tance is converted into electrostatic energy in the capacitance and vice versa.The natural oscillations produced byhe circuit are dam ped graduallyby theeffects of any resistance thatmay be present in the circuit, and since the oscil-latory frequencyof the inductance and the capacitances much greater thanhefrequency of the source the supply voltage may e regarded as being constantduring the time duration of the oscillatory response. These voltage conditions

are represented in figure 1.7. During the interruption of inductive alternatingcircuits, the recovery voltage be expected to reach its maximum value atthe same time at which the current intermpted, since he circuit broken as

the currentapproaches Howeverdue to the nherentcapacitance toground the recovery voltage does not reachts peak at the sam e instant herent is intermpted and therefore, during brief period, a transient responsesobserved in the circuit.

1.4.2.3 Interruption o Capacitive Circuits

The behavior of a purely capacitive circuit during he interruption process sillustrated in figure 1 8. It should be noted that, in contrast with the high de-



Figure 1.7 Current and voltage characteristics during intenuption of an inductive cir-cuit.

gree of difficulty that is encountered during the interruption of an inductivecircuit,when nterruptingacapacitivecircuit, he ystemconditions aredefmitely quite favorable for effective interruption at the instant of currentzero, because the supply voltage that appears across the electrodess increasedat a very slow rate. At the normal current zero where he arc interruption hastaken place the capacitors charged to, approximately, the maximum value ofthe system voltage. The small difference that may be observed is due to thearc voltage, however the magnitude of therc voltage is small in comparisonto the supply voltage andt generally it safely be neglected.

At intenuption, and in the absence of the current, the capacitorwill retainits charge, and the voltage m s s the gap will be to the algebraic ofthe applied voltage and the voltage trappedn the capacitor. The total voltageincreases slowlyfrom an initial value to zero, until one halfof a cy-

cle later the voltage a m s s the gap reaches twice the magnitude of the sup-ply voltage; however there is a relative long recovery period that may enablethe gap to recover its dielectric without reignithg. Under certain cir-

cumstances there may be restrikes that could lead to a voltage escalationcondition. This particular condition will e discussed later in the section deal-ing withhighvoltage transients.



I _.".........."""._.___.

Figure 1.8 Voltage and current characteristicsduring the intenuption of a capacitivecircuit

1.5 Review of Main Theories f ac Interruption

The physical complexityn the behavior ofan electric arc during the interrupt-ing process, have always provided the incentive for researchers to developsuitable models that may describe process. Over the years a variety oftheories have been advanced by many researchers.n the early treatments ofthe interruption theory the investigative efforts were concentratedat therent zero region, which most obviouslys the region when the alternatingrent arc either reignites r is extinguished. Recently, models ofhe arc near the

current maximum have been devised for calculating the diameter of the arc.These models are needed since he arc diameter constitutes oneof the critical

dimensions needed or optimizing the geometry of the nozzles thatare used ingas blasted interrupters. It should be recognized that all of these models pro-vide only an approximate representation ofhe interrupting phenom ena; but,sthe research continues and withhe aid of the digital computer, more advancedand more accurate models that include partial differential equations describincomplex gas flow and thermodynamic relationshipsre being developed.

In the section that follows, a of qualitative descriptionsof someof the early classical theories, andof some ofthe most notable recent oness



tSYSTEM

RECOVERY CASE 1

RE-IGNITION

PQlNT

Figure 1.9 Graphical representation of the "Race Theory."

presented. The chosen theories re those that have proved to closely representthe physical phenomena, or to those that have been used as the basis for thedevelopment of more complex, combined modern day theories.

One of the earlier theories, usually referredo as the wedge theory,is nowlargely ignored, however t is briefly mentioned here becausef the strong in-fluence it had among researchers during thearly research days in the area ofarc interruption and of its application to the emerging circuit breaker technol-

O D .

Slepian's Theory

The first known formal theory of arc interruption was introduced by JosephSlepian in 1928 The Slepian theory is known as the "race theory itsimply states hat successful nterruption s achieved whenever the rate atwhich the dielectric strength of the gap increases is faster than the rate a twhich the reapplied system voltage grows.

Slepian visualized the process of interruption as beginning immediatelyafter a current zerohen electrons are forced away fromhe cathode and when



a zone, or sheath composed of positive ions s created in the space imme-diately near the cathode region. He believed that the dielectric withstand of

sheath had tobe greater than he criticalbreakdown value of the mediumwhere interruption had taken place. The interrupting performance dependednwhether the rate of ion recombination, which results in an increase in thesheath thickness, s greater than he rate of rise of the recovery voltage whichincreases the electric field acrosshe sheath.

The validity of this theory is still accepted, but within certain limitations.The idea of the sheath effects is still important for predicting a dielectricbreakdown failure, which is the type of breakdown that occurs several hun-dreds of microseconds after current zero, whenhe ion densities re low. How-ever thismechanism of failure s not quite as accurate for the case of thermalfailures, which generally occurt less than en microseconds after current zero,and when the ions densities are stillsignificant, nd when the sheath regionsare that they can usuallye neglected. The concept ofhe race theory s

graphically illustrated n figure 1.9.1.5.2 Prince’sTheory

This themy which known as the displacement, or the wedge theory was ad-vocated by Prince in the U.S. and by Kesselring4] in Germany. Accordingto theory, the circuit is interrupted if the length of the gas discharge pathintroduced into the arc increases during the interrupting period to such anextent that the recovery voltage not sufficientlyhigh o produce a breakdown

in this path.According to this heory as soon as the current zero period sets the arc isin two by a blast of cooling gases andhe partly conductive arc halves of

the arc column are connected in series with the column of cool gas which ispractically non-conductive. If it is assumed that the conductivity of the arcstubs is high in comparison to that of the gas, then it be assumed that thestubs be taken as an extension of the electrodes. In that case the dielectric

of the path between the electrodes is approximately equal to the

sparking voltage of a needle point gap in which breakdown is preceded by aglow discharge. At the end of this current zero period, which corresponds tothe instant t, the two halves of the arc are separated by a distance

D = 2 v t

where v is the flow velocityof the cooling medium and is the time durationof the current period. Assuming, for example, the interruption of an air blast

circuit breaker where it is given the current zero time t=

microsecondsand the flow velocity v = 0.3 millimeters per microsecond (which corre-



c3

3>

Figure 1.10 Withstand capability of a pair of plane electrodes in air at atmosphericpressure.

sponds to the velocityof sound in air), then using the previously givenequa-tion for the space of cool air, the distance between the electrodes is D= 60

millimeters. Now referring to figure we fmd that for a 60 mm distancethe withstand capability shoulde approximately 50 kV.

1.5.3 Cassie’sTheory

Among thefm

useful dif€erential equation descriting the dynamic behaviorof an arc was the one presented byA. M. Cassie in [5]. Cassie developedhis equation for the conductivity of the arc based onhe assumption that a ighcurrent arc is governed mainly by convection losses during the high currenttime interval. Under his assumption a more or less constant temperature acrothe arc diameter is maintained. However,as the current changesso does the arc

section, but not so the temperature inside of the arc column. These as-

sumptions have been verified experimentallyy measurements taken

of the vena contractaof nozzles commonly used in gas blast circuit breakers.Under the given assumptions, the steady state conductance G of the model issimply proportional to the current, so that the steady state voltage gradientis fixed. To account for the time lag that is due to the energy storage capacity



Q and thefinite rate of energy losses N he concept of he arc time constant (isintroduced. “Time Constant” is given by:

The following expression s a simplified form of the Cassie equation, th isequation is given in terms of the instantaneous current.

For the high current region, ata collected from experimental resultss in goodagreement with the model. However, around the current zero region, agree-ment is good only for high rates of current decay. Theoretically and practi-

cally, at current zero the arc diameter never decays to zero to resultn arc in-termption. At current zero there s .a remaining filament of an arc with a

diameter of only a fraction of a millimeter. filament still is a high tem-perature plasma that be easily transformed inton arc by the reappearanceof a high enough supply voltage. The Cassie modeln many cases is referredas the high current region model of an arc. model has proved to be avaluable tool or describing the current interruption phenomena specially whenit is used in conjunction with another wellnown model, the Mayrmodel.

1.5.4 Mayr’s Theory

Mayr ook a radically different approach to that of Cassie, and in 1943 hepublished his theory He considered an arc column where the arc diameteris constant and where he arc temperature varies as a function of time, and ofthe radial dimension. He further assumedhat the decay of the tem perature ofthe arc was due to thermal conduction and that the electrical conductivity ofthe arc was dependent on temperature.

From an analysis of the thermal conduction n Nitrogen at 6000 Kelvin andbelow, Mayr found only a slow increase f heat loss rate in relation with theaxial temperature; thereforehe assumed a constant power lossNo which is in-

dependent of temperature or current. The resulting differential equation is:

where:



T he validity of theory during the current zero period is generally ac-

knowledged and most investigators have used the Mayr model near currentzero successfully primarily because in region the radial losses are the

most dominant and controlling factor.

Browne’s Combined Theory

It has been observed that in reality, the arc temperature is generally wellabove the 6000 K assumed by Mayr and it is more likely to be in excess of

K. These temperatures areso high that they lead to a linear increaseofthe gas conductivity, instead of the assum ed exponential relationship. To takeinto consideration these temperatures and in order to have a proper dynamic

response representationt is necessary that, in a model like Mayr’s, the modelmust follow closely an equationf the type of Cassie’s during the current con-

trolled regime.

T. E. Browne recognized need and in [7] he developed a com-posite model using a Cassie like equation to define the current controlled arc

regime and then converting to a M ayr ike equation for the temperature con-trolled regime, and n the event that interruption does not occur athe intended

current zero,he reverted again tohe Cassie model.

The transition point where each of these equations are considered to be

applicable is assumed to be at an instant just a few microseconds around the

point where he current reachests normal zero crossing.In 1958 Browne extended the application of his combined model [S] to

cover the analysis of thermal re-ignitions that occur duringhe first few micro-

seconds following critical post current zero energy balance period.

Starting with the Cassie and he Mayr equations and assuming that beforecurrent zero the current is defined by the driving circuit, and that fter current

the voltage applied acrosshe gap is determined strictly y the arc circuit,

Browne assumed hat the Cassie equation is applicable to the high current re-

gion prior to current zero and also shortly after current zero following a ther-mal re-ignition. The Mayr equation was useds a bridge between he regions

were the Cassie concept was applied. Browne reducedhe Cassie and he Mayrequations to the following two expressions:

1). For the Cassie’s period rior to current zero

2

for the Mayr’s period around current



Experimental evidence [9] has demonstrated that model is a valuabletool which has a practical application and which has been used extensivelynthe design and evaluationof circuit breakers. However its usefulness depends

on the knowledge of the constant (which only be deduced from experimen-tal results. Browne calculated constant from tests of gas blast interruptersand found it to be in the order of one microsecond, which is in reasonableagreem ent with the commonly accepted ranges found by other investigators

W ] .

1.5.6 Modern Theories

In recent years there has been a proliferation of mathematical models; How-ever these models re mainly developments on more advanced m ethodologiesfor performing numerical analysis, using he concepts establishedby the clas-sic theories that have been described.

But also therehas been a num ber f new more complex theories that havebeen proposed by several of investigators. Signifcant contributionshave been made by: Lowke and Ludwigll], Swanson [12], Frind [13], Tuma[14,51, and Hermann, Ragaller,Kogelschatz,Niemayer,ndhade[16,17,18]. Among these works, probablyhe most significant technical contri-bution canbe found in the investigations of Hennann and Ragaller [18].Theydeveloped a model that accurately describeshe performance of and of SF,interrupters. What is m e r e n t in model, from the earlier models, is thatthe effects of turbulence downstreamf the throat of the nozzle are taken intoconsideration.

a) There is a temperature profile which encompasses three regions; the first

one embodies the arc core, the second that covers the arc's surrounding ther-mal layer and he third consistingof the extemal cold gas.b) The arc column around current zero is cylindrical and the temperature dis-triiution is independent of its axialposition, andc) The average gaslow velocities are proportionalo the axial position.

The relatively consistent and close agreement, that has beenobtained,between the experimental results and the theory suggests that, although somerefinements may still be added, is the model thatso far has given he bestdescription and the most accurate representationf the interruption process na circuit breaker.

The models that have been listed in section have as a common de-nominator the recognition of the important role played by turbulence in theinterrupting process. Sw anson l91 for example has shown thatat 2000 A tur-

On model the following assumptions re made:



bulence has a negligible effect on he arc temperature, while at 100 A turbu-lence makes a difference of 4000 Kelvin and at current zero the differencemade by turbulence reaches values of over 6000 Kelvin. In a way these newmodels serve to prober reinforce the validity of the Mayr equation since withthe magnitudes of temperature reductions produced by he turbulent flow thearc column cools downo a rangeof values thatare nearing those assumed bythe Mayr equation where he electrical conductivity varies exponentially with

respect to temperature.

REFERENCES

1. Ayrton H :The Electric Arc(D.Van Norstrand New York, 1902.2. J. Slepian, Extinction of an a.c. Arc, Transactions AIEE 47p. 1398,1928.3. D.C. Prince andW. F.Skeats, The oil blast circuit breaker, Transactions

4. F.Kesselring, Untersuchungen an elektrischen Lichtogen., ETZ. vol. 55,92, 1932.

5. A.M.Cassie, Arc Rupture and circuit severity: A new heory, Interna-tionale des Cirands Reseaux Electriques’a Haute Tension (CIGRE), Paris,France, Report No.102,1939.

6. 0. Mayr,Beitrage urTheoriedesStatisghenunddesDynamishenLichtbogens, ArchivfurElectmtechnik, 37,12,588-608,1943.

7. T. E. Browne, Jr. A study of arc behavior near current zero by means ofmathematical models, AIEE Transactions 67: 141-143,1948.

8. T. E. Browne Jr., approach to mathematical analysis of a-c arc extic-tion in circuit breakers,AIEE Transactions 78 (Part III) : 1508-1517,1959.

9. J. Urbanek, The time constant of high voltage circuit breaker arcs before

10. W. Reider, J. Urbanek, New Aspects of Current Zero Research on CircuitBreaker Reignition. A Theory of Thermal Non Equilibrium Conditions

(CIGRE), Paper 107,1966.11. J.J. Lowke and H.C. Ludwig, A simple modelfor high current arcs stabi-

lized by forced convection, Journal Applied Physics 46: 3352-3360,1975.12. B. W. Swanson and R. M. Roidt, Numerical solutionsfor an SF, arc, Pro-

ceedings IEEE 59: 493-501,1971.13. G. Frind and J.A Rich, Recovery Speedof an Axial Flow Gas Blast Inter-

rupter, IEEETransactions Pow. App.Syst. P.A.S.-93,1675, 1974.14. D.T. Tuma and J.J. Lowke, Prediction of Properties of arcs stabilized by

forced convection, J. Appl. Phys. 46: 3361-3367, 1975.

AIEE 50 pp 506-512,1931.

Proc. IEEE 59: 502-508, April, 1971.



15. D.T. uma and F. R. El-Akkari, Simulations of transient and current

behavior of arcs stabilized y forced convection,IEEETrans. ow. Appar.

16. W. Hennann, U. Kogelschatz, L. Niemeyer, K. Ragaller, and E. Shade,Study of a high current arc in a supersonic nozzle flow, Phys. D. Appl.

17. W.Hermann, . Kogelschatz, K.Ragaller,and E. Schade, Investigationof

a cylindrical axially blown, high pressure arc, J. Phys. D: Appl. Phys. 7:607,1974

18. W.Hermann, nd K. Ragaller, Theoretical description of the current inter-ruption in gas blast breakers, Trans.IEEEPower Appar. Syst. PAS-96:

19. B. W. Swanson, Theoretical models or the arc in the current regime,

Syst. PAS-96:1784-1788,1977.

Phys.7:1703-1722,1974.

1546-1555,1977,

(edited byK Ragaller) Plenum Press, Nework: 137,1978.



2

SHORT CIRCUIT CURRENTS

2.0 Introduction

During its operation an electric system is normally in a balanced, or a steadystatecondition. steadystateconditionpersists as long as nosuddenchanges take place in either the connected supply orhe load of the circuit.

Whenever a change of the normal conditions place in the electricsystem, there s a resultant temporary unbalance, andue to the inherent inertiaof the system there s a required finite period of time needed by he to

re-establish its previously balanced r steady state condition.If a fault, n the form of a short circuit current, occursn an electric system,

and if as a result of such faultt becomes necessary to operate n in te rq t ingdevice; the Occurrence of both events, the fault and the current interruption,constitute destabilizing changeso the system that resultn periods of transientbehavior for the associated currents and voltages.

Interruption of the current in acircuitgenerally place duringatransient condition that has been brought about by the occurrence of a short

circuit.Thenterruptiontselfproduces n dditionalransienthat issuperimposed upon he instantaneous conditionsof the system, and thus tbe recognized hat nterruptingdevicesmustcopewith ransients in thecurrents hathavebeengeneratedelsewhere in the system, plus voltagetransients that have been produced byhe interrupting device itself.

2.1 Characteristics of the Short Circuit Current

Because of their magnitude, and of the severity of its effects short circuit

currents, undoubtedly, represent he most important type of current transientsthat exist in any electrical system.

The main factors hatdetermine the magnitude,andother mportantcharacteristics of a short circuit current are, among others,he energy capacityof the source of current, the impedance of the power source, he characteristicsof the portion of the circuit that is located between he source and the point ofthe fault, and he characteristics of the rotating machines thatare connected tothe system at the time of the short circuit.

The machines, whether a synchronous generatorr either an induction or asynchronous motor, are all sources of power during a short circuit condition.



Since at the time of the short circuit, motors will act as generators feedingenergy into he short circuit dueo the inherent inertiaof their moving parts.

The combination of these factors and he instantaneous current conditions,prevailing at the time of initiation of the fault will determine the asymmetry of

the fault current as well as the duration of the transient condition for th is

current. These two characteristics are quite important, as it will be seen later,for the application of an intem pting device.

2.1.1 Transient Direct Current Component

Short circuit current transients produced by a direct current source are lesscomplex than those producedby an alternating current source. The transientsoccurring in the dc circuit, while either energizing, r discharging a magneticfield, or a capacitor through a resistor, re fully defmed by a simple exponen-tial function.

In ac circuits, the most common short circuit current transient is equal to

the algebraic of a transient direct current component, which, as statedbefore, can be expressed in terms of a simple exponential function, and of asteady state alternating current component thats equal to he fmal steady statevalue of he alternating current, and which cane described by a trigonometricfunction.

The alternating current components created by the external ac source thatsustains the short circuit current. The dc component, in the other hand, doesnot need an external source and is produced by the electromagnetic energystored in the circuit inductance.

A typical short circuit current wave form showing the above mentionedcomponents is illustrated in figure 2.1. In th is figure the directcurrentcomponent is shown as I , he fmal steady state component is shown as IAc

and the resulting transient asymmetrical currentsITotOl

Itshould be noted hat, as aresultand in order to satisfy the initialconditions required or the solution of the differential equationhat defines thecurrent in an inductive circuit, the value of the direct current component isalways equal and opposite to he instantaneous valueof the alternating currentat the moment of the fault initiation. Furthermore t should be noted that t h i s

dc transient current s responsible, and determines he degree of ofthe fault current.

A short circuit currentwill be considered tobe symmetrical when he peakvalues of each half cycle re to each other when measuredn reference toits normal An asymmetrical current will be one which is displaced in

either direction from ts normal and one in which the peak value will bedifferent for each half cycle with respect to he normal



I I

. .. 1 ;

i o ;

I /-.-

i 0

. .

Figure 2.1Transients componentsand response o f an ac short circuit current.

The total current is mathematically described y the following function:

where :

I ,,,= Peakvalue of the steady state acurrent

a= System Time Constant=Rn,

$= fault’s initiation angle

R = System Resistance

L = System Inductance



/ \

Figure 2.2 Volt-Time and current relationships for a short circuit initiated at the

instant of voltage zero.

2.1.2 The Volt-Time Area Concept

To better understandhe physics phenomena involvinghe circuit response o achange of current flow conditions,t is extremely helpful o remember that n apurely inductive circuit the current will always be proportional to the Volt-Time area impressed upon the inductance of the affected circuit. This state-ment implies that he current is not instantaneously proportional tohe appliedvoltage but that it is dependent uponhe history of the voltage application.

The visualization f concept, in terms of what we are calling the Volt-

Time is importantbecause it helps to present the inter dependentvariations of the voltage and the current at its simplest level. This concept isdemonstrated by simply observing the dimensional relationship that exists inthe basic formula which describeshe voltage appearing acrossn inductance:

e = L d i / d t

Dimensionally expression be written as follows:

Inductance Current

TimeVolts=



Solving (dimensionally) or the current we have:

Volts Time

InductanceCUHent =

To illustrate the concept, let s consider some specific examples of a fault.In these examples it will be assumed that the circuit is purely inductive, andthat there s no current flow prior to the momentf the insertion of the fault.

First, let consider the case illustrated in figure 2.2 when the fault is

initiated at the precise instant when the voltages zero. In the figure it can beseen that the currents totally shifted above the and that the current peakis twice the magnitude of the steady state current. The reason for this isbecause, in the purely inductive circuit, theres a 90bhase dif€erence betweenthe current and the voltage. Therefore, the instantaneous current should havebeen at its peakwhen the voltage is zero; furthermore,s it was statedearlier,the value of thedirectcurrentcomponent is andopposite o thealternating current t the same instantf time.

It can also be seen that at time tl the current reaches its peak at the samemoment as the voltage reaches a zero value,his instant corresponds with thepoint where the rea under the voltage e everses.

At time tz it is noted hat both, he current and the voltage reach aninstantaneous value of zero. This instant corresponds with the time when thearea A2 under the voltage m e egins its phase reversal, the two areas beingequal it then follows that the net area value would be zero, which in turn is

the current value.The second example, shown n figure 2.3 illustrates the condition when the

short circuit current is initiated at the instant where the voltage is at its peakvalue. At time t2 it is observed that the area under the voltage m e ecomenegative and therefore the current s seen to reverseuntil at time t3where thecurrent becomes zero since the areas A , and A2 are equal, as it is seen in thecorresponding figure.

The first example describes the worst condition, where the short circuitcurrent is fully displaced from its axis and where the maximum magnitude isattained.

The second example represents the opposite, the most benignant of theshort circuit conditions, the currents about its axis since thereisno direct current contribution and the magnitude of the short circuit is thelowestobtainable for allother aultswhere hevoltageand hecircuitimpedance are the same.

The last example is show in figure 2.4 and is given as an illustration of ashort circuit that is initiated somewhere between the voltage zero and thevoltage maximum.



Figure 2.3 Volt-Time and current relationships for a circuit current initiated atmaximum voltage.

One significant characteristic that takes place, and that is shown in thefigure is the existenceof a major and a minor loopf current about the In

figure we can observe the proportionality of the current and the Volt-Timcurve.

It shouldbe noted that the majorCurrent loop s the resultof the summationof areasA I and A2and the minor loop corresponds to the summationf A3 and

From the above examples and y examining the correspondingigures, theA4

following facts become evident:

1) Peak current always occur t a voltage zero.2) Current zero always occur when theet Volt-Time areas are zero.The magnitude of he current is always proportional to the Volt-Time area

4) At all points on the current wave the slopef the current is proportional to

In all of the examples given above t was assumed that the circuit was purelyinductive. This was done only to simplify the explanation of how the shortcircuit current waves formed.

In real applications, however, condition is not always attainable sinceall reactors havean inherent resistance, and therefore thec component of the

the voltage at that time.



Figure 2.4 Volt-Time and current relationships illustrating major and minor loops ofshortcircuit current.

current will decay exponentially as a result of the electro-magnetic energybeing dissipated through resistance. condition was illustrated earlierin figure 2.1 where the general formof the short circuit current was introduced.

2.1.3 Transient Alternating Current Components

In the preceding discussions the source of current was to be farremoved from he location of the short circuit and thereforehe ac current wassimply definedby a sinusoidal function. However, under certain conditionshetransient ac current may have an additional ac transient component which isthe resultof changes produced byhe short circuit current n the inductance ofthe circuit.

This condition generally happens when a rotating machine is one of thecircuit components that is involved with the short circuit., more specificallyit

refers to the condition where he contacts making he connection to a generatorare closed on a short circuit at the time when the generated voltage passesthrough its peak. Whenever happens, the short circuit current rises veryrapidly, its rise being limited only by the leakage reactance of the generatorstator, or by its sub-transient reactance. This current creates a magnetic fieldthat tends to cancel the flux at the air gap, but to oppose these changesn emf

is induced in the generator winding and eddy currents re induced in the polefaces. The net result is that the ac component is not constant in relation to



Subtransient Component

Transient Component

Steady State Component

W

Figure 2.5 composition a short circuitcurrent ncluding the ac transientcomponents.

time, but instead it decreases from n initial high value to a constant or steadystate value. The particular formf the rate of decrease precludes he use of asingle exponential function, and instead it makes it necessary to divide thecurve into segments and to use a different exponential expression to defineeach segment.This results in the use of very distinctive concepts of reactancefor each exponential function.T h e “subtransient” reactance is associated withthe first, very rapid decrease period;he “transient” reactance, withhe second,lessapid ecrease eriod, nd the “synchronous”eactancewhich is

associatedwith the steady tatecondition,after all the transientshavesubsided. In figure 2.5 the transient components as well as the resulting totalshort circuit currentre shown.

2.1.4 Asymmetry of Three Phase Short Circuit Currents

In all of the material that has been presented so far, on the subject ofcircuit currents, only single phase system currents were escriied . The reasonfor is because in a simultaneous fault, f a three phase balanced or

for that matter in any balanced multi-phase system, only one maximum dccomponent can exist, since only one phase satisfy the conditions for suchmaximum at the instant wherehe fault occurs.



Figure2.6 Threephase short circuit currentcharacteristics.

Furthermore; a symmetrical short circuit current not occur in all threephases of a three phase generator, simply becausef the current displacementinherent to the a multi-phase system. If a symm etrical fault occurs in onephase, the other two phases will have an oppositedirectcurrentcomponents, since in all cases the algebraic sum of all of the dc components

must be equal to zero.The preceding statement is demonstrated by referring to figure2.6, wherethe three phase steady state currents are shown, and where it can be readilyseen that f the peak value of the alternating current is assumed toe to 1

in all the phases, and as it has been established before, the initial value of thedc component in a series LR circuit s equal and opposite to the instantaneousvalue of the ac current which would exist immediately after switching if thesteady state could be obtained instantaneously. Observing figure 2.6 we

see that when theinitial value of the dc component on phaseA is I,,,t meansthat the instantaneous valuef the ac current isat its peak, and the steady statevalue for the currents in phases B and C at the same instant is one half that ofthe peak valuef A.

For a fault that at instant 2, there is no offset on phase since currentB is at that nstant,andwegetanoffsetonphase A that is equal

to-m, hile the offset on phaseis to+mWhen the short circuit is initiated at time we get a condition that is

similar to that which was obtained when the starting pointor the fault was atinstant 1, except that the maximum ffset is observed now on phaseC insteadof on phaseA.



Having hown the conditionswhere the maximum dc component isproduced, let us next look at what happens o the currents some ime later. Anycombination of dc components hat produces a maximum average offsett t =

0 will also produce he maximum offset at any instantthereafter, compared toany other possible combinationsf dc components. This follows from the factthat the decay factor of the dc component is identical for the three currents,since it is determined by the physical components of the system which has

been assumed to be balanced under steady state conditions. If the dc currentdecays at the same rate in all three phases then all of the dc componentswillhave decayed in he same proportion from whatever initial value they had.Therefore it follows that the maximum asymmetryof a fault that started withthe maximum offset on one phase will occur one half cycle ater in the samephase that had the maximum offset.

When considering a fault that hashe maximum offset on phase we seethat the first peakoccurs in phase C approximately (or t = d 3 0 ) after the

inception of the shortcircuit. The secondpeakoccurs on phase B atapproximately 120 (or t = 2d30). Thehird ccurs on phase A atapproximately 180 (or at t = d m ) . Since the ac components are identical wecan readily ell which peak s larger by comparing the dc components. Startingwith phasesB and C, we can see that they at the same value /2). Wecan also see, without lookingt the absolute quantitative values, that sincehedecay at 60 is less than t 120, the peakof the current on phase C is larger.

Next comparing the peaks of phases A and C, we see that the dccomponent on A starts with twice he value of the component of C, but by thetime when A reaches its peak it has decayed more than whathe C componenthas.

These relative values be compared by establishing their ratio.

From result we conclude that he dc components, and thereforehecurrent peaks, woulde ifx/R has a value such that:

When expression is solved for x%R, which is normally the way in whichthe time constant of a power system is expressed, we obtain a value to3.02. Which means hat for values ofx/R > 3.02 the value of the exponential is



Figure 2.7 Typical asymmetric short circuit current as it can be observed from an

oscillogram. The figure illustrates the parametersused for calculating he instantaneousvalues of current.

greater than1 and therefore he peak on phase A at 18O(will be greater than hepeak on phase C at Conversely for values of x/R smaller than thepeak onphase A willbe smaller thanhe earlier peak n phase C.

Practical in general be expectedo ave xi?z ratios

significantly greater than , and th is will cause he peak on phase at the endof a half cycle to e the greatest of the three peaks. Being h is the most severepeak then it is the one of most interest, as it w ill be seen later, for circuitbreaker applications.

2.1.5 Measuring Asymmetrical Currents

The effective, or rms value of a wave form defined as the square rootof thearithmetic mean of the square of the ordinates of a given curve between two

Mathematically, for an instantaneous current, which s a function of time,zero points.

the effective value maye expressed as follows:



In figure 2.7 a typical asymmetric short circuit current during its transientperiod, as it may be seen in a oscillographic record, is shown. In figure,during the transient period, the of the steady state sinusoidal wave has

been offset byanamount to the dc componentof heshortcircuitcurrent, and it has been plotted asurve Lines 1 and 2 define the envelope fthe otal current, and ine 4 represent he rms value of theasymmetricalcurrent. The dimension , represents the peak to peak valuef the sinusoidalwave, A is the maximum value of the current referred to its own ofsymmetry, or in other words is the maximum of the ac component (In); B andC representshepeakvalue of themajor ndminoroopof urrentrespectively.

The general equation for the current wave shown in figure whichrepresents a short circuit current where, for the sake of simplicity, the actransient decrement has been omitted cane written in its simplest forms:

i = I,sin$ + D

where the termD represents the dc component as previously defined in terms

of an exponential function.As given before, he basic definitionof an rms unction is given as:

Now, to obtain an expression that will represent the rms value of the totalcurrent we substitute the first equation, which defines the value of thecurrent as a function of time into the second expression which as indicated

represents hedefinition of the rms value.After implifyingand omemanipulation of the trigonometric functions we can arrive to the followingexpression:

In this equation the term Ig is used to represent the rms value of the totalcurrent. The term effective is synonymous withms,however in he context of

derivation,andonly for thepurposeofclarity term sused odifferentiate between the ms alue of the ac component of the current andhe



rms value of the total current which contains the ac component plus the dccomponent.

By recognizing that the first term under he radical is the nns value of theac component itself, and that D represents the dc component of the wave, it

be realized then that the effective, or total rms asymmetrical current isequal to the square root of the sum of the squares of the rms value of thealternating current component and ofhe direct current component.

If the dc terms expressed as a percentof the dc component with respectothe ac component it becomes equal to% dc * IM/lOOand when substituted intothe equation for &the following expression results.

Furthermore, if in equation the peak value of the ac component issubstituted by the rms value of the same ac component then the equationbe simplified and t becomes:

and is the form of the equation whichis presently shown in the two most

influential and significant world wide circuit breaker standardsl] [2].When a graphical depiction of the wave form of the transient current is

available, the instantaneous values, at a time for the dc component, and forthe rms of the total current be calculated as follows.

Again referring to figure 2.7 the rms value for the ac component isto:

A

I,, =-J?IFurthermore it is seen that:

& = B + C = 2 A

Then it follows that:

And in terms of the rms value of theac component it becomes:



Also from the figure we find that:

B + C

2

A = C + D = -

And solving for the dc componentD we get:

And finally the rms value for the asymmetric current be written in terms of

quantities that are directly measurable fromhe records as:

2.828

2.2 Calculation of Short Circuit Currents

Short circuit currents are usually determined by calculations from data about

the sources of powerand the impedancesof all interconnected inesandequipment up to the point of the fault. The precise determination of shortcircuit currents specially in large interconnected systems, generally requirescomplicated and laborious calculations. However in most cases there is noneed for those complicated calculations and within a reasonable degree ofaccuracy the approximated magnitude f the short circuit currents be easilydetermined within acceptable limitsor the applicationof circuit breakers andthe settingsofprotectiverelays.Since the resistance in a ypical electric

system is generally low in comparison to the reactance, it is safe to ignore theresistance and to use only the reactance for the calculations. Furthermore forthe calculation of he short circuit current n a system,all generators, and both,synchronous and induction motors are considered as sources of power, loadcurrents are neglected, and when several sourcesf current are in parallel, it sassumed that all the generated voltages are n phase and that they re equal inmagnitude at the time of the short circuit.

While computations of fault currents may be made with reactances ex-

pressed in ohms, a number of rules must be observed to take care of machineratings andof changes in the system voltages thatre due to transformers. Thecalculation will usually e much easier f the reactances are expressed in terms



of per cent or per unit reactances. h o t h e r simple methodof calculation is themethod.

2.2.1 The per Unit Method

The per unit method, basically, consistsof using the ratings of the equipmentas the units for measuring the quantities appearingn problems that involvehesame equipment. It is equivalent to adopting a set of units that are tailored to

the system under consideration. If we, for example, consider the load on atransformer expressed n amperes, it will not tell us how much we are loadingthe transformer in relation to what could be considered sound practice. Beforewe are justified in saying that the load is too much, or is too little, we have tocompare it with the normal or rated load for that transformer. Suppose thathetransformer in our example is a three phase,2300-460volts, 500 KVAtransformer and hat it is delivering200 amperes to the load on he low voltageside. The 200 ampere value standing alone does not tell the full story, but

when compared to he full load current (627 amperesn case) it does havesome signX1cance. In order to compare these wo values we divide one byheother obtaining a value equal to 0.3 18.his result has more significancethe plain statement of the amperes value of the actual load, because it is arelative measure. It tells us that the current delivered by the transformer is0.318 times the normal current. We shall call it 0.318 per unit, or 31.8 percent.

Extending concept to the other parameters ofhe circuit,namely; volts,

currents, KVAs, and reactances, we develop the whole theory for the perunit method.

Let V = actual or rated voltsand V, = base or normal volts

Then, the per unit volts expresseds V, is defined by:

VI/ =-

p v,

Using asmiarnotation, we defineer unit KVA as:

Per unit amperes:



IIpu=-

Ib

and per unit reactance:

X

The normal values or all of the above quantities can e defined with only onebasic restriction: The normal valuesf voltage, current, andKVA must satisfythe following relationship:

which when solvedor it gives:

The base or normal values for reactance, voltage, and current are related byOhm's law as follows:

' b

and when substituting the value ofb we obtain:

Which when solvingor the per unit valuet becomes:

From the above it follows that a device is said to have a certain per centreactance when the reactancedrop of the device, operating t its rated KVA is

that certainper cent of the rated voltage.o describe a reactance, lets say, as5%, is to say thatat rated KVA, or whenfull load rated current s flowing, thereactive voltage drop is equal to 5% of the rated voltage. Expressing the



reactance in per cent is to say that for a rated load, the voltage drop due tohereactance is that number (of volts) per hundred volts of rated voltage. Whenthe reactance is expressed in per unit (pu) number represents the reactivevoltage drop at rated current load per unitf rated voltage.

The per cent and he per unit values when referred to he same base KVA

are related by the following simple expression:

P A =x, * 100

When the reactancesare given in ohms they canbe converted toper unit usingthe following relationship:

In order that he per cent or per unit valuesmay be used for the calculationsofa given circuit,t is necessary thatall such values e referred to the same VAbase. The choice for base KVA be absolutely arbitrary, it does notneed o be tieddown oanything in the ystem.However it will beadvantageous to choose as the KVA base a particular piece of equipment in

which we re interested.When the givenper cent, or per unit values represent the ratingsf a piece

of equipment that have different base thanhe one that we have chosen ashebase KVA for our calculations t will be necessary to translate this informationto the same basis. his is readily accomplished by obtaining the proper ratiosbetween the valuesn question, as it is shown below.

2.2.1.1 General Rule sfor Use ofper Unit Values

The following rules, tabulated below, are given to provide a quick source ofreference for calculations that involve theer unit method.

When all reactances are expressed in per unit to the same base KVA, thetotal equivalent reactances

a) for reactances in series:

x,,,=x, x, . .+xi



b) for reactancesnparallel:

I

-+-+ .......+-Totcrf= 1 1 I

x. x ,

c) To convert reactances from delta to wye or vice versa:

Delta toWye

x, =x,+x, x3

Wye toDelta



2.2.1.2 Procedure or Calculating Short Circu it Currents Using thep er Unit

The complete procedure used to calculate a short circuit current using theerunit method be summarized in the following six basic

Choose a base KVA2. Express all reactances n per unit values referred to the chosen baseVA

Simplifyhe ircuitby ppropriately ombining all of thenvolvedreactances. The objectives to reduce he circuit to a single reactance.

4. Calculate the normal, or rated current at the rated voltage, at the point ofthe fault corresponding to the chosen baseVA.

Method

( W A ) ,

f i x vI, =

5. Calculate the per unit short circuit current corresponding to the erunit system voltage divided byhe per unit total reactance.

6. Calculate the fault current magnitude by multiplying theer unit current

(Ip3 imes the rated currentIJ.

The following example s given to illustrate the applicationf the per unitmethod to the solution of a short circuit problem. The circuit to be solved isshown asa single line diagramn figure 2.8.

Choose the value that has been giveny the utility as the baseMVA This

2 Calculation of the reactances for the differentportions of the system yieldsvalue is MVA.

the following values:

For the 69 kV ine Xbm, = 5.9592

For the kV line X,,, == o.24



SYSTEM69 kV800 MVA

Tl=3.5 G1

25 MVAX” = 0.1

30 MVA0.08

T2=T3=T47.5 T3 T4X = 0.06

M1 =M25000 HPX” = 0.15

FAULT

Figure 2.8 Single lineaagram or distribution system solved in the example problem.

For the 4.16kV ine x,,, =4.162 x

800,000= 0.022

3. Simplifying the circuity refero figure 2.9 (a,b,c,d,e); obtaining the perunit reactances of each component we haves shown n (a):

System (1) = X I = I . 0



800MVA E =l

F

Figure2.9Block diagram showing reducing processof circuit in figure2.8.



Transformer Tl(3 )

(KVA)6me 800,000

(KVA)ra&=x3 =x x = 0.08 X- 2.13

30,000

Generator G1 (4)= X4 =0.1 x 800,000

25,000

= 3.2

Transformers T2, T3,and T4 (5,6,8)

=X5, X6,X8 =0.06 x 800,000

7,500= 6.4

Motom M1, andM2 (7,9) = X7,X9 =

0.15 x 800,000

5,000 = 24

Continuing the processof reducing the circuit, and referring to figure 2.9(b),the series combination of 1,2,3s added arithmetically to give:

X1,2,3 (1.1) = 1.0+ 0.65 + 2.13 = 3.78

The two branches containing the components6 , 7 and 8, 9 are combinedindependently.

(6. l) , (8.1) = 6.4 +24 = 30.4

Next, the just reduced erieselements are combinedwithheparallelcomponents, obtaining the new components which are represented by 1.2 and6.2 in figure 2.9 (c).

(12) =X1,2,3x X 4 - 3 . 7 8 ~.2

Xl,2,3 +X 4 - 3.78+3.2= 1.73 and,

(6.2) =X6,7 x X8,9 30.4 x 30.4

X6,7+X8,9 30.4+30.4- = 15.2

Continuingwith

the process in figure 2.9 (d) we have:

(13) =(12) x (6.2)1.73 x 15.2

(1.2)+(62) - L73 +15.2- = 155



and finallywe add (1.3)+ (5) to obtain the total short circuit reactance:

(1.4) =Xsc =7.95

We then calculatehe value of Ipll which is to:

I- is calculated as:

800,000

4.16baw(4.16) = = 111,029

Now we proceed with the calculation of the short circuit current at thespecified fault locationF).

I , = Ipu Ibae = 0.126 111,029= 13,966 Amperes

Alternatively the short circuit current also be calculated by figuring theequivalent short circuitMVA and then dividingt by the square root timesherated voltage t the point of the short circuit.

2.2.2 TheW A Method

The MVA method is insrealitya variation of the per unit method. It generallyrequires a lesser number of calculations, which makes method somewhatsimpler than the per unit method. The MVA method is based in the fact thatwhen it is ssumed that a short circuit currents being supplied fromn infinity

capacity source, the admittance, which is the reciprocal of the impedance,represents the maximum current, or Volts Amperes at unit voltage, which can

flow hroughacircuit or an ndividualcomponentduringashortcircuitcondition.



With the MVA method, the procedure is similar to that of the per unitmethod. The circuit is separated into components, and then these componentsare reduced until a single component expressed termsf its MVA is obtained.The short circuit MVA for each component is calculated in terms of its owninfinite bus capacity. The value or the system MVA is generally specifled bythe value given by the utility system. For a generator, or a transformer, a lor cable, the MVA is equal to the equipment rated MVA divided by its ownimpedance,andby hesquare of the ine to linevoltagedividedby heimpedance per phase, respectively. The MVA values of the components arecombined according to the following conventions:

1. For components in series:

1

1 1 1ATotd =

+-+...-( M A ) 1 ( M 4 2 ( M A )

2. For components in parallel:

Toconvert romDelta oWyeor vice versa the same ules hatwerepreviously stated n section 2.2.1.1for the per unit method are applicable.

The same example, shown n figure 2.8, that was solved before, using theper unit method, will now be solved using the MVA method. Referring tofigure 2.9 we use the same schematics while using theer unit method.

The MVA values for each componentare calculated as follows:

For the system (1) theMVA =-For the line (2) the A - 1,360

g2

3.5

For transformerT l (3) theM A =-0

0.08 - 375

For the generator G14) theM A =- 2505

0.1

For the transformers T2,3 and (5,6,8) the MVA =- 1255

0.06



For the motors M1,2 (1,2)he W A =- 33.30.1 5

The MVA value for the combined 1,2,3 is:

MvA(l.l) =1

1 1 1-+- +-800 136075

= 465

In figure 2.9@), he W A or (6.1) and (8.1)s:

W A 6.1), (8.1) = = 26.325 33.3

125+ 33.3

In 2.9 (c) the reduced circuit is obtained by combining (6.1) + (8.1) to give((6.2) whoseM vA value is:

(6.2) W A 26.3 +26.3 = 52.6

Next, as shown n (d) he MVA for (1.3) is equal to he parallel combinationof(1.2 and (6.2) which numericallys equal to:

(1 .3 )WA = 465 + 52.6 = 518

Finally combining heW A S f (1.3) and (5) we obtain:

(1.4) =WA , = 518 125 = 100.69518+125

Now the value of the short circuit currentan be calculated as follows:

Is' = 4.16100*69 = 13,975Amperes

It is now left up to the reader to chose whichever methodspreferred.

2.3 Unbalanced Faults

The discussion far has been based under the premise that the short circuitinvolved all three phases symmetrically, and therefore that fault had setpa new three phase balanced system where onlyhe magnitude of the currents



had changed. However, it is recognized that other than three phase balancedfaults can happen in a system; for instance thereay be a line to ground or aline to line fault.

Generally it is the balanced three phase fault where the maximum shortcircuit currents can e observed. In a line to line fault very seldom,f ever thefault currents woulde greater than those occurringn the three phase balancedsituation. A one line to ground fault obviously is of no importance if the

system is ungrounded. Nevertheless n such cases, because f the nature of thefault a new three phase system, where he phase currents, and phase voltagesare unbalanced s set up.

Analytical solution of the unbalanced system is feasible but it is usuallyhighly involved and oftenery difficult. The solutionf a balanced three phasecircuit, as it has been shown, is relatively simple, because all phases beingalike, one can be singled out and studied individually as if it was a singlephase. It follows then,hat if an unbalanced three phase circuit could somehow

be resolved into a number of balanced circuits then each circuit might beevaluated based on its typical single phase behavior. The result f each singlephase circuit evaluation could then e interpreted with respect to the originalcircuitusing heprinciple of superposition.Such tool for the solution ofunbalancedfaults is afforded by the technique known as the symmetricalcomponents method.

2.3.1 Introduction to Symmetrical Components

The method of symm etrical components s based on Fortesque’s theorem,whichdeals ngeneralwith the resolution of unbalanced ystems ntosymmetrical components. By method an unbalanced three phase circuitmay be resolvedntohree alanced omponents.Each omponent issymmetrical in itself and therefore it can be evaluated on the basis of singlephasenalysis.Thehreeomponentsre:he ositive-phase-sequencecomponent,henegative-phase-sequence omponent, nd the zero-phase-sequence component.

The ositive-phase-sequence omponent ompriseshree urrents (orvoltages) all of equal magnitude, spaced 120 (apart and in a phase sequencewhich is the sam e as that of the original circuit.If the original phase sequenceis, for exampleA, B, C, then the positive-phase-sequence components also

B, C. Refer to figure 2.10 (a).The negative-phase-sequence component also comprises three currents (or

voltages) all of equal magnitude spaced 120(apart, and in a phase-sequenceopposite to he phase sequenceof the original vector, thats C, B, see figure

2.10 (b).



Bn

A0

Bo

CO

Figure 2.10. Symmetricalvector system, (a) Positive-phasesequencevector, (b)

Negative-phasesequence vector;(c) Zero-phase sequence vector.

The zero-phase-sequence component is constituted of three currents (orvoltages) all of equal magnitude, but spaced O(apart, figure 2.10 (c), inzero-sequence the currents or voltages, as it can be observed, are in phase witheach other and in reality they constitute a single phase system.

With the method of symmetrical components, currents and voltagesn each

phase-sequence interact uniquely; these phase-sequence currentsr voltages donot have mutual effects with the currents or voltages of a different phase-sequence, and consequently he systems defined by each phase-sequence maybe handled quite independently and their results can thenbe superimposed toestablish the conditionsf the circuit s a whole.

Defining impedance as the ratio of the voltage to its respective currentmakes possible to define impedances that can be identified with each of hephasesequencecomponents.Therefore, here is apositive-phase-sequenceimpedance, a negative-phase-sequence impedance and a zero-phase-sequenceimpedance.Each of thesempedances may also be resolvedntoheirresistance and reactance components.

The olutionof the short ircuitusinghemethodof ymmetrical

components is carried out in much of the same manner as with the per unit orMVA methods. The difference is that the negative and zero reactancesare included in the solution.

In a balanced circuit the negative-phase sequence, and the zero phase-sequencecomponents are absent,andonly the positive-phase-sequence ispresent, therefore the solution is reduced to the simplified m ethods that weredescribed earlier.In the unbalanced circuit he positive and he negative-phase



sequencecomponentsarebothpresent,and nsomecases hezero-phase-sequenceomponentmaylso be present.heero-phase-sequencecomponent will generally be present when there is a neutral or a groundconnection. The zero sequence components, however are non-existent in anysystem of currents or voltages if the vector of the original vectors s equalto zero. This also implies thathe current of a poly-phase circuit eedi ig into adelta connection s always zero.

2.3.1.1 Reactances or Computing Fault Currents

As stated earlier, there is a reactance which is idenwlable with each phasesequence vector, just as any other related parameter, such as the current, thevoltage, the impedance. The reactance can e expressed under the same set ofrules described and consequently;he reactances can e identified as follows.

The Positive-Phase-Sequence Reactance, Xp, is the reactance commonlyassociated and dealt with in all circuits, it is the reactance we are already

familiar from our applications of the simplified per unit method. In rotatingmachinerypositive-phase-sequencereactancemayhave hreevalues;sub-transient, transient and synchronous.

The Negative-Phase-Sequence Reactance,Xn, s present in all unbalancedcircuits. In all lines and static devices, such s transformers, the reactances topositive and negative-phase sequence currents is equal, that is, X p = Xn forsuch type of apparatus. For synchronous machines and rotating apparatus in

general it is reasonable to expect hatdue o the rotatingcharacteristics,

reactances to he positive and he negative-phase-sequence currents will noteequal. In contrast with the possibility of three -dues for the positive-phasesequence, the negative-phase sequence has but one or rotating machinery, itsmagnitude is nearly the same as that of the sub-transient reactance for thatmachine.

TheZero-Phase-SequenceReactance, Xo, dependsnotonlyupon theparticular characteristicof the individual device, Which muste ascertain foreachdeviceseparately. but also it dependsupon the way the device isconnected.Forransformers, for example, the value of the zero-phase-sequence reactance s given not only by he characteristics of its windings butalso by he way theyare connected.

2.3.1.2 Balanced Three Phase Faults

For balanced hree phase faults he value for the total reactanceX t is given bythe value of the positive-sequence reactance p alone.

2.3.1.3 Unbalanced Three Phase FaultsFor unbalanced three phase faults, the same equations m ay be used with thespecial understanding s to what the value ofX t is for each case.



Line-to-LineFault:

Xp +Xnx t =-5

This kind of fault seldom causes a fault current greater thanhat for a balanced

three phase fault because ofheusual

relationship of X p and Xn. It is robable,therefore, that investigation of the balanced faultwill be sufficient to establishthe possible maximum fault current magnitude.Line-to-Ground Fault.

however, if the system is grounded we havehe following:Obviously type of fault s of no importance or an ungrounded

x t =X p + X n + X o

3

This equation, usually may e simplified to:

x t =2xp+x0

2.4 Forces Producedby Short Circuit CurrentsThe electromagnetic forces that are exerted between conductors, wheneverthere is a flow of current, is one of the most significant, well known, andfundamental phenomena thats produced by he electric current. Bent bus bars,broken support insulators and in many instances totally destroyed switchgearequipment is the catastrophic result of electrodynamic forces that are out ofcontrol. Since the electrodynamic forces are proportional to the square of the

instantaneous magnitude of the current, then, it is to be expected that theeffects of the forces produced by a shortircuit current wouldbe rather severeand oftentimes quite destructive.

The mechanical forces acting between the individual conductors, parts ofbentconductors, or contactstructures within switchingdevices, attainmagnitudes in excess of several thousand pounds (newtons), per unit length.Consequently the switching station and ll of its associated equipment must edesigned in eitherf two general wayso solve the roblem

One way would be to design all the system components so that they areN l y capable of withstanding these abnormal forces. The second alternativewould be to design the current path in such a way as to make the electro-



magnetic forces to balance each other.In order to use either approach and toprovide the appropriate structures, it is necessary to have at least a basicunderstanding about he forces that are acting uponhe conducton.

The review that follows is intended as a refresher of the fimdamentalconcepts nvolved.Itwillalsodescribeapracticalmethod to aid in thecalculation of he forces for some of the simplest, and most comm on caseshatare encounteredduring the design of switchgear quipment.Relatively

accurate calculations for any type of conductor’s geometric configurations repossible with a piece by piece approach using the methods to be described,however the process will be rather involved and tedious, for m ore complexarrangements it will preferable to use one of the ever increasing number ofcomputer programs developed to accomplishhe

2.4.1 Direction of the Forces Between Current Carrying Conductors

This section will be restricted to establish, only in a qualitative form, the

directionof the electromagnetic orces in relation to the instantaneousdirection of the current. Furthermore, and for the sake of simplicity, onlyparallel, or perpendicular pairsof conductors will e considered.

To begin with, it would be helpful to restate the following well h o w nelementary concepts:

1. A current carrying conductor that is located within a magnetic field issubjected to a fo rce that tends to move the conductor. If the field intensity is

perpendicular to the current, the force is perpendicular to both, the magneticfield, and he current.The relative directions f the field, the current, and he force are d e s m i d

by Fleming’s left-hand rule; which simply says that,n a.three directionalthe index finger points in the direction of he field, the middle finger points inthe direction of the current and he thumb point n the direction of he force. Inreality relation is all that is needed to determine the direction of the forceon anyportionof onductorwhere the magneticieldntensity is

perpendicular to he conductor.2. The direction of he magnetic field around a conductor s described by theright-hand rule.This rule requires that the conductor be grasped with he righthand with the thumb extended in the direction of he current flow, the curvedfingers around he conductor, then will establishhe direction of he magneticfield.

Another useful concepto remember is that; the field intensity at a point n

space, due to an element of current, s considered as the vector productof thecurrent and the distance to the point. This concept will be expanded in thediscussions that re to follow.



Y

Figure 2.11 Graphicalepresentation of theelationship between current,

electromagnetic field, nd electromagnetic forceor a pairof parallel conductors(Biot-

SavartLaw).

4. For the purpose of he discussions that are to follow, the directions in spacewill be representedbyasetofcoordinatesconsisting of threemutuallyperpendicular These axes will e labeled as the Y, and Z

2.4.1.l ParallelTo begin reviewsection, the simplest,andmostwellknowncase,consisting of two parallel conductors, where each conductor is carrying a

current, and where both currents are flowing in the same direction has beenchosen. Assuming,as it is shown in figure 2.11wo parallel conductors,Y ndY’ carrying the currents 4 and i,,.. It is also assumed that these currents areboth flowing n the same direction, also shownn 2.11. Taking a small portiondy’ of the conductor Y’, the current in that small portion of the conductor

produces a field everywhere n the space around t; therefore, the field intensityproduced by that elementf current on a point P located somewhere alonghesecondconductor Y can be representedbyavectorequal to dB (vectorquantities willbe represented by bold ace characters). Since vector is theresult of the cross product of two other vectors, and furthermore, since theresult of a vector product s itself a vector, whichs perpendicular to eachof the multiplied vectors; its direction, when the vectors are rotated in adirection that tends to make the first vector coincide with the second, will

follow the direction of the advance of a right handed screw. Applying theseconcepts it can be seen that the field intensity vector dB at point P must bealong a line parallel to the Z since that is the only way that it can be



perpendicular to dy’ and to r, both of which are contained in the X:Y plane.Rotating dy’ in a counter clockwise direction will make it coincide with thevector r. A right handed screw with its parallel to 2 would move in theupwards direction when so rotated; hen he field ntensity dB is directedupwards as shown in figure 2.11. The direction thathas ust been determinedfor the field intensity can alsoe verified usinghe right hand rule.

What hasjust been determined, for a particular element dy’, s applicable

to any other element along Y’; therefore, all of the small sections of theconductor Y’produce at point P field intensities that re all acted in the samedirection, as previously identifled. Furthermore what has been said about apoint P on conductor Y , is also applicable to any other point along Y andtherefore it be stated that the field along Y due to the current in Y’ iseverywhere parallel o the Z

To find the direction of the force that is acting upon the conductor Y,

simply apply Fleming’seft hand rule, when thiss done, it be seen that he

force is parallel to the X as it is shown in the corresponding figure.Thisresult is not surprising and it only confirms a principle which is already wellknown; that is, two parallel conductors, canying currents in the same directionattract each other.

If the direction of the current i,,. is reversed, the vector product will thenyield a field intensity vector dBhat is directed downwards and then Fleming’srule gives that the forces directed in such a wayas to move Y away fromY’.

Now, by eaving lowing in its originaldirection,whichwillkeep dB

pointing upwards, and reversing he direction of iyFleming’s rule indicates aforce directed so as to move Y away from Y’. Therefore, once again it isverified, the well known, fact that two parallel conductorscarrying currents inopposite directions repel each other.

2.4.1.2 Perpendicular Conductors a Plane

For the arrangement where there re two perpendicular connectorsin a plane,it is possible to haveour basic cases, Forall cases it w ill be assumed that the

conductors coincide withhe X and as shown in figure 2.12.Case I. For case number1, observing figure 2.12 (a), the field dB at point

P, due to the current in the segment dx, s parallel to the Y is so,

because of the result of he vector cross product, since bothx and r are in theX-2 plane. If dx is rotated to makeit coincide with the vector r, the directionof the rotation is such as to make a right handed screw, parallelo Y to moveto the ight. This determines the direction of the field,which is readilyconfirmed by he use of the right hand rule.

The above applies to any dx element, and therefore, the to ta l field B atpoint P coincides withdB. Since applies to any pointP on the Y then



' I F

X

Figure 2.12 Electromagnetic field andorceelationshipsorerpendicular

conductors as a functionof current direction.

the field at any point on Z, due to the current i, has the direction shown infigure 2.12 (a). Again, Fleming's left hand rule be used to determine thedirection of the force at any point along the conductorZ , he direction of theforce for this particular case s shown in figure 2.12 (a).

Case 2. For case number 2,f the direction of the current, is reversed, dBstill remains parallel to but now it points to he left, as shown on figure 2.12(b). The force F is then still parallel to the X but it is reversed from thedirection corresponding to case1 shown in 2.12 (a).

Case 3. If i,is reversed, leaving with its original direction,dB s identicalto the irst case butF is reversed as shown n 2.12 (c).



X

i

X

3.1111

- 1 I' TF

Figure 2.13 Graphical showing the direction of the electromagneticforce as a function of urrent direction

Case 4. Finally if both i, and i, are reversed the field dB is reversed in

comparison to case1but the direction f the force remains unchangeds itbe seen on figure .12 (d).

The four cases described above have one common feature, thats related tothe directionof the current flow. In every case the direction of the force onhe2 conductor is such as to try to rotate t about theorigin0 to make its current,i, coincide with he current

Thecurrent in Z also producesa orceon X. Its direction bedetermined by applying the principle just described. A graphical



X X’

Figure 2.14 Electromagnetic field and forces for a pair of perpendicular conductorsnot in the same plane.

showing the directions of the forces acting on tw o conductors at right anglesand in the same plane s given in figure.13 (a), (b), (c), and(a).

2.4.1.3 Perpendicular Conductors in th e SamePlane

This particular case is represented in figure 2.14 where it is shown that the

field dB at point P due to the current i, is in the Z-Y plane, since it must beperpendicular to dx being the vector product ofdx and r.Because of the cross

vector product relationship it must also be perpendicular to r. Now, a linethrough pointP on he Z-Y plane and perpendicular o r is also perpendicularto r’, which is the projection of r on the Z-Y, be done visually bylifting any of the acute angle comersf a drafting triangle).

T h e vectors like r joining P with all the elements dx’ along the conductor

X’ have a common projectionr’ on the Y-Z plane. Therefore the components

dB at P all coincide, and the total field intensity B has a directionas inthe figure.



Figure 2.15 Field intensity along theZ axis due to currents n theX’ xis.

Figure 2.16 Two dimensional view of the force direction. View looking intothe endof the Y axis.



The directions of the field intensities at other points along 2, due to thecurrent i, in X , have directions tangential to circlesrawnwith their centers at

as shown on figure .15. The direction for B is established by applying the right hand rule. At the

point onZ nearestX’ the field intensity o coincides in direction withZ. Thereis no force applied on the conductor at that point, since B does not have acomponent perpendicular to the conductor. At any other point along the

field B has a component perpendicular to the conductor, and thereforecapable of producing a force on the conductor. In figure 2.15, it can be seenthat above0, the nearest point to ’ on Z, the perpendicular componentf B

directed towardsX, and below0, away fromX.Putting t h i s information back into a three dimensional diagramhe direction ofthe forces be determined by applyinghe left hand rule. Figure .16 whichis drawn ooking into he end of the Y will serve to clarifyhe results.

The forces on the conductor Z are directed so that they tend to rotate theconductor about00’ so as to make , coincide in direction with,. If any of thecurrents are reversed principle still applies.

2.4.1.4Direction of Forces Summary

The following observations represent a simplified of the resultsobtained in the preceding section that are related to the determination of theforce direction between conductors thatre carrying current.

1. The preceding discussion covered parallel of conductors, as well asconductors hathave ectangularcorners,andconductorscross-overs,which may be found in common switchgear construction.

2. In the case of parallel arrangements, the conductors simply attractor repeleach other.In the case of perpendicular arrangements, the forces on the conductorssimply tend to flip over the conductors so as to make their respectivecurrents to coinciden direction.

Once it is known how a first conductor acts on second, the basic principleof action and reaction f forces, will define how the second conductorcts onthe first, it pushes back when being pushed, andt pulls when being pulled.

2.4.2 Calculationof Electrodynamic Forces Between Conductors

As it isknown, the mechanical forces acting n conductors are produced by heinteraction betweenhe currents and heir magnetic fields. The attractionorcebetween two parallel wires, where both re carrying a common current hat is

flowing in the same direction s used to define the unit of current, he ampere.According with th is defintion 1 ampere is to the current which will



produce an attraction force to 2x10' newtons per meter between twowires placed1meter apart.

Thecalculation of the electrodynamic orces, hat are acting, on theconductors is based on the law of Biot-Savart. According to h is law the force

be calculated by solvinghe following equation:

F Pd l xi2 - (4.xIO-") it

x

i2-=" = 2 x 10"1 ewton per meterx i 2I 2nd 27t x d d

where:

F = Force in newtonsI = length in metersd = distance between wiresn meters

il & i , = current in amperes= permeability constant= 4nxl0' (weber / amp-m)

From the above equation, androm the previous discussions,descriiing thedirection of the electromagnetic forces, it is seen that the force between twoconductors is proportional to the currents, il and 12, lowing hrough theconductors, to the permeabilityconstantand oanotherconstant that isdefined only by the geometrical arrangement of the conductors. As indicated

before, calculations or complex geometric onfgurations should be made withthe aid of any of he many computer models thatre available.

The total force acting on a conductor cane calculated as the summation ofthe component forces that are calculatedor each of the individual sections ormembers of theconductor.However heseforcecomponentsdonotexistindependently by themselves, since in order to have current flow a completeelectriccircuit is needed.Theuse of conductormembers in pairs is aconvenient way of calculation, but conducting mem ber in the circuit

must be taken in combination with each other membern the circuit.The force calculation for the bus arrangements that follow only represent

some of the most typical, and relatively simple cases that are found in theconstruction of switchgear equipment.

2.4.2.1Parallel Conductors

The general equation given byhe Biot-Savart law s directly applicable or thecalculation of he electromagnetic forces acting between parallel conductors,f

the conductors are round and they have an in f i i te length. The equation isapplicable if the ratio between the length of the conductor and the distance



10

1

0.1

0.M0 2 4 6 a 10 12 14

Figure 2.17 Multiplying factorA for calculating the distributed force along apair

short conductors in parallel.

separating the conductors is more than in which case the resulting error in

the calculated force is less than per cent, and therefore it is generallyacceptable ouse th is approximatevalue for makingestimates for theconductor strength requirements.

For conductors ofa finite length, the following relationship given y C.W.Frick [4] should be used.

i, x i ,d

A x 1

where:

The numerical values or the factorA be obtained from figure Inthe caption box enclosedn figure the relationships between the lengths,and of the conductors, and their spacing, is shown.



1

1.2

0.8

0.8

0.4

0.2

0

Figure 2.18 Correction factorK for f lat parallel conductors.

l

In the case of rectangular conductorshe force be determined using he

same formula thats used for round conductors, except hat a shape correctionfactor K is added to the original formula This correction factor takes intoaccount the width, thickness, and spacing of the conductors, and accounts forthe known fact, that the electromagnetic forces between conductors are notalways the same as those calculated under the assumption that the current isconcentrated at the center of the conductor. The values for K have beencalculated by H.B. Dwight [5] and are given in figure 2.18. For arrangementswhere the (d-h) / (h+b) > 2 (refer to figure 2.18) the error introduced is

not sufficiently ignificant and the correction factor maye omitted.2.4.2.2 Conductors Right Angles

In figure 2.19 a multiplying factorA is plotted for M e re n t lengths l of one ofthe conductors. The distributed force,er unit length, for the second conductorat a distance from the bend is calculated using he following formula:

- - 1 x 1 0 " x i , x i ~ x ~ newtons

Iwhere:



a

1

0.1

0.01

1 10

Y

100

Figure 2.19 Multiplying factor A used to calculate the distributed orcealongdifferent lengths of conductors at right angles to each

IA=asg iven inf igure2 .19= Y 1 +Y and

y and I are as shown in figure 2.19.The direction of the forces were previously determined in section 2.4 of

th is chapter. Butnow for convenience, the general rulewill be restated here,asfollows:

When in a simple bend he current flows from oneeg into the other leg ofthe bend, the force will act n a direction away from he bend. If the currentsflow into the bend from both sides or flow away from the bend from bothsides, the forces are directed towards he inside the bend, as if trying to make astraight line out of both conductors.

When one conductor oins another at right angles the force distribution ineach conductor can be calculated using the method given above. To find the

total effect of the electrodynam ic forces, theorce on each eg of the conductoris calculated and then the summation of these forces will yield the desiredresult.



100

1

1

Y

100

Figure 2.20 Values ofA’ for calculating the total force ata point on a conductorat

right angle.

To calculate the stresses on the individual parts it will be necessary tocalculate the force acting on a certain section of the conductor, and in mostcases it will also be required to find the moment produced by h is force withrespect to some given point. The total force acting on the conductor can becalculated using the same equation given for the distributed forces; however,

the multiplying factorA winow become AThis new factorA is then definedas:

where:l = l e n g t h of one of he conductor’s legy = l e n g t h of the other leg



4

3 . 5

2 . 5

2

1 . 5

0 . 5

0

Figure 2.21 Multiplying factorD for calculating the moment of the forceon a conductor with respect toanorigin point

yo= 0.779 x conductor radius (for round conductors) and0.224 x (a +b) or square bars where a and bre the sectiondimensions ofhe bar

”ypicalvalues for the multiplying factorA that have been calculated fordifferent lengths f conducton are shown in figure 2.20.

To calculate the moment of the total force with respect to the center ofmoments o, which in this case corresponds o the center point of he bend, thefollowing eqyationsusd.

M o = 1 x 1 0 - 7 x i , x i , x D

where:D = A multiplying factor whose valuegiven in figure 2.21

When the origin point of the desired moment not at the center of thebend, then the moment with respect to a different point be calculated byfirst calculating the moment M, with respect to the point which b e



regarded as the product of the force F and a certain perpendicular distance,wherep is related to then the moment of the same forceF, with respect o apoint at a distancep+n from will be

where n is the distance from to p , and therefore

M p = M o + F x n

Distance n is to be considered either positive or negative according to theposition of p with respect o the force and to the point that is; nis positivewhen it is to b e added to p and it is egative when t is to be subtracted from p.

2.4.3 Forces on Conductors Produced Three Phase CurrentsAs it has been shown before, when a fault occurs on a three phase circuit,because of their difference in phase, all three currents not be equallydisplaced,. It is also known that at least two of the currents must be displacedin relation to their normal axis, and in many instances all three phases maywell be displaced from he normal

Because of the interaction between ll conductors, the forces acting n each

of the conductors of a set of three parallel conductors, that are h a t e d in thesame plane and thatre equally spaced cane defined as follows:The force on conductor 1 is equal to the force due to conductor 2 plus he

force due o conductor 3,(F,= F1,2+

The force on conductor 2s equal to he force due to conductor 3 minusheforce due to conductor 1, (F 2 = Fz3+Fzl).

The force on conductor is equal to the force due to conductor 1plus theforce due o conductor 2, = F3,1 +F3,*).

When these forces re calculated, and heir mathematical maxima s found,assuming that the exhibit a phase sequence 1, 2, 3,(when looking atthe conductors from left to right), and that the current in phase 1 leads thecurrent in phase2, while the currentn phase 3 lags he current in phase 2, thefollowing generalized results are obtained:

The maximum orce on the outside conductorss to:

F,&Fqml = 12.9x lo" - newtons per meter( 9The maximum orce on the center conductor is:



Figure 2.22 Diagram representing the forces on a phase parallel conductor ar -rangement. Short circuit initiated on phase No.1 at 75(&er current zero in that phase).

.2= 13.9 IOd7($) newtons per meter

where:i, = rmsvalue of he symm etrical currentd = center to center distance fromhe middle to the outside conductors

Figures and show aplot of the maximum forces or a fault that isinitiated in the outsidepole,and for one nitiated in the centerpole,respectively.

It is interesting to note that or any of the three conductors he force reachesits maximum one half of a cycle after the short circuit is initiated. Also it

should be noted that the value of the maximum force s the same for either ofthe outside conductors; howeveror the force to reach that maximum,he shortcircuit must occur 75 electrical degrees before current zero in the conductorcarryingacurrent that is leading the current in the middleconductor, oralternatively it mustoccur75 lectricaldegrees @er the urrent erocorresponding to the phasewhich is lugging the current in the middleconductor.The maximum force onhe center phase (conductor will reach a maximumif the short circuit is initiated 45 electrical degrees either,before or @er, theoccurrencefurrentero in the middlehase. The maximum



ms

Figure 2.23 Diagram representing the forces on a phase parallel conductorarrangement.Short circuit initiatedon center phase (phaseN0.2) t 45(before current

in that phase.

force on the middle conductor representshe greatest of he forces on the threeconductors.

From all of this, another nteresting fact develops; if theshortcircuithappens 75 degrees uJer the current zero for the current in the conductor thatis lugging the current in the middle conductor, that instants 45 degrees before

the current zero in the middle conductor, therefore under these conditions herespective forces on both of the outside conductors will reach their maximumand th is maximum will be reached simultaneously on both phases one half

cycle after the initiation of the short circuit. Similarly if the short circuit isinitiated 75 degrees before current zero in phase 1 that instant corresponds o45 degrees uJer current zero in phase and therefore he respective forcesonphases 1 and 2 will reach their maximum and they will do so simultaneously.one half cycle fter the of the short circuit.

For a conductor arrangement wherehe spacing is such as in

an equilateral triangular arrangement, the maximum value of the resultantforce on any conductor will be:



The force on any of the conductors in geometry will reaches itsmaximum if the instant when the short circuit begins is 90 electrical degreeseither before or after the occurrence of the current zero in that conductor. Theforce reaches a m aximum value under those conditions, 180 electrical degreesor one half of a cycle after the initiation of the fault. The direction of themaximum force in case is perpendicular to the plane determined by theother two conductors, and he maximum force is directed away from hat

plane.

REFERENCES

1.

2.

3.

4.5.

6.

7.

8.

9.

ANSVIEEEC37.09-1979 TestProcedure for ac High-VoltageCircuitBreakers Ratedon a Symmetrical Current Basis.International Standard IEC 56 High Voltage Alternating Current Circuit

Breakers. Publication 56: 1987.C. L. Fortesque,Methodof ymmetricalcoordinatesappliedo thesolution of polyphase networks, AIEE Transactions, Vol 37 (part 11 ):

C. W. Frick, General Electric Review 36:232-242, May 1933.H. B. Dwight, Calculation of Magnetic Forceon Disconnection Switches,AIEE Transactions,Vol39: 1337,1920.E. W. Bohnne, The geometry of Arc Interruption, AIEE Transactions Vol

ANSVIEEE C37.010-1979 Application Guide or ac High Voltage CircuitBreakers Ratedon a Symmetrical Current Basis.J. L. Blackburn, Symm etrical Components for Power Systems Engineer-ing, Marcel Dekker, Inc. 1993.C. F. Wagner, R. D.Evans, Components, McGraw-HiU 1933.

1027-1140,1918.

60524-532, 1941.

10. M. H. Yuen, Short Circuit- ABC-Leam It an Use It Anywhere,Memorize No Formula, IEEETransactionsonIndustrialApplications:

11. B. Bridger Jr., All Amperes Are Not Created Equal: A Comparison ofCurrent Ratings of High-Voltage Breakers Rated According to ANSI andIEC Standards, IEEE Transactions n Industry ApplicationsVol29, No. 1:

195-201, Jan.-Feb. 1993.12. C. N. Hartman, Understanding IEEE Transactionson Industry

Applications Vol 1A-21No. 4: 842-848, Jul.-Aug. 1985.13. G. F. Corcoran, R. M. Kerchner, Alternating Current Circuits,John Wiley

& Sons. 3rd. Edition, 1955.14. D. Halliday, R. Resnick, Physics, Part I and 11, John Wiley & Sons,2nd.

Ed. 1967.

261-172, 1974.



3

TRANSIENT RECOVERY VOLTAGE

3.0 Introduction

At the beginningof Chapter it was stated that ll current interrupting devicesmust deal with current and voltage transients. Amonghe current transients, ofspecial interest, are those which are the direct result of sudden changes in theload impedance, such as in the case of a short circuit. Current transients pro-duced by a short circuit re dependent upon events that re part of the systemand therefore they are consideredas being transients that are induced by thesystem. The voltage transients, in the other hand, are the result of either the

initiation, or the interruption of current flow. These transients are initiated bythe switching device itself, and therefore they be considered as being tran-sients that are equipment induced. However, the characteristics of these tran-sients do not depend on the type of equipment, but rather, they depend uponthe parameters, and the specific location f each of the components of the cir-

cuit.What follows is an introduction to all important subject dealing with

voltage transients. Knowledge about he nature and the characteristicsf tran-

sient voltages, s an essential necessity or all those involved n the design, theapplication, and the testing of interrupting devices. Transient voltage condi-tions, especially those occurring following current interruption, muste prop-erly evaluated before selecting an interrupting device, whether it is a circuitbreaker, an automatic recloser, a fuse, a load breaking switch,r in general anykind of fault interrupting r load breaking equipment.

Whenever any of the just mentioned devices is applied, it is not sufficient

to consider, and to specify onlyhe most common system parameters suchas:

available fault current, fault impedance ratio(m),oad current level, systemoperating voltage, and dielectric withstand levels, butt is imperative that therequirements imposed byhe transient voltage e truly understood and properlyacknowledged to insure correct application ofhe selected switching device.

Voltage transients, as stated earlier, generally occur whenever a circuit is

being energized, or de-energized. In either case these transients be quitedamaging specially to transformers, reactors and rotating machinery that m aybe connected o the circuit. The transientsoccurringduringclearingofa

faulted circuit and which are referred here as, Transient Recovery Voltages(TRV),will be the first to be considered. In a later chapter other typesf volt-age transients such as those produced by switching surges, current chopping,restrikes and prestrikes wille covered.

77



CB

TRV-I

A A

Figure 3.1 Graphical representation of an Electric Network illustrating the sourcesthe Transient Recovery Voltage.

3.1 Transient Recovery Voltage: General Consideration

All types of circuit interrupting devices be considered as being a link thatis joining two electrical networks. On one side of the device there s the elec-trical network that is delivering power and which be identified as thesource side network. In the other side there is an electrical network that is

consuming power and consequently it can be identified as the load side net-work, as is illustrated in figure 3.1.

Whenever the interrupting device s opened, the two networks are discon-

nected and each f the networks proceeds to redistributets trapped energy. Asa result of this energy redistribution, each network will develop a voltage thappears simultaneously at the respective terminals of the interrupter. The al-

gebraic sum of these two voltages then represents the Transient RecoveryVoltage, which normally s simply referredas "RV.

A comprehensive evaluation ofhe recovery voltage phenomena hat takesplace in any electrical system should e based upon the conditions prevailingat the moment of the interruption of a short circuit current. As minimum re-

quirements to be taken into consideration for evaluation are: the type ofthe fault, the characteristics of the network connections, and he switching ar-rangement used.



Depending upon the different combinations of these conditions, it is obvi-that the transient recovery voltage can have any different characteristics;

it can exhibit a single frequency,or a multi-frequency response. It can be ex-pressed in the form of a sinusoidal function, a hyperbolic function, a triangularfunction, an exponential function, r asa combinationof these functions; it alldepends as it has been said, upon he particular combination of he many fac-tors which directly influencehe characteristics of the TRV.

If all factors are taken into consideration exact calculations ofhe TRV incomplex systems is rather complicated, and generally these calculations arebest made with he aid of a digital computer program, suchs the widely usedElectro-Magnetic Transients Programr EMTP.

For those applications where a somewhat less accurate result will suffice,E.Boehne [l], Greenwood [2] andP.Hammarlund amongothers,have shown that it is possible to simplify the calculations by reducing theoriginal system circuits to an equivalent circuit which has a simple mathe

cal solution. Nevertheless when these simplifed calculation method is em-ployed the problem of how to properly select the equivalent circuits and thevalues of the constants to e used in the calculations still remains.

The selections are only practical equivalents containing lumped compo-nents that approximately describe the way in which the actual distributed ca-pacitances and inductancesre interrelated in the particular system under con-sideration. Furthermore the calculation procedures still are somewhat tedious;which again points outhe fact that, even or moderately complex systems,t is

advantageous to usehe modern computer aided methodsf calculation.However, there is something to be said about simple methods or approxi-

mated calculations of TRV, andt the risk of oversimplifying the problem, it ispossible to say that n the majority of the cases a first hand approximation ofthe TRV is generally all that is needed for the proper initial selection, andorjudging the adequacy of prospective applicationsof circuit breakers. For someparticular cases t is possible to considerust the most basic conditions found inthe most common applications. In most cases these are the conditions thathave been used s the basis for establishing standards that definehe minimumcapability requirements f circuit breakers.

A simplified calculation approach can alsoe of help in determining if therated TRV of a circuit breaker is sufficient for the application at hand and inmany cases the results obtained, with such simplified calculation, can e usedto determine if there is a need for further more accurate calculations. Anotherpossible application of the simplified calculation approach is that it can beused to evaluate possible corrective actions thatay be taken to match the ca-pability of the device with he characteristics of the circuit. One corrective ac-tion is the additionof surge capacitors o modify the inherent TRV f a system.



1.

2.

4.

5.

3.1.1 Basic Assumptions or TRV Calculations

The following assumptionsare generally made when calculating the transient re-covery voltage oftransmission,or a distribution high voltage power

Only three phase, symmetrical, ungrounded terminal faults, need to beconsidered, is because the most severe TRV appears across the firstpole that clears an ungrounded three phase fault occurringt the terminals

of the circuit breaker,The fault is assumed to be fed through a transformer, whichn is be-ing fed by an infinite source. This implies that a fault at the load sideterminals of a circuit breaker allows he full rated short circuit current toflow through he circuit breaker.The current flowingn the circuit is a totally reactive symmetrical currentwhich means thatat the instant when the current reaches its zero, the sys-

tem voltage wille at its peak.

The voltage a m s s the circuit breaker contacts, as the current approacheszero, is equal to the arc voltage of the device and t is assumed to be neg-ligible during the TRV calculation, since the arc voltage, when dealingwith high voltage circuit breakers, represents only a small radion of thesystem voltage. H owever this may not be the case for low voltage circuitbreakers where the arc voltage, in many instances, represents a significapercentage of the system voltage.The recovery voltage ate represent the inherent TRVof the circuit, and itdoes not include anyf the effects that the circuit breaker itself may haveupon the recovery voltage.

3.1.2 Cu rrent Injection Technique

A convenient contrivance employed or the calculation of the "RV s the in-troduction of the current injection technique. What this entitless the assump-tion that a current equal and opposite tohe short circuit current, which wouldhave continued to flow in the event that interruption had not occurred, isflowing at the precise instant of the current zero where he interruption of theshort circuit current takes place. Sincehe currents, at any time, are equal andopposite, it is rather obvious that the resultant value of the sum of these two

currents is zero. Consequently the most basic condition required for currentinterruption is not being violated.

Furthermore it is possible to assume thathe recovery voltage exists onlysa consequence of current, which is acting upon the impedancef the sys-

tem when viewed fromhe terminals ofhe circuit breaker.Additionally, since the frequency of the TRV wave is much higher than

that the power frequency, it is possible to assume, without introducing any



L L L L

Figure 3.2 Schematic representation f the elementsof a Transmission Line.

significant error, that the injected currenti) be represented by a linearrent ramp s defined by:

where:

I,, = rms value of the short circuit currento= d =

t = time in seconds

As it will be seen later concept will be used extensively for the calcu-

lation of Transient Recovery Voltages.

3.1.3 Traveling Wavesand the Lattice DiagramTo better understand someof the important characteristics relatedo thesient voltage phenomena taking place duringhe execution of switching opera-tions involvinghigh voltage equipment, it would be beneficial to have t leasta basic knowledge about he physical nature and behaviorof traveling wavesthat are present on the transmission linesuringthese times.

One important characteristic of transmission liness that since their resis-tance is generally neglected, they can be represented as being made up of acombination of distributed inductive and capacitive elements. The inductiveelements are all connected in series, and the capacitive elementsre distributed



along the line in parallel as is shown in figure When an electrical systemis visualized in his fashion, it can be seen that f a voltage s applied to he endof the line, the first capacitor willbe charged immediately, andhe charging ofthe capacitors located downstream from the point where the voltage was ini-

tially applied will be sequentially delayed as a consequence of the inductors,that are connected in series between the capacitors. The observed delay willbe proportionally longer t each point down the line.

If the applied voltage s in the form of a surge signal that at zero andthat to zero in a short time, thent is reasonable to expect hat the volt-age across the capacitors will reach a maximum value before returningo zero.As this pattern is repeated, at each capacitor junction point along the line, itcan easily be visualized that the process semes as a vehicle to propagate theapplied surge in the form of a wave which moves along the line. During thepropagation of the wave the original characteristics of the surge signal remainbasically unchanged in terms of their amplitude and waveform.

Since, in order to charge he capacitors, at each connection point alongheline, a current must flow through the inductances that are connecting the ca-pacitors; then, at any point along he line, the instantaneous value of the volt-age e(t) willbe related to the instantaneous valuef the current i(t) by he fol-lowing relationship:

e(t) = Zi(t)

Where the constant of proportionality Z represents the surge impedancefthe line, which s given by:

In the above relationship L and C are the inductance and the capacitance,

per unit length, of the line. The numerical value ofhe surge impedanceZ is aconstant in the range of 300 to 500 ohms. A value of 450 ohms is usually as-

sumed for single overhead transmission conductors and ohms for bundledconductors.

Dimensionally, the surge impedance is given in ohms, and is in the natureof a pure resistance; however, it is important to realize that the surge imped-ance, although it is resistive in nature, t can not dissipate energyas a normalresistive element It is also important o note that he surge impedanceof a

line is independent of the length of such line, is so because any point lo-cated at my distant part of a circuit can noth o w that a voltage has been applied somewhere inhe line until a traveling wave reaches that point.



Traveling waves, as is the case with any other electromagnetic disturbancein will propagateat the speed of light, that is 300 meters per microsecond,or approximately 1000 feet per microsecond. the wave passes from a linethat has an impedance to Z1nto another circuit element, possibly but nnecessarily another line, which has an impedancequal to Z2,ew waves willpropagate from the junction point, traveling back into Z1,nd through thejunction intoZ2. he new waves are shaped identically as the incident wave,

but their amplitude and possibly their signsre changed.The coefficients usedo obtain the new voltage waves are:

Reflection (fromZ2 ack intoZ,):

Refraction (fromZ1nto Z2):

If the line termination is a short circuit, thenZ2 and the above equation

becomes:

For a reflected wave: KM = -1

For a refracted wave: Km = 0

If the line end is an open circuit, thenZ2 and the expressions are:

For a reflected wave: KRo=

+l

For a refiacted wave: Km = +2

The back and forward moving wavespass each other along theline, and the potential at any point along such line isbtainedby adding the poten-tials of all the waves passing hrough he point initherdirection.

With the aid,o f a lattice diagram, figure it is possible to keep track of

all waves passing through a given pointt a given moment.A lattice diagram,can be constructed by drawing a horizontal ine from “a” to “ b which repre-



cca,,

4T 2ndFORWARD REFLECTION

Figure Typical construction f a Lattice Diagram.

sents, without any scale, the length of the transmission line. Elapsed time isrepresented in a vertical coordinate thats drawn downwards from the abscissa,this time is given by the parameterT which symbolizes the time required bythe wave to travel from one endof the line to the other end. The progress fthe incident wave and ofts multiple reflections is then trackeds shown with

their corresponding labels by the zigzag linesn figureThe next step is to determine the relative amplitudes of the successive re-

flections. It be seen that an incident wave, whether it is a current or a

voltage wave, whichs entering point “a” from the left isfunction of time f(tand it progresses undisturbed to point“b”where thefirst back reflection takesplace. This first back reflection is equal toKR$(f), where KRb s the same co-efficient that was earlier defined using the values of Z on both sides of “b”.

When this wave reaches point “a7’, the reflection back towards point “b” isobtained by using the coefficient KRawhich also was defined earlier, but,time the valuesof 2 on both sides of “a” are used. The refraction beyond thpoint “a” is calculatedy the coefficientKTa which is evaluated using the samerelationship given earlier for evaluating the coefficient KT. The process is re-peated for successive reflections, and the amplitudef each successive wavesexpressed in terms of these coefficients. The above coefficients be substi-



tuted by the corresponding numerical values defined below and the values canthen be used to obtain the actual amplitudef the wave.

For a line terminating on a shorted end: KRb= -1 ;Kn =

For a line terminating on an opened end:Ra= + l ;CTa =

3.2 Calculation of Transient Recovery VoltagesFor any high voltage transmission r distribution network, t is customary andalso rather convenient to identify and to group the type of short circuits or

faults as either, terminal,or bolted faults ands short line faults.A terminal fault s one where he short circuit takes lace at, or very near,

the’terminals of the circuit breaker, while a shortine fault (SLT) is one wherethe short circuit occurs t a relatively short distance downstream fromhe cir-

cuit breaker onts load side.Depending upon he characteristics of the network andhe type of he fault,the typical TRV can be represented by either single frequency,or double fre-quency waveforms for the terminal faults, and by a multi-frequency, hat in-cludes a sawtoothed waveform componentor the short line ault.

It is also important o recognize that the type of the fault has an importantsignificance on the performance recovery of the circuit breaker. Following a

short line fault the circuit breaker is more likely to fail in what is called the

thermal recovery region. Which is a region that comprises approximatelyhefirst ten microseconds following the interruption of the current, and wherethermal equilibrium has not yet been re-established.In figures 3.4 (a) and (b)

the oscillograms of a successful and an unsuccessful interruption are includeFor a terminal ault it is more likely thatif any failures to interrupt do oc-they will be in the dielectric recovery region, whichs the region located

between a range from approximately 20 microseconds up to about 1 millisec-ond depending on he rating of the circuit breaker. In figure (a) a dielectric

region failure s shown while in figure (b) the recovery voltage correspond-ing to a successlid interruptions shown.

3.2.1 Single Frequency Recovery Voltage

A single frequency TRV ensues when during the transient period the electricenergy is redistributed among a single capacitive and a single inductive ele-ment. In general condition is met when the short circuit s fed by a trans-former and when no additional transmission lines remain connectedt the bus

following the interruption f the short circuit. This condition generally occursonly in distribution systemsat voltages lower than 2.5 kV, where in the ma-



I

Figure 3.4 Transient Recovery Voltagen the Thermal Recovery Region.Oscillographic traces of a Line Fault. (a) Successful interruption.(b) Dielectricfailure after approximately



mFigure 3.5 Transient Recovery Voltagen the Dielectric Recovery Region.Oscillographic traces of a Terminal Fault (a) Successful interruption (b) Dielectricfailure after approximately 60



jority of cases the fault current is supplied by step-down transformers andwhere becausehe characteristics of the lines, thatare connected to he bus, aresuch that when considering he transient response f the circuit theyare betterrepresented by their capacitance rather than by their surge impedance. aconsequence of condition the circuit becomes underdamped, and it pro-duces a response which exhibits a typical one minus cosine waveform,

The simplest circuit that serves as an illustration for the single frequency"RV response is shown in figures 3.6 (a) and (b). Referring to h is figure, andafter opening the switch, the following very basic equation be written todescribe the responsef the circuit shown n figure

= L - + - j i d ti 1

dt C

where the initialvalues are:

1,= 0 since is a basic requirementor interrupting he current, and

V, = 0 since it was chosen o disregard the value of the arc voltage.

Rewriting the above equation,using the Laplace we obtaim

where S represents theLaplace transformoperator.Solving for the current

Since the TRV s equal to the voltage a m s s the capacitor, voltagewhen shown in the Laplace notations equal to:

Substituting the value of I(s) in the above equation and collecting terms:



Gen. L CB

LCB

Figure 3.6 Typical simplest circuit which producessingle frequency response.

Letting\I'=aoLC

and then obtaining the inverse transform, the followingequation s obtained:

TRV=-v c o s a t - c o s o o t

LC [ ag - 02 1



TRV=V(1-coso0t)

where:

V= 1.88EmM

The value 1.88 s used as a constant following he recommendations made,for the purpose of standardization, by the Association of Edison IlluminatingCompanies [4]. This recommendation is based on the fact that at the time ofcurrent zero, on an ungrounded three phase terminal fault, the voltage at thesource terminal f the breaker is equal to 1.0 per unit while the voltage onheload side terminal s equal to0.5 per unit so the net steady state voltage m s sthe circuit breaker is equal to 1.5 per unit. However during the transient pe-riod, if the effects of any damping are neglected, he voltage can oscillateo amaximum amplitude equal to.0 per unit, and therefore it woulde reasonableto say that in any practical application he maximum peak of the TRV a m s sthe fm t pole to interrupt a three phase short circuit current could e between1.5 to per unit. When regulation and damping factors, which have beenobtained from digital studies, and which have subsequently been verified byfield tests were factored the final recommended valuef 1.88 was chosen.

3.2.2 General Caseof Double Frequency Recovery Voltage

In the majority of the actual system circuits that are connected to theterminals of a circuit breaker be represented by relatively simple equiva-lentcircuitscomposed of umpedcapacitiveand nductiveelements.Thesubstitution of the distributed capacitance and reactance of transformers andgenerators makes it possibleo convert complex circuits into simple oscillatocircuits, which may be easier to handle mathematically. One such system cir-cuit, which s often found in practices shown in figure3.7 (a), and its simpli-fied version in (b).

As it can be recognized, finding he response of circuit is not a difficulttask since the two frequencies are not coupled together, and in fact, they aretotally independent of each other. The solution of circuit is given by thefollowing relationships that define each one of the two independent fi-equen-cies; their resulting waveform whichs obtained as the summationof the wave-forms that are generated by each independent frequency, andhe total voltageamplitude for the recovery voltage.

The frequencies re given by:



(b)

Figure 3.7 Schematic of simple double frequency circuitused as comparison basis forthe calculationof other simplified circuits.

The magnitude and waveorm for the total voltage s proportional to the induc-tances, and is givenby:

=V[aL(l - cosaLt)+ a$ - cosQ st)]

where:



1=-

The above equation, describinghe Transient Recovery Voltageof the cir-cuity is applicable only during the few hundred microseconds followingthe interruption of the currenty until the power frequency source voltage beto change by more than a few percent from its peak value. h ddition, theequation is accurate only or purely inductive circuits, sincehiswas one of heoriginal assumptions. If the power factor of the power source is such that thephase angle between the current and the voltages not exactly 90(a more exacexpression could e used as shown below.

where:

= 377 for a 60Hz ower frequency

3.2.2.1 CircuitSimpli j icdon

The aim of he circuit simplification processs to obtain representative circuitfrom which relationships can be established relating the inductances and thecapacitances of these circuits with those the model circuit described above.For example he relatively complicated network scheme thats shown in figure

(a) can be simplified as shown in figure (b). This simplified circuitthen be mathematically related to the circuit of figure 3.7 by using thefollowing relationships [2],[3].



Figure 3.8 Example of circuit simplification.

For thefrequency relationships

For the amplitude relationships

( a + b - a 2 ) ( a + b )

aL =



a, = l-a,

where:

a=--(1+p+a2)

To illustrate the procedure the following numerical example is given bysolving the above equations withhe following values for the variouscircuit elements:

LA= 1.59mH = source inductanceLB

=mH

= reactor inductance= 0.01 pF = capacitance of all equipment on he busCB 500pF = capacitance of breaker and reactor

The resultantTRV is shown graphically in figure 3.9

3.2.2.2 Circuit Simplification Procedures

Recognizing that absolute guidelines for sim pl ifjh g a circuit are not quite

possible because one of the major difiiculties in the procedure is the properchoice of those circuit components which have a significant influence on thetransient phenomena under consideration.



" ~ M b m w ~ o 9 c h O

TIMEmicro-seconds

Figure 3.9 ResultantTransientRecovery Voltage from circuit in figure

This selection in many instances is solely based on experience that has beenacquired through practice. Despite the limitations, a few generalized rules,that are designedo facilitate the circuit reduction task, cantill be provided.

The initial circuit is constructed, at least initially, with all the principalcomponents, such as, generators, cables, reactors, transformers and circuitbreakers.Starting at the location of the fault and going towards the circuit power

source, choose a point where the systems fairly stable. Denote pointas an infinite source, r a generator withero mpedance.The distributed capacitances of generators and transformers are shown aslumped capacitances. The capacitance of each phase of the generatorbe substituted by one halff its total capacitance.The inductanceof each transformer n the original circuits replaced by a

type circuit. One half of the total capacitance to ground of the phasewinding should be connected at each end of the transformer coil, which

corresponds to the leakage inductance of the transformer. The capaci-tances to groundf both windings muste considered.



5.

6.

7.

8.

9.

Whenever two or more capacitances are located in close proximity to eaother and are joined by a relatively low impedance, in comparison toherest of the circuit, the capacitances may be combined ntoasingleequivalent capacitanceThe value of inductance is calculated from the total parallel reactance ofall of the reactances that are connectedo the bus barsof the transformers,generators and reactors coils, with the exception of the reactance of the

faulted feeder. If cables are used and if they are very short their induc-tance willbe negligible comparedwith that of the generator andhe trans-former, and t can be ignored.A capacitance representing he of one half of the faulted phase reac-tor's capacitance to ground, plus one half of the circuit breaker capaci-tance to ground and the total capacitance to ground of the connectingfeeder is connected at the breaker terminals.Show the total capacitance to ground of the cables in the particular phase

under consideration. Generally this capacitance is much larger than allothers, unless the cables are very short, in which case a capacitance hatincludes all connected branches and the equivalent capacitancesf the re-actors, transformers, and generators is considered. This capacitance is al-ways equal to one half of the total capacitance of the circuit componeAt the connection point of the generator, the cables and the transformer,the individual capacitances are combined into a single equivalent capaci-tance.

10. If the transformer is unloaded, the magnetizing inductance of the trans-former is much greater than thatf the generator, and thereforehe genera-tor's inductance be ignored.

11. If motors are included in the circuit and they are located remote to fault,their impedances are shown as the single equivalent impedanceof all themotors in parallel

3.2.2.3. Three Phase to Ground Fault a Grounded System

In a three phase system, when the neutral of the system is solidly grounded,and a three phase fault to ground occurs, each phase will oscillate independ-ently it is therefore possible to calculate the response of the circuit using thesolution that was previously obtainedor a simplified single phase circuit haing a configuration s it was shown in figure 3 .8 (b).

3.2.2.4 Three Phase IsolatedFault a Grounded

As it is already known, the worst case"RV s always observedn thefirstphase to

clear the fault.In the event of an isolated phase fault occurringwithin a sol-

idlygrounded the nfluenceexerted n he ecoveryvoltageby he



Figure 3.10 Schematic diagramof the equivalent circuit for a phase isolated faultin a grounded system.

other two phases must e taken into consideration.In figure 3.10 a representative three phase circuit is shown.. What is sig-

nificant in this new circuit is the addition of the capacitance shown as C ,which is equal to he total capacitance

to ground from the location the load inductance to the fault location,capacitance includes one half f the capacitance to ground of any reactorspresent in the In figure 3.10 phase A is assumed to be the firstphaseto clear the fault, and therefore t is shown as being open; while phasesB andC are assumed to still be going through he process of interrupting the currentclosed and therefore they can be assumed to be still be closed. As it can beseen the current flowing through phase A, in the direction the arrow, willretum through the parallel paths of phases B and C, and therefore these two

phases canbe represented by a single ath having one half he inductance andtwice the capacitance to ground of the original phases. This is shown in thecircuit of figure 3.11 (a).



LG LrCB

LR

(b)

Figure 3.11Circuit simplification from riginalcircuit on figure

The circuit can be even further simplified as shown in figure (b),

where it can be observed that the portionof the diagram, to the left hand sideof the switch, is composed of two oscillatory circuits which similar to oneof the circuits for which a solution has been provided.



TRANSMISSIONLINES

FA

U

LT

Figure 3.12 system configurationused as basis for defming TRV ratings ofhigh voltage transmissionclass circuit breakers.

The right hand side of the diagram is made up of four oscillatory circuits,which makes the calculation portion of the circuit rather difficult. As arule, further simplification generally be achieved by omitting he smallestinductances and,or capacitances andby assuming that such omission will havea negligible effect n the amplitude of the oscillation.

3.2.3 Particular Case of Double Frequency RecoveryVoltage

From more practical point f view of circuit breaker application,t is ratherusefid to evaluate the transient recovery voltage that appears on a typicaltransmission circuit which is commonly found in actual high voltage powersystem applications. typical circuit is the one that has been used for es-tablishing standard basis f ratings for circuit breakers, the circuit is shown infigure 3.12. As it be seen in the figure, one the characteristics of thiscircuit is that the fault is fed by a source consistingf aparallel combinationofone or more transformers nd oneor more transmission lines.

It isposslile to reduce this original circuit o a simpler circuit consisting fa parallel combination f resistive, inductive and capacitive elements.



!QUIVALENT SYSTEMLINES SURGE REACTANCEIMPEDANCE Ls ETRV

PFigure 3.13 Equivalent circuitfor the transmission system configuration shown in fig-ure

In such a circuit the inductanceL is the leakage reactanceof the transformer,and the capacitance C corresponds to the total stray capacitance of the instal-lation. The resistance, in case, represents the total surge impedance ofthe transmission lines and is equal to the individual surge impedance ofeach line divided by he “n” numberof lines interconnectedn the

In the majority of the applications, the parallel resistance the surge im-

pedance of the lines is such that it effectively s w a m p s the capacitance of thecircuit, and therefore,it is a com mon practice o neglect the capacitance. Theresulting equivalent circuit, which is shown in figure is then one thatconsists of a parallel combination of only inductanceL)and resistance ZJ.

The operational impedance, or type of circuit, is given by the follow-ing expression:

Now using the injection current technique an expression is obtained for thevoltage that appears across the just found circuit impedance. The resultingvoltage, which happenso be the TRV the circuit is given by:



The solutionof equation in he time domain giveshe following result:

V @ )= (Ims)OL

where:

Qt)=hv he exponential componentf the total response.

Since the voltage for the first pole to clear is equal to 1.5 times the maxi-mum system voltage hen the corresponding transienl recovery voltageor thisportion of the total envelope is:

where Emted) s equal to the rated maximum voltage ofhe device.

This voltage, which initially appears across the f im 'phas e that clears the

fault, also appears in the form of a traveling wave; beginning at the bus andtraveling down along eachf the connected transmission lines.As it is alreadyknown, the first reflection of the traveling wave takesplace as the result of adiscontinuity in the line, from wherehe traveling wave s reflected back tohebreaker terminal, where the traveling wave voltages added to thenitial expo-nen tkl voltage wave.

Using the attice diagram method, previously discussedn section thevalue of the reflected wave is determined tobe equal to the product of the co-

efficient of reflection KRb,which is determined by the line termination, timesthe peak valueof the El mv envelope Once again, and according tohe travel-ing wave theory, after the reflected wave arriveso the point where the fault islocated it encounters a terminating impedance andts effective value becomesequal to the product of the reflected wave times the coefficient for the re-fracted wave or KRb KTa E ,o. When simple reflection coefficients,equal to minus one for a shorted line and plus one for an open line, are usedand if the line terminal impedances are resistive in nature; thenhe coefficients

become simply amplitude multipliers.The TRV calculation can then be reduced to first finding the initial expo-nential responseE , and then adding, t a time equal to microseconds



TIME microseconds

Figure 3.14 Typical TransientRecovery Voltage corresponding to the circuit shown infigure

per mile, a voltage equal to KRa KTa E, mv. A typical waveform is illus-trated in figure 3.14.

Primarily, to facilitate the testing of circuit breakers that are subjected toparticular orm of TRV, TheAmericanNationalStandards nstitute

has chosen a composite EX-COS model waveform, see figure 3.15,which approximate the actual transient recovery voltage waveform. The iniexponential component f the response is calculated in the same way s before.

The equation describing the-cosine portion of he response is written as:

E , = 1.76 Eekd )(l - Q of)



Figure3.15 Equivalent wave form for esting TRV requirements of circuit breakers.

where:

oo -

t2

The 1.76 multiplier is specified by the American National Standards nsti-

tute following the recommendations madeby the Association of Edi-son Illuminating Com panies [4]. multiplier is the product of an statisti-

cally collected value of 1.51 times an assumed damping factor of 0.95 andtimes the now familiar 1.5 factoror the ftrst pole to clear the fault.

The peak value of the voltages by [5] as the requim

m a t for transmission class circuit breakers, (which the class of circuitbreakersabove 72.5 kv), the specify T2 s the timemmimsecondsre-

quiredtoreach the peak of the voltage. Both numerical valuesreas a fimction of the specific rating of the circuit breaker5 ] .

3.2.3.1

Initial Rateof Rise

The initial slope or initial rate of rise of the T R V ,which is also specifiedn thestandards documents s an important parameter hat defines the circuitbreaker



capabilities for it represents one ofhe limiting values or the recovery voltage.The initial slope is obtained by taking the first derivative of the exponentialcomponent of the recovery voltage waveform.

Substituting t = 0 into the equation the followings obtained,

InitialRate = R o = 1.5 &'(ImsoZn)

Inspection of the above equation suggests that the initial TRV rate is di-rectly proportional o the fault current interrupted and inversely proportionalothe number of transmission lines that remain connected to the bus. In a sub-transmission class system the feeders are generally in a radial configurationand therefore the fault current does not depend uponhe number of connectedlines; thus, as the numberof lines decreases he fault current remains constant.In the case of transmission class systems,as the number of lines is decreased,so does the fault current.

3.2.4 Short Line FaultRecovery Voltage

A short line fault is a short circuit condition that occurs a short distance awa

from the load side terminals of a circuit breaker. This short distance is notprecisely defined, butt is generally thought tobe in the range of several hun-dred meters up to about a couple kilometers. What makes type of faultsignificant is the fact that, as it is generally recognized throughouthe industry,it imposes the most severe voltage recovery conditions upon a circuit brea

The difficulties arise because the line side recovery voltage appears as asawtooth wave and thereforehe instantaneous, and rather steep initial rampfvoltage imposes severe stresses on he gap of an interrupter before it has hadenough time to recoverts dielectric withstand capability.

When dealing with the recovery voltage thats due to a short line fault it musbe realized that when fault occurs at some finite distancerom the terminals of tprotective device, is always a certain amount of line impedance involved.This line impedance reduces, to some extent, the faulturrent,but it to

some of the system voltage. Theurtheraway from the terminals the faulocated, the greater the fraction of the voltage sustained by the lineacross the load terminals of the circuit breaker.

Consequently, since an unbounded charge can not remain static, then fol-lowing the interruption of the short circuit current the voltage trappedalong the line will begin o re-distribute itself inhe form of a traveling wave.



cca,,

K,K,e(t) = - e(t)

-- K,K2,e(t) =+ e(t)

4T

-&e(t) = 0

<

evaluate the characteristics of the traveling wave it will be rather con-venient to use a lattice diagram, figure 3.16, where the numerical valuesf thecoefficients, for the reflected and transmitted waves are used to shown theamplitude multiplieror the voltage wave s it travels back and forth alonghe

line.fmally determine he waveform of the line side voltage a simpleraphi-

method, shown in figure 3.17, can be used. In graph, the horizontalscale represents the time elapsed since the interruption of the ament, and isdivided in time intervals which are multiples of T; where T is the one waytraveling time from the breaker to the location of the fault. On the vertical

representing volts at the breaker terminal, the divisions which are set atvalues corresponding to = ZmR.

When the values from theattice diagram in figure 3.16 are transferred intofigure 3.17, one can notice that theirst rising rampof voltage is a straight linestarting at the origin and passing through the locus points= 2T and e =2E.



2=2nd. Reflected Wave3=3rd. Reflectedave4= 4th. Reflected Wave5= 5th. Reflected Wave

-.

‘.8.

16 - Resultant Sawtooth Wave-10 ‘.

ELAPSED TRAVEL TIME

Figure 3.17 Resultant composite wave for short fault produced by the traveling wavephenomena.

line has the slope Z d hich corresponds to the voltage that is sus-

tained at the breaker terminalsntil the first reflection returns.At t = 2T another wave from the open end, which according to the

lattice diagram of figure 3.16 has a voltage value equalo -2e, = This

wave has a zero alue at t = 2T and a slopeof -2 Zd , which is double that ofthe preceding ramp andn the opposite or negative direction. ramp is rep-resented by he line drawn rom the coordinates T, and 4T, 4E.

The lattice diagram further shows that a third wave at t = 4T with apositive double slope. Adding up the ordinates of the successive weobtain the expected typical sawtooth wave which characterizes the load siderecovery voltageof a short ine fault.

The fiequency of the response is dependent upon the distance to the fault

and to the travel time of the wave. The voltage amplitude is also dependentupon the distance to the fault and onhe magnitude of the current.



Specific equations describing the rate of rise and the amplitude of the saw-tooth wave are given by:

RL = F I m lod k V lp and

e = d I f i ( O . 5 8 V ) kV

where:

d = an amplitude factor, generally givens 1.6

The time required to reach the voltage peaks given by:

The total Transient Recovery Voltage s to the of the load sidetransient voltage associated with the traveling wave plus, either the- cosineor the exponential-cosine waveform representing the source side componentthe TRV. The choice of the source side waveform depends n the type oftem under consideration ands calculated in accordance to the guidelines pre-

sented previously.3.2.5 Initial Transient Recovery VoltageITRV)

The term Initial Transient Recovery Voltage (ITRV) refers to the conditionwhere during thefirst microsecond, or so following a current interruption, herecovery voltage is influenced by the proximityf the system component con-nections to the circuit breaker. Among those components re the buses, isola-tors, measuring transformers, capacitors, etc.

Following interruption a voltage oscillations produced which is similar tothat of the short line fault, but, this new oscillation has a lower voltage peakmagnitude, but the time to crest has a shorter duration due to the close disbetween the circuit breaker and the system components.

The traveling wave will move wave will move down theus up to the pointwhere the first discontinuity is found. In an IEC report [6], the frrst disconti-nuity is ident5ed as being the point where a bus bar branchesff,or the pointwhere a capacitor oft least one nanofarads connected.

The following expression for the ITRVs given in reference [12].



where

= Surge impedance= 260 OhmsTi=Wave travel timen microsecondsI = Fault currentkAa= nf

The above expression shows that:

1. The first peak of the ITRV appears at a time to twice the travelingtime, of the voltage wave, from the circuit breaker terminals to the firstline discontinuity.

2. The initial slope of the ITRV depends only on the surge impedance of thebus and the ate of change of current (W dt at I=O).

The above statements suggest that since the travelim e is a function of thephysical locationof the component, it is practically impossible to define a gen

form of ITRV. One can expect that there would be as many ITRV varia-tions as therere station layouts.

Nevertheless, representative values for various voltage installations havebeen established[7]. These values are tabulated n table 3.1.

TABLE .1

Rated Maximum

Voltage kV rms

1.1.0.8.6.5.4.3oltagePeakTi(s)

Time to First

800 562 242 169 145

~ ~ ~~~ ~~

The second item, deals with the rate of change of current. It suggests that

if the slopeof the current s modified by the action of the breaker during intruption then t can be expected that the ITRV would alsoe modified.

All th is indicates that at least in theory the ITRV exists. However in prac-tice there are those who question their existence saying that the ideal breakerassumed for the calculations does not exist.

It appears thatt is better to say that thereay be some circuit breakers thamay be more sensitive than others to the ITRV. More sensitive breakers arethose that characteristically produce a low arc voltage and that have negli

or no post arc current, in other words, an ideal circuit breaker.



REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

E. W.Boehne, TheDetermination of CircuitRecoveryRates,AIEETransactions, Vol54; 530-539. 1935.Allan Greenwood, Electrical Transientsn Power Systems, John Wiley andSons Inc., 1971.P. Hammarlund, Transient Recovery Voltage Subsequent to Short Circuit

Interruption,Proc.RoyalSwedishAcademyEngineeringSci.No.189,1946.Transient Recovery Voltage on Power Systems, Association of Edisonl-luminating Companies,New York, 1963.

C37.06-1979, Preferred Ratings and Related Required Capabilitiesfor ac High Voltage Circuit Breakers Ratedn a Symmetrical Current Ba-sis.IEC 56-1987 High-voltage alternating-current circuit-breakers.

ANSL/IEEE C37.04-1979 Standard Rating Structure for AC High-VoltageCircuit Breakers Rated on a Symmetrical Current Basis.ANSVEEEC37.011-1979ApplicationGuide for TransientRecoveryVoltage for ac High Voltage Circuit Breakers Rated on a Symm etricalCurrent Basis.0. Naef, C.P. Zimmerman, J. E.Beehler,ProposedTransientRecov-eryVoltageRatings for PowerCircuitBreakers, EEETransactions,Power Apparatus and Systems, V0176 Part11; 1508-1516,1958.

10.C. L. Wagner,H. M. Analysis of TransientRecoveryVoltage(TRV) Rating Concepts, IEEE Transactions n Power Apparatus and Sys-tems, Vol PAS 103, No. 11; 3354-3363, April 1984 .

11. R. G. ColclaserJr., D. E.Buettner,TheTravelingWaveApproach toTransient Recovery Voltage, IEEE Transactions on Power Apparatus andSystems, Vol PAS 86; 1028-1035, June 1969.

12. S. R. Lambert, Application of Power Circuit Breakers, IEEE Power En-gineering Society, 93 E H 0 388-9-PwR, 32-37,1993.

13. G. Catenacci, Electra 46 ,39 (1976).



SWITCHING OVERVOLTAGES

4.0 IntroductionAmong the most common reasons for dielectric failures n an electricaside from lighting strikes,re the overvoltages produced y the switching thatis normally required or the ordinary operation ofhe electrical network.

Switching overvoltages can be produced by closing an unloaded line, byopening an isolating switch, or by interrupting low currentsn inductive or ca-

pacitive circuits wherehe possibility of restrikes exists.Switching overvoltages are probabilistic in nature andheir appearances in

a system depend mainly upon he number of faults that mustbe cleared on aline and on how frequently routine switching operations are performed on aparticular system. This implies that not only opening operations that are in-tended for interrupting a short circuit current are responsible for switchingovervoltages; but, also the many routine operations that are performed, some-times daily, in a system. These routine operations are fully capablef produc-ing overvoltage effects by virtuef them alteringhe system confguration.

As it has been said repeatedly, overvoltagesn transmission and distributionsystems can not be totally avoided, but their effects can be minimized. Gen-erally the occurrence and the magnitude of the overvoltage can be limited bythe use of appropriate measures suchs the use of series or parallel compensa-

closingresistors,surgesuppressors;such as metaloxidevaristors, orsnubbers containing combinations of resistors and capacitors, and in somecases by simply following basic established procedures or the proper designand operation of a system. [l]

It is appropriate at point to emphasize that although circuit breakersparticipate in the process of overvoltage generation they do not generate thesevoltages, but rather these voltagesre generated by the system. Circuit break-ers, however can provide means for decreasing, or controlling, these voltagesurges. They do either by timing controlsr by incorporating additionalhardware suchas closing resistorsas an integral part of the circuit breaker de-sign. [2]

4.1 Contacts Closing

The simple closingof a switch or of a circuit breaker produce significantovervoltages in an electric These overvoltages are due to the system

111



adjusting itself toan emerging different configuration of components s a re-sult of the addition of a load impedance. Furthermore, there are charges thatare trapped in the lines and in the equipment that is connected to the systemand these charges now muste redistributed within the system.

In addition, and wheneverhe closure of the circuit occurs immediately f-

ter a circuit breaker opening operationhe trapped charges left over from hepreceding opening can significantly contribute tohe increase in the magnitude

of the overvoltages that may appear in he system. It is important to note thatin most cases the highest overvoltages wille produced by the fast reclosing oa line. It should also be realized that the higher magnitudes the overvolt-ages produced byhe closing or the reclosing operation a circuit breaker wilalways be observed at the open endof the line.

Although the basic expressions describing the voltage distribution acrossthe source and the line are relatively simple, defininghe effective impedancethat controls the voltage distribution withinhe elements of the circuit is rather

difficult and generally only be adequately handled with the aidof a com-puter.

Because of the complexity of the problem no attempt will e made here oprovide a quantitative solution. The aim of chapter will be to describequalitatively the voltage surges phenomena that takelace during a closingora reclosing operation, and during some special cases of current interruption.The upper limits of overvoltages that have been obtained either experimeor by calculationwill be quoted but onlys general guidelines.

4.1.1 Closing of a Line

A cable that is being energized from a transformer represents he simplestcase of a switching operation as is shown in figure 4.1 (a). For the sake of

simplicity, the transformer has been represented by its leakage inductance;while the cable is represented by its capacitance. a result simplifi-cation the equivalent circuit can take the formhe circuit illustrated in figur4.1 (b).

The transient voltage, shown n figure 4.1 (c), oscillates alonghe line at arelatively low single frequency and has an amplitude that reaches a peakalueapproximately equal to twicehe value of the system voltage that was presentat the instant at which the closure of the circuit took place.

Although the above described circuit may be found in some very basicapplications, in actual practice t ismore likely to expect that a typical systemwill consist of one or more long interconnected overhead lines, depicted in

figure 4.2 (a). The equivalent circuit, and he transient responseof system

is shown in figure 4.2 (b) and (c) respectively. The transient response, as beseen in the figure, is by the combined impedance the transformer



Tr CB

3E 0 4

b

4.1 Representation of the simplest case of closing into a line. (a) Single line

schematic, (b)Equivalent circuitand (c) Transient surge.



Tr

CB

Figire 4.2 Switching surge resulting from energizing a complex system. (a) Single

l i e schematic the system, @) Equivalent circuitand (c) Surge voltage.



that is feeding the system and by he total surge impedance of the connectedlines. The total surge impedance, as it be recalled, is equal to the surgeimpedance of each individual ine divided by he number of connected lines.

The total closing overvoltages given by the of the power frequencysource overvoltage andhe transient overvoltage being generatedt the line.

The overvoltage factoror the source is given by he following equation.

where:

f = power frequencyL = positive sequence inductanceer length of lineC = positive sequence capacitanceer length of line

1 =linelengthX = short circuit reactance of source2 = surge impedance f the line

It is evident, by simply observation of the above equation, that a higherpower fiequency overvoltage factor cane expected as a result of the follow-ing occurrences:

1. When the length of the lines increase

2. When the source reactance increases3. When the surge impedance of the line is lowered as a direct result of an

4. When the power frequency is increased, which means hat the overvoltageincreased number f connected lines and

is higher in a 60 Hz system than n a 50 Hz one.

The overvoltage factoror the transient response portionf the phenomenais not as easy to calculate m anually and a simple formulas in the precedingparagraph is just not available. However, t ispossible to generalize and tbe said that he overvoltage factor or the transient response s proportional to:

1. the instantaneous voltage difference between the source voltage and the

2. the damping impedance of the lines connected at the source side of the

3. the terminal impedance of the unloaded line/ lines being energized

line voltage as the contacts of the circuit breaker close,

circuit and

In any case whats important to remember is:

1. When switching a number of lines the amplitude factor f the overvoltageis always reducedas the size of the system increases, and



The reduction of the amplitude factor is not due to the damping effects ofthe system but rather to the superposition f the individual responses eachhaving a different frequency.

4.1.2 Reclosing of a Line

Since in order to improve the stability of the system it is desirable to restoreservice as quickly as possible, it is a common operating practice to reclose a

circuit breaker a few cycles after it has interrupted a fault.If the interrupted fault happens to e a single phase to ground fault, thent

is possible that a signifcant voltage may remain trapped in the unfaultedphases. happens because the three phases represent a capacitor that hasbeen switchedoffat current zero and therefore, because ofhe inductive natureof the system, coincideswith the instant where a maximum voltages pre-sent in the line.

Since the closing of the contacts may take lace at any point in the voltage

wave, it could then be expected that when reclosing the circuit, the circuitbreaker contacts may close at the opposite polarity of the trapped charge,which, when coupled with the voltage doubling effect producedy the travel-ing wave, leads tohe possibility of an overvoltage acrosshe contacts that canreach a magnitude s high as 4 per unit.

4.2 Contact Opening

The opening of a circuit was previously discussedn the context of interruptinga large magnitude of current wherehat current was generally considered toethe result of a short circuit. However, therere many occasions where a circuitbreaker is required to interrupt currents thatare in the range of a few amperesto hundred amperes, and where the loads are characterized as beingeither purely capacitive r purely inductive.

The physics of the basic interrupting process; hat is the balancing of thearc energy is no different whether the intermpted currents re small or large.However, since lower currents will contribute less energyo the arc it s naturalto expect that interrupting these lower currents shouldbe a relatively simpler

but, is not always he case because,as it will be shown later, he veryfact that the currents re relatively low in comparison to a short circuit currentpromotes the possibility of restrikes occurring acrosshe contacts during inter-ruption. Those restrikes be responsible for significant increases in themagnitude of the recovery voltage.

According to standard established practice, a restrikes defined as being an

electrical discharge that occurs one quarterf a cycle or more after he initialcurrent interruption. A reignitions defined as a discharge that occurs notaterthan one eighthf a cycle after current zero.



1

0.50

a2

-0.50

l4? -1.00

-15 0

-2.00

I a \

Figure 4.3 Recovery voltage resulting from the switchingf capacitorbanks

4.2.1 Interruption of Small Capacitive Currents

The switching of capacitor banks and unloaded lines requires that the circuitbreaker interrupts small capacitive currents. These currents are generally lessthan ten amperes for switching unloaded lines and most often less than onethousand amperes or switching capacitor banks.

Interruption, asalways, takes place t current zero and thereforehe systemvoltage, for a ll practical purposesis at its peak. This as it should be recalledmakes current interruption relatively asy but, again as it was before, thisis not necessarily because those low currents may be interrupted when thegap between the circuit breaker contactss very short and consequently,a fewmilliseconds later as the system recovery voltage appears the circuitbreaker contacts he gap is still rather small andt may be very difficult or thecircuit breaker to withstandhe recovery voltage.

At the time when current interruption takes lace the line to ground voltagestored in the capacitor n a solidly grounded circuits equal to 1.0 per unit.



The source side, n the other hand, will followhe oscillation of the powefrequency voltage and thereforen approximately onehalf of a cycle the volt-age across the contacts would reach its peak value but, w ith a reversalof itspolarity. At this time then the total voltagea m s s the contacts reaches a valueof 2.0 per unit which corresponds tohe algebraic of the capacitor voltagecharge and he source voltage s is shown in figure

If he circuit has an isolated neutral connection thenhe voltage trapped in

the capacitor, for the first phase to clear, has a line to ground value f 1.5 perunit and he total voltage acrosshe contacts one half f a cycle later will thenbe equal to 2.5 per unit.

Restrikes can be thought as being similar to a closing operation where hecapacitor is suddenly reconnected to the source, and therefore it is expectedthat there wille a flow of an inrush current which due to the inductancef thecircuit and n the absence of any damping effects will orce the voltage in thecapacitor to swing with respect to the instantaneous system voltage to a peak

value that is approximately equal to the initial value at which it started butwith a reversed polarity.f the restrike happenst the peakof the system volt-age, then he capacitor voltage will attaincharge valueof per unit. Underthese conditions,if the high fiequency inrush currents interrupted at the zerocrossing, which some circuit breakers are capablef doing so, then the capaci-tor will be left with a charge correspondingo a voltageof per unit and onehalf of a cycle later there w ill be a voltage of per unit applied a m s s thecircuit breaker contacts. f the sequence is repeated, the capacitor voltage will

reach a per unit value, as is illustrated in figure 4.4. Theoretically, and if

damping is ignored, the voltage across he capacitor build up according toa seriesof . . . nd so on without limit.

4.2.2 Interruption of Inductive Load Currents

When a circuit breaker that has an interrupting capability of several tens ofkiloamperes is called upon to interrupt inductive load currents that are gener-ally in he range of a few tens to some hundredsf amperes, as for example in

the case of arc furnace switching, those currents are interrupted in a normalfashion, that is at current zero. However, and again due to the high intermpt-ing capacity of the circuit breaker those small inductive currents cane inter-rupted rather easilyn a manner similaro that which was describedn connec-tion with the intermption of small capacitive currents.

At the time of interruption the gap between he contacts may be very short,and since he voltage is at its peak, then in many cases he small gap may notsufficient to withstandhe full magnitude of the recovery voltage which begi

to appear across he contacts immediately followinghe intermption of therent. As a consequence the arc may restrike resulting in a very steep voltagechange and n signifcant overvoltages.



Figure 4.4 Voltage escalation due to restrikes during a capacitance switching opera-tion.

However, because of the randomness of the point at which the restrikestake place and due to he inherent dampingof the circuit, it is very unlike thatthe upper limitof these overvoltages will exceed aalue of 2 per unit.

There are however special cases that arise when a circuit breaker has ex-ceptional capabilities for intermpting high frequency currents, such as thosegenerated by a reignitionr a restrike. Whenever he high frequency currentsinterrupted the normal power frequency recovery voltage reappears acrosshecontacts and in some casest is ossible that a restrikemay occur again.

During the interval between the two reignitions the contacts have movedthus increasing he gap distance and therefore a higher breakdown voltages tobe expected. Nevertheless, during this interval more magnetic energy s accu-mulated in the inductance of the load and consequently additional energy isavailable to trigger a breakdown which would occurt a voltage that s higherthan the previous one. This process may repeat itself as successive reignitionsoccur amss a largergapand at increasedmagneticenergy evels,and



Figure 4.5 Voltage surges caused by successivereignitionswhen interrupting low in-

ductive currents.

therefore, at higher mean voltage levels resultingn a high frequency series ovoltage spikes suchs those shown n figure 4.5.

Because of the statistical nature of phenomenon it is not possible toestablish an upper limit for the overvoltage; however, it is advisable to beaware of the potential isk and to use protective devices suchs surge arrestors.

4.2.3 Current ChoppingCurrent chopping is the result of the premature extinction of the power fre-quency current before a natural current zero is reached that is due to the arcinstability as the current approaches zero.

It is commonly believed that vacuum circuit breakers are capable ofchopping currents. However, this is not the case, all types of circuit breakerscanchop.Nevertheless,what is different is that he nstantaneouscurrentmagnitude at which the chopping occurs varies among the different types ofinterrupting m ediums and indeed t is higher for vacuum interrupters. [4], [ 5 ]



(c)

Figure 4.6 TypicalCurrentChopping. (a) Equivalent circuit, (b) Choppedcurrent

across the breakernd (c)Transient voltage acrosshebreaker.



In theory, when current chopping occurshe current is reduced instantane-ously from a small inite value to but, in reality does not happensuddenly simply because f the inductance that is present in the circuit andasit is well known, current can not change instantaneously in an inductor. It istherefore, tobe expected that some smallinite element of time must elapseorthe transfer of the magnetic energy thats trapped in the system inductance.

At the instant when current chopping occurs he energy stored in the load

inductance is transferred to the load side capacitance and thus creating a con-dition where overvoltages be generated. In figure 4.6 (a) the simplifiedequivalent circuit is shown and in (b) the voltage and current relationshipsreillustrated.

Referring to the equivalent circuit the energy balance equations can bewritten as:

1 1 1

- CE i = -CE,Z +-Hi2

and the overvoltage factorK s given by:

where:

E,,,= Overvoltage peakEo= Peak voltage t supply sideE, = Capacitor voltage t instant of chopIo= Instantaneous value of chopped current

L-= Surge impedance ofhe circuitC

As it be seen, the magnitude of the overvoltage factor K is highly de-

pendent uponhe instantaneous value f the chopping current.

4.2.3.1 Current Chopping Circu it Breakers other than Vacuum

For air, oil, r SF, interrupters, the arc instability that leadso current choppingis primarily controlled by the capacitance of the system. The effects of the

system capacitance on he chopping level are illustrated in figure 4.7 [6]. Theeffects of the capacitance on vacuum interrupterss also included in figurefor comparison purposes.



0.001 0.01 0.1

SYSTEM CAPACITANCE (micro farads)

1

Figure 4.7 Current Chopping Level as Function System Capacitance forMinimum0il.Circuit Breakers(MOCB), SF, Gas CircuitBreakers(GCB),AirBlastCircuit

Breakers (ABCB), and Vacuum Circuit Breakers (VAC).

For gas or oil circuit breakershe approximate value of the choppingrent is given byhe formula

where:

h=Chopping number

The following re typical values or chopping numbers:

ForMinimumOilcircuitbreakers 7 to 10 lo4For Air Blastircuitreakers 15 to 40 lo4

For SF, puffer circuitreakers 4 to 17 lo4

The values f the system capacitance cane assumed to be in the range of10 to 50 nano-farads.



4.2.3.2 Current Chopping in Vacuum Circuit Breakers

In contrast to other typesof circuit breakers he current instability n vacuuminterrupters is not strongly influenced by the capacitance of the system (seefigure 4.7), but is dependent upon the material f the vacuum contacts and ythe action of the anode spot created byhe vacuum arc. There is no choppingnumber for vacuum interrupters but, insteadhe chopping current itself can e

specified as follows:ForCopper-Bismuthcontactscurrentchopping 5 to 17 AmperesForChromeCoppercontactscurrentchopping 2 to 5 Amperes

4.2.4 Virtual Current Chopping

Virtual current chopping in reality is not a true chopping phenomenon butrather it is the normal interruption of a fast transient current. V irtual choppinis a phrase that has been coined to describe the condition illustrated in the

simplified circuit shownn figure 4.8.Referring to this figure, the power frequency currents are shown as I , IB

and IC. Assuming that or example, a current reignitionccurs shortly after theinterruptionof the power frequency current in phase he reignition current,will then flow o ground through he line to ground capacitanceC, in the loadside of the breaker in phase A and the components i b and i, flow in phases Band C due to the coupling of their respective lineo ground capacitances.

The high frequency transient current produced by the reignition superim-

poses itself on the power frequency; furthermore, the high frequency currentcould be larger in magnitude than he power frequency current and thereforetcan force current zeroes t times other than those expected to occur normallywith a50 60Hz current.

As it has been stated before therere some types of circuit breakers whichare capable to interrupt these high frequency currents, and thereforet is pos-sible to assume that n some cases he circuit breaker may clearhe circuit at acurrent zero crossing that has been forced by the high frequency current and

that the zero crossing occurs at a time prior to that of the natural zero of thepower frequency current. Whenhishappens, as far as the load is concerned, itlooks the same as if the power frequency current has been chopped since asudden current zero has been forced.

Since the high frequency current zeroes will occur at approximately thesame time in all three phases he circuit breaker may intermpt the currents inall three phases simultaneously thus giving rise to a very complicated seqof voltage transients that ay even include reignitionsn all three phases.

Considering that, when compared in a "normal" current chopping, we fithat the instantaneous valuef current, from whichhe load current s forced to



Figure 4.8 Virtual current chopping (a) Circuit showing the flow of the induced cur-

rents. @) Relationships between the three phase currents (fromef.7).



zero, is significantly higher but, that the surge impedance is somewhatlower, then the ine to ground overvoltage could e assumed to be at about thesame orderof magnitude as the overvoltages thatare generated by the conven-tionalcurrentchopping;however, in the worstcase, if the neuttal is un-grounded onehalf of the reignition current would return each of theother two phases and they both will be in phase but with opposite polaritieswhich will result n the line to ine overvoltage of the two phases being twicetheir corresponding line to ground overvoltage.

4.2.5 Controlling Overvoltages

Circuit breakers themselves do not generate overvoltages, but they do initiatethem by changing the quiescent conditionsof the circuit. As it has been statedbefore the switching overvoltages re the result of two overvoltage componentsthe power frequency overvoltage, and the transient overvoltage component.Limiting the m agnitude of the f m t is usually sufficient to reduce the total

overvoltage to within acceptable limits. However, this does not exclude thepossibility of using appropriate measures to additionally limithe magnitude of

the overvoltage by limitinghe transient response.Among the measures that can be taken to reduce the magnitudes of the

power frequency overvoltages are:

(a) Provide polarity controlled closing(b) Add closing andor opening resistors acrosshe breaker contacts(c) Provide a method combining polarity control and closing resistors(d) Add parallel compensation(e) Reduce the supply side reactance

The ransient overvoltage factor cane controlled by:

(a) removing the trapped charges from theine(b) synchronized closing which be accomplished either by closing at a

voltage zero ofhe supply side or by matching the polarityf the line andthe supply side.

(c) synchronized opening which optimizeshe contact gapat current zero(d) using pre-insertion resistors

From all the listed alternatives only resistors can be considered to be an

integral partof a circuit breaker.The practice of including closing resistors spart of a circuit breaker s relatively commonfor circuit breakers intendedorapplications at voltages above 123V.

REFERENCES1. AllanGreenwood,ElectricalTransients in Powerystems,Wiley-

Interscience, New York, 1971.



2.

4.

5.

6.

7.8.

9.

R. G. Colclaser Jr., Charles L. Wagner, Edward P. Donahue, MultistepResistor Control o f Switching Surges, IEEE Trans. PA&S Vol. PAS-88

G. W.Stagg,A. H. El-Abiad,ComputermethodsnPowerSystemAnalysis, McGraw Hill New York 1968.S. Berneryd,Interruption of SmallInductiveCurrentsSimplePhysicalModel and Interaction with Network.TheUniversity o f Sydney. Ab-

Symposium on Circuit Breaker Interruption and Power Testing,May 1976.Small inductive Current Switching,IGRE WG 13.02, Electra 72M. Murano, S. Yanabu, H, Ohashi, H. shizuka and T. Okazaki, CurrentChopping Phenomena of Medium Voltage Circuit Breakers IEEE Trans.PA&S Vol. PAS-97 No.1, Jan-Feb 1977, 143-149.Virtual current chopping, Electra, 72,87-90.Thomas E. Brown, Circuit Interruption, Theory and Applications, Marcel

Dekker Inc., New York, 1984.W. S. Meyer,T.H.Liu,ElectromagneticTransients Program RuleBook, Bonneville Power Administration Portland Oregon, 1980.

NO. 7, July 1969, 102-1028.



5

TYPES OF CIRCUIT BREAKERS

5.0 Introduction

Circuit breakersare switching devices which according o American NationalStandards Association C37.100 [l] are defined as: “A mechanical de-vice capable of making, carrying and breaking currents under normal circuitconditions and also making, carryingor a specific time and breaking currentsunder specified abnormal circuit conditions suchs those of short circuit.’’

Historically, as the operating voltages and the short circuit capacities of the

power systems have continued to increase,ighvoltage,highpower circuit breakershave evolved trying to keep pace with the of the electric powerNew technologies, primarily those involving the use of advanced interrupting medums, ave been developed and continue to betudied

To achieve current interruption some of the early circuit breaker designssimply relied on stretching the arc across a pair of contacts in later arcchute structures, including some with magnetic blow-out coil were incorpo-rated, while other devices used a liquid medium, including water but more

generally oil,as

the interrupting medium.Some of those early designs have been significantly improved and varia-tions of those typesof circuit breakers are still in use specially in low voltageapplications where presently plainir circuit breakers constitute he dominanttype of circuit breakers used.

For indoor applications,t system voltages that range from approximatelkV up to 38 kV, air magnetic circuit breakers were the breakers of choice inthe US until the mid nineteen seventies, whilen Europe and other installations

outside of the US, minimum oil circuit breakers were quite popular. Bulk oiland air blast circuit breakers where quite common untilhe mid seventies foroutdoor applications t voltages ranging from 15V up to 345 kV. For indoor,medium voltage applications minimumil circuit breakers enjoyedgreat dealof popularity inEurope.

With the advent of vacuum and sulfurhexafluoridehe older typesof circuitbreaker designs have been quickly superseded and today they can be effec-tively considered s being obsolete technologies.

What we just described suggests that circuit breakers can e used for dif-ferent applications, that they can have different physical design characteristicsand hat heyalsocanperform heir nterruptingdutiesusingdifferentquenching mediums and design concepts.

129



5.1 Circuit Breaker Classifications

Circuit breakers be arbitrarily using many different criteria suchas; he intended voltage application,he location where they re installed, theirexternaldesigncharacteristics,andperhapsandmost mportantly,by themethod and the medium usedor the interruption of current.

5.1.1 Circuit Breaker Types by Voltage Class

A logical starting point for establishing a classification of circuit breakers isthe voltage level at which the circuit breakers are intended to be used. Thisfirstbroad classification divideshe circuit breakers into two

Low voltage circuit breakers, whichre those thatare rated for use a t volt-

2. Highvoltagecircuitbreakers.Highvoltagecircuitbreakers are thoseages to volts, and

which are applied r that have a ratingf volts or more.

Each of these is further subdivided andn the case of the high volt-age circuit breakers theyare split between circuit breakers that are ratedkV andaboveand hoserated kV andbelow.Sometimes hese wo

are referred as the transmission class, and the distribution or mediumvoltage class f circuit breakers, respectively.

The above classification of high voltage circuit breakerss the one that iscurrently being used by international standards such s ANSI and

the International Electrotechnical Commission [4].

5.1.2 Circuit Breaker Types by Installation

High voltage circuit breakers cane used in either indooror outdoor installa-tions. ndoorcircuitbreakers are defined in ANSI [l] as those“designed for use only inside buildingsr weather resistant enclosures.”

For medium voltage breakers ranging from kV to kV it generallymeans that indoor circuit breakers are designedor use inside of a metal clad

switchgear enclosure.Inpractice, the only differences between indoor and outdoor circuit breakis the external packaging or the enclosures hat are used, as illus-

trated in figures5.1 and 5.2. The internal current carrying parts, thentenupt-ing chambers and the operating mechanisms, in many cases, are the same forboth types of circuit breakers provided that they have the same currents andvoltage ratings andhat they utilize the samentenupting technology.

5.1.3 Circuit Breaker Types by External Design

From the point of view of their physical design, outdoor circuitbreakers can be identified as either dead or the live type circuitbreakers. These two breaker types are shownn figures and 5.4.



Figure5.1 Indoor type circuit breaker.

Figure5.2 Outdoor type circuit breaker.



Dead circuit breakers re defined in C37.100 [l]as “ A switch-

ing device in which a vessel@) t ground potential surrounds and containsheinterrupter@) andhe insulating mediums.”

A live circuit breaker s defined in he same standard as: A switchingdevice in which the vessel@) housing he interrupter@) s at a potential aboveground.”

Dead circuit breakers are the preferred choice in the US and in mostcountries that adhere to published standards. These circuit breakers aresaid to have the following advantages overhe live circuit breakers:

1. Multiple low voltage bushing type currentransformers be instalied at

2. They have a low, more aesthetic silhouette.3. Their unitized constructionoffers a high seismic withstand capability4. They are shipped factory assembled and adjustmentsre factory made

both, the line side and the load sidef the circuit breaker.

In applications where the E C standardsarefollowed l i e circuitbreakers are the norm. The following advantages are generally listed for live

circuit breakers:

1. Lower cost of circuit breaker (without current transformers)Less mounting space requirements

3. Uses a lesser amount of interrupting fluid

5.1.4Circuit Breakers

Typeby Interrupting Mediums

In the evolutionary processof the circuit breaker technology,he factorsthat have dictated he overall design parametersf the device are, he interrupt-ing and insulating medium that used, together with the methods that areutilized to achieve the proper interaction betweenhe interrupting medium andthe electric arc.

The choice f air and oil,as the interrupting mediums, was madet theof the century and it is remarkable how well and how reliably these mediums

have served the industry.Two newer technologies, one using vacuum and he other sulfurhexafluo-

ride (SF,) gas as the intempting medium, made their appearance t about thesame time in the late 1 9 5 0 ’ ~ ~nd are now what is considered to be the newgeneration of circuit breakers.

Although vacuum and SF, constitute today’s leading technologies, acussion of air and oil circuit breakers is provided because many of these de-vices are still n service, moreover, because manyf the requirements specified

in some of he existing standards were based onhe operating characteristics fthose circuit breakers.



Figure 5.3 Dead tank type circuit breaker (grounded enclosure).

Figure 5.4 Live type circuit breaker (ungrounded enclosure).



5.2 Air Magnetic Circuit Breakers

A plain break knife switch operated in free air, under normal atmosphericconditions, is one of the earliest versionsknown of a circuit breaker. However,

simple device had a very limited capacityn terms of voltage andof inter-rupting current.

The interrupting capability limitations of such switch investiga-

tions that led to the development of improved designs. Those improvementsinvolved the inclusion of a number of components whose function was to en-hance the cooling of he arc.

The most signifcant advancement was the development of the arc chute,which is a box ike component device that contains a numberf either metallicor insulated plates. Additionally, in most of the designs, when intended formedium voltage applications, therc chute includes a magnetic blow-out co

It is a well-known fact that ir at atmospheric pressure has a relatively lo

dielectric strength, it is also known that in still air nothing accelerates theprocess of recombination and therefore the time constant for deionization isfairly large.

It should be recalled that a low time constants highly desirable, in fact isindispensable for high voltage devices; However, a high time constant is ac-ceptable in low and at the lower end of the medium voltage devices, wheretoffers the advantages of substantially reducing the possibility of generatingswitching ovemoltages, and n many cases of modifying the inherent TRV of

the system. This influence may be such that for ll practical purposes this typof breakers be considered to e insensitive toTRVproblems.

Since current interruption across anir break is based only uponhe naturaldeionization process that takeslace in the air surrounding the arc; then, in or-der to improve he interrupting capabilityt is necessary to enhance the deioni-zation process by meansf some appropriate external cooling method.

We should recall that in order to maintain the ionization of the gas, whenthe arc is effectively cooled, the m agnitude of the arc voltage m ust increase.

What this means is simply that as the arc cools, the cooling effectively in-creases the deionization of the arc space, which in urn increases the arc resis-tance. As a consequence of the increase in resistance, the short circuit currentand the phase angleare reduced and thus, he likelihood of a successful inter-ruption is significantly enhanced.

In an air circuit breaker, increasing the resistance of the arc in effect in-creases the arc voltage Thus, o effectively increase the rc voltage any of thefollowing means cane used.

1. increase the length of the arc, which increases the voltage drop a m s s thepositive column f the arc.



split the arc into a number of shorter arcs connected in series. Whatdoes is that instead f having a single cathode and anodet the endsof thesingle arc column there are now a multiplicity of cathode and anode re-gions, which have additive voltage drops. Although he short arcs reducethe voltage of each individual positive column, the summation of all thevoltage drops is usually greater than thatof a single column; finthermore,if the numberof arcs is large enoughso that the summationof these volt-age drops is greater than the system voltage a quick extinctionf the arc ispossible.Constricting the arc,bycoflstraining it betweenverynarrowchannels.This in effect reduces the section of the arc column and thus in-creases the arc voltage.

With both of the last two suggested methods there is an added benefit,which is the additional coolingof the arc as the result of the high energy stor-

age capacity provided byhe arc chute plates thatre housed inside of the arcchute itself.

5.2.1 Arc Chute Type Circuit Breakers

arc chute be described as a box shaped structure made of insulatingmaterials. Each arc chute surrounds a single pole of the circuit breaker inde-pendently, and it provides structural support or a setof arc plates and n somecases, when o equipped, it houses a uilt in magnetic blow-out coil

Basically thereare two types of arc chutes where each types characterizedprimarily by the material of the arc plates that are used. Some arc plates are

made of soft steel and n some casesare nickel plated.In type of arc chutethe arc is initially guided insidehe plates by means of rc which aresimply a pair of modified arc horns. Subsequently the arc moves deeper intothe arc chute due to the forces produced by the current loop andhe pressure ofthe heated gases..

To enhance and o control the motionof the arc, vertical slotsare into

the plates. The geometrical pattern of these slots varies among circuit breakermanufacturers, and although there may be some similarities in the plate de-signs, each manufacturer generally has a unique design ofts own.

When the arc comesn contact with he metal plates t divides into anum-ber of shorter arcs that burn across a set of adjacent plates. voltage dropthat is observed a m s s each of these short arcs s usually about to 40 volts.The majority of voltage beingdue to the cathode and anodedrop of eacharc. The voltage drop of the positive column depends in the plate spacing,

which in determines the length f the arc’s positive column.A

schematicrepresentation of type of arc chute, which is used almost exclusively inlow voltage applications,s shown in figure 5.5.



I

I

Short Arcs

.. . .I................... Metal

Plates

Contacts,

d

i

Figure 5.5 Outline of a plain arc chute used in low voltage circuitbreakers.

A second type ofarc chute is one that is generally synonymous with mag-

netic blow su t assist and which is used in circuit breakers intended for appli-cations at medium voltages of up to 15 kV and for interrupting symmetricalfault currents of up to 50U. his type of arc chute almost invariably uses in-sulated arcing plates that are made of a variety of ceramic materials such as

zirconiumoxideoraluminumoxide. An example of type ofcircuitbreaker is shown in figure 5.6.

With particular type of arc chute the cooling of the arc and its final

quenching is effected by a combination of processes. First, therc is elongated

as it is forced to travel upwards and through a tortuous path thats dictated bythe geometry and the locationf the insulating plates and their slits.

Simultaneously the arc is constricted as it travels through the slots in thearc plates and as the arc fills the narrow space between the plates. Finally,when the arc gets in contact with the walls of the insulating plates the arc iscooled by diffusion to these walls.

The diffusive cooling is strongly dependent uponhe spacing of the plates,dependence has been shown byG. Frind [ 5 ] and his results are illustrated

in figure 5.7.

Since the arc behaves like alexiile and stretchable conductor, its possiile todrive he arc upwards orcing it into hespacesbetween hearcingplates,



Figure 5.6 Typical 15 kVAir Magnetic circuitbreaker.

and thus rather effectively increasing the length ofhe arc and ts resistance.Motion of the arc is forced by the action the magnetic field produced by a

coil that s generally found embedded intohe external supporting plates ofhearc chute. In figures 5.8 and 5.9 a complete arc chute assembly and a coilhatis to be potted are shown.

The coil, which is not a part of the conducting circuit during normal con-

operation, is connected to he ends of an arcing gap as shown schemati-cally in figure 5.10 (a). When he circuit breaker begins o open, the current

from the main contacts to he arcing contacts where, uponheir sepa-ration, the arc is initiated (figure5.10 (b)) As the contacts continue o increasetheir separation, the arc is forced into the arc runners where the coil is con-nected, and in so doing the coil s inserted into he circuit (figure5.10 (c)) Themagnetic field created by the coil will now except a force upon the arc thattends to move the arc up deeper inside he arc chute, figure5.10 (a).

The heatingof the arc plates and their ir spaces results inhe emission of

large amountsof gases and vapors, that muste exhausted through the opening

at the topof the arc chute.



100x10

inQ

0.1

w

= 0.01

10 100

PLATE DISTANCE (0.001 IN.)

Figure 5.7 Recovery time constant n air as a function of the spacing between thechute plates.

The mixture of gases and vapor is prevented from flowing back and intothe contacts by the magnetic pressure that results from the interaction of the

arc current and the magnetic field. As long as the forces produced byhe gasare lower than the magnetic forces,he flow of the gases willbe away from thecontacts.

important requirement for the connection of the coil is to theproper polarity relationshipso the arc is driven upwards and intohe intermpt-ing chamber. It is also important to have a phase lag between the magneticflux and the current being interrupted so that at current zero there is still aforce being exertedon the extinguishing arc.

Because at low current levels the magneticorce is relatively weak mostirmagnetic circuit breakers include some form of a puffer that blows a small

stream ofair into the arc, as the arcing contacts separate,o help drive he arcupwards and intohe plates.

In most designs, to avoid the possibility of releasing hot, partially ionizedgases which may cause secondary flashovers, a flat horizontal stack of metalplates is placed at the exhaust portof the arc chute.

Finally, it should be noted that even though is one of the oldest tech-

niques of current interruption and that in spite of all the theoretical and ex-perimental knowledge that has been gained over the years, the design of arcchutes remains very much s an art. Theoretical evaluationof a design is very

difficult and he designer still haso rely primarilyon experimentation.



Figure 5.8 Complete assemblyof a 15 kV arc chute.



Figure 5.10 Blow-out coil insertion sequencef an air magnetic circuit breaker:

(a) circuit breaker closed, coil by-passed and (b) circuit breaker during arcing period

coil inserted.



5.2.2 Air Magnetic Circuit Breakers: Typical Applications

One important characteristic f air magnetic circuit break& is that their inter-rupting capability s greatly influenced byhe magnitude of the voltage.

It has been demonstrated that the interrupting current capability increasesas the voltage decreases and consequently these breakers can be referred asbeing a voltage controlled interrupter.This characteristic, as it will be shown

in a later chapter, is reflected in some existing performance standards wherewithin some specific limitshe intermpting current capabilitys given by:

Because of the high arc resistance that is characteristically exhibited bythese circuit breakers, theyre capable of modifying the normal wave form of

the fault current to the point that they may even advance the occurrence of acurrent zero. This represents a significant characteristic f these breakers spe-cially for applications in circuits where the fault current asymmetry exceeds100% and where there ay not be currents zeroes for several cycles.his highasymmetry condition is common in applications related to the protection oflarge generators.

Among some of the significant disadvantages of these circuit breakerswhen compared to modem type breakers f the same ratingsare their size and

their cost. Other disadvantages include;hort interrupting contact ife which isdue to the high energy levels hat are seen by the interrupter, their need for a

relatively high energy operating mechanism and the riskshat are posed by thehot gases when they are released into the switchgear compartment followingthe interruption of a short circuit current.

5.3 Air Blast Circuit Breakers

Although there was a patent issuedn 1927,air blast circuitbreakers werefirst

used commercially sometime around he year 1940, and for over five decadestechnology has proved toe quite successful.Air blast circuit breakers have been applied throughouthe complete high

voltage range, and until he advent of SF6 circuit breakers, they totally domi-nated the higher endof the transmission voltage class.n fact, at one time theywere the only type of circuit breakers available for applications at voltageshigher than 345V.

In reality, air blast circuit breakers shoulde identified as a specific type ofthe more generic class of gas blast circuit breakers becauseir is not necessar-ily the only gas that cane used to extinguish the arc, other gases suchs

trogen, carbon dioxide, hydrogen, fieon and of c o m e SF6 can be used. Fur-



thermore, it iswell known, and as isgenerally agreed, the interrupting processis the same for all gas blast circuit breakers andost of the differences in per-formance observed between ir blast and SF6 circuit breakersare the result ofthe variations in the cooling capabilities, and thereforen the deionization timeconstant of each of the gases. For reason the detailed treatment of the in-terrupting process willbe postponed to later in chapter, when describingthe more modemSF, technology.

Because there are some differences in the basic designs of air blast inter-rupters, and becausehe newer concepts or gas blast interrupters have evolvefrom the knowledge gained with the ir blast circuit breaker, thereare a num-ber of subjects which need to be addressed in section if nothing else, toprovide a historical framef reference.

In all of the designs of blast circuit breakers the interrupting processsinitiated by establishinghe arc between two receding contacts and by, simul-taneously with he initiation of the arc, opening a pneum atic valve which pro

duces a blast of high pressure that sweeps the arc column subjecting it tothe intense cooling effects f the flow.

5.3.1 Blast Directionand Nozzle Types

Depending upon he direction of the air flow in relation o the arc column [6]

there are,as shown in figure 5.11 (a),(b) and (c), three basic typesf blast ori-entations.

1. axialblast2.radialblast, andblast

From the three blast types, the axial the radial type are generally pre-ferred for extra high voltage applications, while the cross blast principle has

been used for applications involving medium voltage and very high interru

To effectively coolhe arc the gas flow in anxial blast interrupter muste

properly directed towards he location of the arc. Effective control of theflow is achieved by using a D’Laval typef a converging-diverging nozzle.

These nozzles be designed either as insulating, or as metallic or con-

ducting nozzles. Additionally and depending in the direction of flow for theexhaust gas each of the nozzles in these two groups, can be either what iscalled a single r a double flow nozzle.

A conducting single flow nozzle, s shown in figure 5.12, is one where themain stationary contact assembly serves a dual purpose. It carrieshe continu-

ous current when he circuit breaker s closed and as the circuit breaker opensthe arc is initiated across onef its edges and ts corresponding mating movincontact. After the gas flowis established, the pneumatic force exertedby the

ing currents.



1.2.

4.

ContactsNozzleArcDirectionof Gas Flow

Figure 5.11 Air blast direction: (a) axial direction, @) radial directionand (c) cross

blast or transverse direction.



i

Figure5.12 Outline of a conducting singleflownozzle.

gas on the arc effectively transfer the arc to a stationary arc terminal, or arc

catcher that is disposed longitudinally t the center of the nozzle.It is easily observed that with design the arc length can be increased

considerably andat a faster rate than that which s possible with an insulatingnozzle where he arc is initiated directly across a air of receding arcing con-tacts. Under these conditions the time needed for the arc to reach its finallength is dependent upon the final contact gap and consequently on the open-ing speedof the breaker contactshat is normally in the range f 3 to 6 metersper second (10 to 0 feet per second).

An axial insulating nozzles geometrically similar o the conducting nozzleas shown in figure 5.13 and as its name implies, the insulating nozzle s madeof an insulating material. The materialf choice is generally teflon, either s apure compoundor with some ype of filler material. Fillers are used to reducethe rate of erosion of the throat of the nozzle.

It should be noted that once the arc is properly attached to the intendedarcing contacts, the gas flow characteristics, and the interaction between thegas and he arc are the sameor both typesof nozzles.

The cross blast design s among one of the earliest concepts that was usedon air circuit breakers. As shown schematically in figure 5.11 (c) the arc isinitiated across a air of contacts and s subjected to a of air that flowsperpendicular to the of the arc column. It was contended that a consider-able amountof heat could be removed from the arc since the whole length of tarc is in contactith the airflow. However, is not the case, mainly because thcore of the arc has a lower density the surroundingairand therefore at the cen-tral part of the arc column theres very little motion between the arc and theas.Nevertheless, at the regions lying alongsidehe contacts where he roots of thearc are being elongated the gas flows in an axial direction and a substantial



Insulating Nozzle

GasFloW

Figure 5.13 Outline of anaxial insulating nozzle.

amount of cooling can be achieved. Most of the circuit breakers were madeusing a blast were applied onlyt medium voltages andigh

5.3.2 Series Connectionof Interrupters

Without a doubt, a single break interrupters the simplest and most economical so-

lution. However, significant improvementsn the interrupting capacity of circuitbreaker can e achieved by connecting number of interruptersn

It is to see that by connecting a number f interrupters in series, therecovery voltage, at least in theory, s equally divided a m s s each interrupter,it also be seen that the number of deionizing chambers is increased andthus the energy balancing processs increased proportionally tohe number ofinterrupters connectedn series.

One of the main difficulties encountered withhis type of application s to

ensure that during the transient period of he interruption process eachof theinterruptersoperatesunder the exactsameconditionsofeachother.Thissamenessrequirementapplies to both, the aerodynamicand the electricalconditions. From the aerodynamic point of view he flow conditions mustbemaintained for each interrupter. This generally requires the use of individualblast valves or each interrupter. It also requires that the lines connecting eachinterrupting chamber are properly balanced to avoid pressure drops that ayaffect the gas flow.

Electrically, the restriking voltages and consequentlyhe recovery voltages,must divide evenly a m s s each set of contacts. However, in actual practiceth is does not happen and an uneven distribution of voltage occurs due to he



unbalanced inherent capacitance that exists a m s s the interrupting device it-self, and between the line and the grounded parts of the circuit breaker.improve the voltage distribution across the gapst is a common practice to usegrading capacitorsor resistors connected between theive parts and ground.

5.3.3 Basic Interrupter Arrangements

Medium voltage air blast circuit breakers were normally dead designs.

Air circuit breakers intended or applications at systems voltages greater than72.5 kV were almost without exceptionf the live design type.

In some of the earlier designshe blast valve was located insidef the highpressure storage at ground level, whilehe interrupters were housed t theend of insulating columns. The tripping operation was initiated by openinheblast valve which in momentarily pressurized the interrupter causing hecontacts to move. The blast valve was closed later, n about 100 milliseconds,and as the pressure inside of the interrupters was decreased the breaker con-

tacts reclose.maintain system isolation in the open position of the circuit breaker

there was a built-in plain air break isolating switch hat also served as a dis-connecting switch or the grading capacitors,or resistors. The isolating switchwas timed to open in about 40 to 50 milliseconds after he opening time of the

contacts.The disadvantages associated with h is type of design were he larger gap

length required by he isolator at the higher system voltages. As it can be ex-

pected the greater required additional operating time and therefore fastreclosing times were difficult to achieve. Furthermore, since the exhaust wasopen to he atmosphere the consumption was relatively large.

A number of improvements where made focused primarily w ithhe objec-

tive of removing the need for the air isolating switch. Many different ar-rangements of blast and exhaust valves where used untilhe present design inwhich the interrupters are maintained ullypressurized at all timeswasadopted. In version tripping of the circuit breaker is executed by first

opening the exhaust valves and then sequentially openinghe contacts. M e r afew milliseconds the exhaust valves are closed while the contacts still are in

the open position. For closing the contacts are depressurized while the valvesare held closed.

With design, the air consumption was substantially reduced and whats

more important, since the intermpting chambers were held at the maximumpressure at all times, the breaking and he withstand capacity of he interrupterwas optimized. One last advantage that should be mentioned is that since theair consumption was reduced was the operating noise level of the inter-rupter. This is significant because air blast circuit breakers any notorious fortheir high operating noise level.



5.3.4 Parameters Influencing Air Blast Circuit Breaker Perform ance

There are many factors that influence the performance of a gas blast circuitbreaker. However, from all those factors there are some that be easilymeasured suchas the operating pressure, the nozzle diameter andhe interrupt-

These parameters have been usedo establish relationships 7], [S], that are

related to the voltage recovery capability of the interrupter in the thermal re-gion. These relationships are included in their graphic form in figures 5.14,5.15, and 5.16.

The significanceof these relationships s not in the absolute values hat arebeing presented because, depending onhe specific design f the nozzle and nthe overall efficiency ofhe interrupter, the magnitude of the variables change;however, the slope of the curves and therefore the exponentf the correspond-ing variables remains constant thus indicating a performance tr&d and givin

a point of reference for comparison between interrupter designs. Furthermoreand as it will be seen later there is a great deal of similarity between thesecurves and those obtainedor SF, interrupters.

ing current.

10

38&

W

La

=1 A . . A . . . . .

1 10 100

NOZZLE THROAT DIAMETEX(mm)

Figure 5.14 Voltage recovery capability in the thermalregion as a function of nozzlethroat diameterfor an airblast interrupter.



Figure 5.15 Voltage recovery capability in the thermal regions a function of pres-sure for an air blast interrupter.

10

RATEOFCHANGEOF dYdt

(Amperes mlcm-sec)

Figure 5.16 Voltage recovery capabilityn the thermal regions a functionof rate ofchange of current for n airblast interrupter.



5.4 Oil Circuit Breakers

From a historical perspective, the oil circuit breaker is the first design of abreaker for high power applications. It predates the blast type by severaldecades.

One of the f m t designs of an oil circuit breaker on record inhe US, is theone shown n figure 5.17. circuit breaker was designed and built by N.

Kelman in 1901. The breaker was installed on a 40 kV system that was capa-ble of delivering a maximum short circuit currentf to 300 amperes. Rec-ords indicate that the circuit breaker was in service from April 1902 untilMarch of 1903, whenfollowing a numberf circuit interruptions,at short timeintervals, blazing oil was spewed over onhe surrounding woodwork, startingfire which eventually spread tohe power house [9].

The design of circuit breaker was extremely simple. It consisted oftwo wooden barrels filled with a combination of water and oil. The contacts

consisted of two vertical blades connectedat the top and arrangedo that theywould drop into the stationary contacts to close the circuit.

Figure 5.17 Oil circuit breaker builtn 1901 by Kelman (reference9).



From these relatively humble beginnings the oil circuit breaker was refand improved but, throughoutll these mutationsit maintained its characteris-tic simplicity of construction and ts capability for intermpting large currents.

Oil circuit breakers where widely used and presently therere many still inservice. However, they have suffered the same fate as did the air blast circuitbreaker, they have been made obsolete byhe new SF6 technology.

5.4.1 Properties of Insulating OilThe type of oil that has been used in virtually all o il circuit breakers is onewhere naphthenic base petroleum oils have been carefblly refrned to avoidsludge or corrosion that may e produced bysulfuror other contaminants.

The resulting insulating oils identifed as type 10-C transformer oil. It ischaracterized byan excellent dielectric strength, by a good thermal conductiity (2.7 lo4 caVsec cm C) and by a high thermal capacity0.44 CaVg "C).

Some designsof oil circuit breakers take advantagef the excellent dielec-

tric withstand capabilities of o il and use the oil not only as intermpting me-dium but also as insulation within the live parts of the circuit breaker and toground.

Insulating oil at standard atmospheric conditions, and for a contactgap, is far superior than air or SF, under the same conditions. However, oilcan be degraded by small quantitiesof water, and by carbon deposits that rethe result of the carbonization of the oil. The carbonization takesplace due tothe contact of the oil with the electric arc.

The purityof the oil usually can e judged by ts clarity and transparency.Fresh oil has a clear amber color, while contaminated oil s darken and thereare some black deposits thathow signs of carbonization. The condition ofheoil normally is evaluated by testing for its withstand capability. The tests aremade using a spherical spark gapith two spheres 20 mm in diameter and t a

Fresh oil should have a dielectric capability greater than35 kV. For usedgap Of IlUn.

oil it s generally recommended hat capability be no less than 15V.

5.4.2 Current Interruption in Oil

At the time when the oil circuit breaker was invented no one knew that arcsdrawn in oil formed a bubble containing mainly hydrogen and that arcs buring in a hydrogen atmosphere tend tobe extinguished more readily arcsburning in other typesf gases. The choiceof oil was then indeed a fortuitouchoice that has worked very well over the years.

When an arc is drawn in oil the contacting oil surfaces are rapidly vapor-

ized due to the high temperature ofhe arc, whichas we already know is in therange of 5,000o 15,OOO"K. The vaporized gas then forms a gas bubble, wtotally surrounds he arc.



LIQUID I IFigure Gasbubble produced by an arc that is surroundedby oil.

It has been observed that he approximate compositionof bubble is 60

to 80% hydrogen, acetylene (C;H2) and the remainder consists of smallerproportions of methane and other gases.Within the gas bubble, shown in figure 5.18, therere at three easily iden-

tifiable zones. In the innermost zone, which containshe dissociated gases andis the one in direct contact with he arc, it has been observed thathe tempera-ture drops to between 500 to SOO(K. This gaseous zone is surrounded by ava-por zone where he vapor is superheated in its inside layers and s saturated atthe outside layers. The f i a l dentifiable zone s one of boiling liquid where t

the outside boundary the temperature of the liquid is practically equal to therelative ambient temperature.Considering that the arc in oil circuit breakers is burning in a gaseous at-

mosphere it would be proper to assume thathe theories of interruption devel-oped for gas breakers re applicable to he oil breaker. This assumption isbeen proven o be correct and therefore he performance of both, gas blastir-cuit breakers,as well as oil circuit breakers, be evaluated by applying he

theories ofarc interruption that were presented in chapter.

It has been demonstrated hat hydrogen is probably the ideal gas for inter-ruption, but the complications and of a gas recovery system make ts ap-plication impractical.



Figure 5.19 Thermalconductivity ofhydrogen.

Even though that comparatively speaking,he dielectric strength of hydro-gen is not particularlyhigh, its reignition voltage s 5 o 10 times higher thanthat of air. Hydrogen also has a very high thermal conductivity that is fasterduring the period of gas dissociation, as shown in figure 5.19, which resultsna more rapid cooling and deionizing ofhe arc

5.4.3 Types of Oil Circuit Breakers

In the earlier designsof oil circuit breakers he interrupters consisted f only

plain break andno consideration was given to include special devices o con-tain the arc or to enhance the arc extinguishing process.n hose early designsthe arc was merely confined within the walls of a rather large oil anddeionization was accomplished by (a) elongation the arc, (b) by the in-

creased pressure produced byhe heating of the oil in the arc region and (c) bythe natural turbulence that is set by the heated oil. plain break circuitbreaker concept s illustrated in figure.20.

To attain a successful intem ption , under these conditions,t isnecessary todevelop a comparatively long arc. However, long arcs are difficult o control,and in most cases leads to long periods of arcing.



Oil level

I Oil Bubble

Figure 5.20 Outlineof a plain break oil circuit breaker.

The random combinationsof long arcs, which translate into high rc volt-ages, accompanied by long arcing times make unpredictablehe amount of arcenergy that has tobe handled by the breaker. This unpredictability presents aproblem because it is not possible to design a device that handle such awide and non well-defined range of energy.

Plain break oil circuit breakers where generally limited onheir application

to 15 kV and maximum fault currents of only about 200 amperes.Moreover, these circuit breakers were good onlyn those situations where herate of rise of the recovery voltage was low.

The developmentof the explosion chamber,or interruptingpot, constituteda significant breakthrough for oil circuit breakers. It led to the designs of theso called “suicide breakers”. Basically the only major change made on theplain breaker design as the additionof the explosion pot, whichs a cylindri-cal container fabricated from a mechanically strong insulating material. Thiscylindrical chamber s mounted in such a way as to fully enclose the contactstructure. the bottom of the cham ber there is an orifice through which hemoving contact rods inserted.

The arc as before is drawn across the contacts, but now it is contained in-side the interrupting pot and thus the hydrogen bubble s also contained insidethe chamber. As the contacts continue to move and whenever the movingcontact rod separates itself from the orifice at the bottom of the chamber an



Figure 5.21 Outline of an explosion chamber type of oil interrupter. (a) Contactsclosed, (b) arc is initiated as contacts move, (c) gas escapes through interrupter potopening.

exit similar to a nozzle becomes availableor exhausting the hydrogen that istrapped inside he interrupting chamber. A schem atic drawing f design isincluded in figure .21.

One of the disadvantagesof design is its sensitivity to the point on thecurrent wave where the moving contact rod is separated from the interrupterchamber. If the first urrent zero occurs too early beforehe contact leaves hebottom orifice then the interrupter m ust wait for the next current zero which

may come a relatively long time fter the contact had lea the pot and conse-quently when the pressure inside the pot has decayed to an ineffective valuedue to the venting through he bottom orifice.

Another drawbackof this interrupter chamber s itsdependency on currentmagnitude. At high values of current the corresponding generated pressure shigh and may even reach levels that would result in the destruction of thechamber. Sometimes the high pressure has a beneficial quasi-balancing effectbecause the high pressure tends to reduce the arc length and the interrupting

time, thus, decreasing the rc energy input. However, with lower valuesfrent, the opposite occurs, he generated pressuresare low and the arcing times



Figure 5.22 Cross baMe interrupter hamber.

increase until a certain critical rangef current, reached whereit is diflicult toachieve interruption. This current level is commonly idenW1ed as the “critical

current.”Among the alternatives developed to overcome these limitations are the

inclusion of pressure relief devices to limit the pressure due to the high

rents. For the low current problem, the impulse breaker was developed. Thisdesign concept providesa piston pump intended to squi r t oil into the contactsat the precise time when interruptions taking place.

To reduce the sensitivity to the contact position at current zero thebaffle nterrupterchamberdesignwascreated. This designrapidlygainedpopularity and t became the preferred designor all the later vintage o il circuitbreakers. A typical interrupting chamberf type is shown in figure 5.22.

The design consists f a numberof specially designed insulated plates that

are stacked togethero form a passageor the arc that is alternately restricted



Figure 5.23 Oilbreaker interrupting chamber showing lateral vents.

and then laterally vented,s shown in figure 5.23. This design permits he lat-venting of the pressure generated inside f the chamber. This arrangement

subjects the arc to a continuous strong cross flow which has proven to bebeneficial for extinguishing the arc.

Further developments f the interrupting chambersed to some designshatincorporated cross blast patterns, while others included whats known as com-pensating chambers wherean intermediate contact s used to establish the arcsequentially. The first contact draws the arc in an upper chamber which pre-heats the oil prior to opening the second contact.

A typical relationship between the arcing time as a function of the inter-

rupted current and as a function of the system voltage was established by F.Kesselring [lo] and is shown in figure 5.24 (a) and@).

5.4.4 Bulk Oil Circuit Breakers

The main distinguishing characteristicf bulk oil breaker types s the fact thatthese circuit breakers use theil. not only as the intermpting medium but alsoas the primary means to provide electrical insulation.

The original plain break oil circuit breakers obviously belonged tohe bulk

oil circuit breaker type. Later, when the newly developed interrupting cham-bers where fitted to the existing plain break circuit breakers with no sig-nificant



7

9 5

F 4

$ 2$ 1

W 3

20 40 50 60

ARCING (

Figure 5.24 (a) Oil breaker arcing time as functionof current at constantvoltage.

W

a!i

20 25 30 35 40

ARClNG (milli-sec)

Figure 5.24 (b) Oil breaker arcing ime as unction of voltage at constant current.

modificationsbeing made specially to the oil tank, and because of the good ac-ceptance of this type of design, specially in the US; the bulk oil type concept

was simply continued o be fabricated.



Figure5.25 (a) 15 kV single tank oil circuit breaker .In many cases, at voltages that generally extendedup to kVathree

poles were enclosed into a single of oil, However a numberf breakers in

the medium voltage range had three independent as did those circuit

breakers with voltage ratings greater than45 kV. The three poles were gangoperated by a single operating mechanism. The single and the multiple tankcircuit breaker designsre shown in figure 5.25 (a) and(b).



Figure 5.25 (b) kV multitank oil circuitbreaker.

To meet the insulating needs of the equipment, and depending on themagnitude of the application voltage, adequate distances must be providedbetween the live parts of the device and he grounded containing the insu-

lating oil. Consequently type of design required large and large vol-umes of oil, for example for a kV circuit breaker approximately li-ters, or about gallons of oil were required, and for a kV circuitbeaker the volume was increased to 50,000 liters, or approximatelygallons.

Not only the size of the breakers was very large but the foundationpads where he breakers were mounted had toe big and quite strong. n orderto withstand the impulse forces developed during interruptiont is usually re-

quired that the pad e capable of supporting a force to up o times theweight of the circuit breaker including the weightf the oil. This in the caseof a kV circuit breaker amountedo a forceof about 50 tons.



5.4.5 Minimum Oil Circuit Breakers

Primarily in Europe, because of the need to reduce space requirements and tscarcity andhigh cost of oil, a new circuit breaker which usesery small vol-umes of oil was developed. This new circuit breaker is the one known by any

of the following names; minimum oil, low oil content, or oil poor circuitbreaker.

The main difference between the minimum oil and the bulk oil circuitbreakers is that minimum oil breakers use oil onlyor the interrupting functionwhile a solid insulating material s used for dielectric purposes, as opposed obulk oil breakers whereoil serves both purposes.

In minimum oil circuit breakers a small oil filled, arcntenupting chamberis supported within hollow insulators. These insulators re generally fabricatedfrom reinforced fiber glass for medium voltage applications and of porcelainfor the higher voltages.

Figure 5.26 Typical 15 kVminimum oil circuit breaker.



The use of insulating supports effectively u a l i i design as a livebreaker. By separating the live parts from ground by means of he insulatingsupport the volume of oil required s greatly decreasedas it can be seen in fig-ure 5.26 where a typical 15V low oil breakers shown.

5.5 Sulfurhexafluoride

Considering the fact that oil and air blast circuit breakers have been aroundoralmost one hundred years; sulfurhexafluoride circuit breakers are a relativelynewcomer, having been commercially introducedn 1956.

Although SF6 was discovered in 1900 byHenry Moissan [l l] The firstre-ports of investigations made exploringhe use of SF, as an arc quenching me-dium was published in 1953 by T. E. Browne, A. P. Strom and H. J. Lingal[12]. These investigators made a comparison ofhe intermpting capabilities ofair and using a plain break intermpter.

The published results, showing the superiorityf SF, were simply astound-ing. As it can be seen in figure 5.27 SF, was 100 timesbetter than air. In thesame report t was shown that he addition of even moderate rates of gas flowincreased the interrupting capability by aactor of 30.

60 80

Figure 5.27 Comparisonof ntemptingcapability betweenSF, and air (from ref. 12).



SF6 circuit breakers, n their relatively short existence already have comeocompletely dominatehe highvoltage circuit breaker mark& andn the processthey have made obsoletehe air blast and oil technologies. Almost without ex-ception SF, circuit breakers used in all applications involving system volt-ages anywheren the range of72.5 kV to 800 kV.

Inmedium voltage applications, fromkV nd up o about 20 kV, F6has

found a worthy adversary in another newcomer the vacuum circuit breaker.

Presently neither technology has becomehe dominant one, althoughhere arestrong indications hat for medium voltage applications vacuum maye gain-ing an edge.

5.5.1 Properties of SF6

SF6 is a chemically very stable, non-flammable, non-corrosive, non-poisonouscolorless and odorless gas. It has a molecular weight of 146.06 and is one ofthe heaviest known gases. The high molecular weight and its heavy density

limits the sonic velocityof SF6 to 136 meters per second which is about onethird thatof the sonic velocityf air.

SF, is an excellent gaseous dielectric which, under imilar conditions, hasmore than twice the dielectric strengthf air and at three atmospheres of abso-lute pressure it has about the same dielectric of oil (figure 5.28).Furthermore, it has been found thatF, retains most of its dielectric propertieswhen even with substantialproportions of air, or nitrogen.

250

200

150

100

50

I V Y W

b b b b b

0 200 400 60000 1000 1200

ABSOLUTEPRESSURE n kPa

Figure5.28 Dielectric of SF, as function ofpressure.



.0.1 ,

* . . . . . a . . . . .10 100

TEMPERATURE (Deg. C )

Figure 5.29 Heat transfer characteristics of SF,.

Because of its superior heat transfer capabilities, whichre shown in figure5.29, SF, is better than air as a convective coolant. It should be noted thatwhile the t h m a l conductivity of helium is ten times greater than that fthe laterhas better heat transfer characteristics due to the higher molar heatpacity of SF, which together with its low gaseous viscosity enablest tofer heat more effectively.

SF6 is not only a good insulating gas but it is an efficient electronscavenger due to its electron or electronegativity. property isprimarily responsible for its high electric breakdown strength, but t also pro-

motes the rapid recovery of the dielectric strength around the arc region fol-lowing the extinction the arc.

Because of its low dissociation temperature andts high dissociative energySF, is an excellent arc quenching medium. Additionally, the outstanding arcextinguishing characteristics of SF, are also due to the exceptional ability of

gas to recover its dielectric strength very rapidly following a period ofarcing,and o tscharacteristicallysmall imeconstantwhichdictates hechange conductance near current

The first characteristic is important for bus terminal faults while the secondis essential for the successful interruptionf short line faults.



5.5.2Arc Decomposed By-products

At temperatures above SOO(C. SF, will begin to dissociate. The process ofdissociation canbe initiated by exposing F, to a flame, electrical sparking,ran electric arc. During process the SF6 molecules will be broken downinto sulphur and fluorine ionst a temperature f about 3000(C.

It shouldbe recalled that duringhe interruption process the core ofhe arc

will reach temperatures well in excess of lO,OOO(K; However, after the arc isextinguished and he arc region begins to cool down and whenhe temperaturedrops below approximately l,OOO(C the gas will begin to recombine almosttotally, and only a small fraction will react with other substances..

The small amountsof gas that do not recombine react with ir, with mois-ture, with the vaporized electrode metal and with some of the solid materialsthat are used in the construction of the circuit breaker. These decompositionby-products may be gaseous or solid, but they essentially consistf lower

fur fluorides, andof metal fluorides of whichhe most notablesare CuF,, A F 3 ,

Among the secondary sulphur-fluorides compounds that are formed [l31

are S2F2 andSF,, but they quickly react with moistureo yield hydrogen fluo-ride sulfur dioxide (SO,) and other more stable oxyfluorides such as

thionyl fluorideSOF,.The metallic fluorides are usually present in the form of a fine non-

conductive dust powder thats deposited on he walls and in the bottom of the

breaker enclosure. In the case of copper electrodes he solid substances appearas a milky white powder that acquires some blue tinges when exposed to he

atmosphere due to a reaction which yields a dehydrated salt.

W S , CF4.

5.5.2.1 CorrosiveEffects of F,

Sulphur-hexafluoride on its pure and uncontaminated form is a non-reactivegas and consequently theres no possibility for any type of corrosionhat may

be directly attributableo SF6.When the by-products of arced SF, come in contact with moisture some

corrosive electrolytesmay be formed. The most commonly .used metals gen-erally do not deteriorate and remain very stable. However, phenolic resins,glass, glass reinforced materials and porcelain cane severely affected. Othertypes of insulating materials such s polyurethane, Teflon(PTEE) and epoxies,either of the bisphenol A orhe cycloaliphatic type, are unaffected.

It is therefore very important to take appropriate measuresn the selection

of materials, and the utilization of protective coatings. Corrosion also beprevented by he elimination of moisture.



5.5.2.2 By-productsNeutralization

The lower fluorides and manyf the other by-productsare effectively neutral-ized by soda lime (a 50-50 mixture of NaOH + CaO), by activated alumina(especially, driedA1203),r by molecular sieves.

The preferred granule size for soda lime or alumina is 8 to 12 mesh, butthese do not exclude the possible useof other mesh sizes. The recommended

amount to be used is approximately equal o 10 of the weight of the gas.Removal of the acidic and gaseous contaminants is accomplished by

circulating the gas through filters containing the above described materials.These filters can eitherbe attached to the circuit breaker itselfor they may beinstalled in specially designed but commercially available gas reclamationCartS.

If it is desired to neutralize SF6which has been subjected o an electric arc,it is recommended that the parts be treated with an alkaline solution of lime

(Ca(OH),), Sodium Carbonate (N&C03), or Sodium bicarbonate (NaHCO,).5.5.3 SFs Environmental Considerations

The release of human made materials into the atmosphere has created twomajor problems. One is the depletion of the stratospheric ozone layer and heother is the global warmingor “green house effect.”

Ozone Depletion Agent. SF6 does not contribute o the ozone depletion fortwo reasons: First because due to the structure of the ultraviolet absorption

spectrum of SF6 the gas not be activated untilit reaches the mesosphere atabout 60 kilometers above the and altitude is far above the strato-spheric one which is in the range of about to 45 kilometers. The secondreason is the fact that SF6 does not contain chlorine which is the principalozone destroying agent.

Green house eflects Agent. SF, has been labeledas the most potent green-house gas ever evaluated by the scientists of the Intergovemmental Panel on

Climate Change @CC) 141 [151.What makesSF6 such a potentiallyow& contriiutor to global warming

is the fact thatSF6, l i e all the compounds in he fully fluorinated famjly, has asuper stable molecular structure.This structure makes these compounds o bevery long lived, to the extent that within human time frames these gases areindestructible.

SF6 is a very good absorber of infrared radiation. This heat absorptioncharacteristic combined with its long life (3,200 years) [l61 has led scientists

to assign an extremely high Global W arming PotentialGWP rating to SF6.The GWP rating is a comparative numerical value that is assigned to a

compound. The value is arrived at by integrating over a time span the



tive forcing value produced y the release of 1 kg of the gas in question andthen dividing value by the value obtained with a similar procedure withCO2 Because CO2is considered the most common pollutant, it has been se-lected as the basis of comparison for assigning GWP values to other pollutants.

The radiative forcing, accordingo its definition, is the change in net irra-

diance in watts per square meter.The GWP values for CO2and for the most common fully fluorinated com-

pounds integrated over a one hundred years time horizonre given in Table 5.1(taken from reference [14]).

TABLE 5.1

Global Warming Potential (GWP)for most commonC's

compared to

COMPOUND LIFETIMEYEARS GWP

"200

6,300

3,2004,900

3.200 h

12,500

Presently the concentration of SF, has been reported as being only about

parts per trillion by volume This concentration is relatively low,but it has been observed that it is increasing at a rate of about 8 % per year.This means that if the concentration continues to increase at this rate, in lessthan 30 years the concentration could e about 50 pptv.

More realistically, assuming a worst case scenario [16], the concentrationof 50 pptv is expected to be reached by the year 2100. A more optimistic es-timate is 30 pptv. At these concentrations the expected global wanning at-tributable to SF, has been calculated as and 0.014 for the most pessi-

mistic and he most optimistic scenarios respectively. Additional ata indicatethat the expected global warming due to SF, through the year 2010 is about0.004 In comparison withan increase of 300 parts per million by volume(ppmv), ofCO2,the expected change in the global temperatures 0.8 "C.

It is apparent that based on the estimated emission rates the concentrationof SF, would be very small 17]. Nevertheless, because f the long life time ofSF, there is a potential danger, speciallyf the rate of emissions where to in-crease rather thano reach a level value. It s therefore essential that all types

of release of SF, into the atmosphere be eliminated or at least reduced to anabsolute minimum. This can be done by strict adherence to careh1 gas han-d in g procedures and proper sealingor all new product designs.



Figure 5.30 Electric arc, adial temperatureprofile.

5.5.4 Current Interruption in SFs

As we know, the electric arc is a self-sustaining discharge consisting of aplasma that exists in an ionized gaseous atmosphere. We also know that theplasma has an extremely hot core surrounded byn atmosphere of lower tem-perature gases.

Figure 5.30 represents the typical temperature profile of an arc as a func-tion of its radius, when the arc is being cooled by conduction. The figureshows that theres a relatively central regionof very high temperature cor-responding to the core of the arc. It also shows the existence of a broader,lower temperature region and the transition point between these two regions,where there s a rather sharp increase in temperature.

characteristic temperature profile simply indicateshat the majority ofthe m e n t is canied by the hottest region of the arc's core which is located

close to the central the reason being, as we well know hat an increase intemperature corresponds to an increase in electrical conductivity.

Since the arc always tries to maintain ts thermal equilibrium, its tempera-ture will automatically adjust itselfn relation to the m e n t magnitude. How-ever, once full ionization is attained further increases n current do not lead toincreases in temperature.Nevertheless, as the currentapproacheszero thetemperature about he core of the arc begins to drop and consequently he re-gion losing its conductivity.

The peak thermal conductivity ofSF,, as is seen in figure occurs ataround 2,000 OK; herefore, near current zero, when rapid coolings needed for



2 8 124

TEMPERATUREDEG. ~ ~ 1 0 3

Figure5.31 Themal conductivity of SF,.

intermption, SF, is extremely effective because at this temperature electricalconductivity is very low.

At the other side ofhe spectrum, at high currents he thermal conductivityof SF, is not much different from that of other gases and therefore the arc

cooling processin that region is about the same regardless of the kind of gasthat is being used.

The main difference between interruption inir and in is then the tem-perature at which maximum thermal conductivity takes place. These tempera-tures are about 6,000 "K for air and 2,000 OK for SF,.

difference translates into the fact that is capable of cooling muchmore effectively than air at the lower temperatures and therefore t is capableof withstanding higher recovery voltages sooner.n other words he time con-

stant ofSF, is considerably shorter than thatf air.The assigned time constant for is 0.1 microseconds while for air is

greater than 10 microseconds. The significance of this time constant is appre-



ciated when considerations given to applications wherea high rate of "RV sexpected, suchas in the case of short line faults. Experience indeedhas shownthat SF6 withstand higher recovery rates thanir.

5.5.5 Two Pressure SF6 Circuit Breakers

The first SF6 circuit breaker rated for application at voltages higher thankV and a current interrupting capabilityf kA was commercially introduced

by Westinghouse in 1959.The original design of type of circuit breaker was n adaptation of the

air blast and oil circuit breaker designs, and thus the axial blast approach,which was described before when discussing air blast breakers, was used.Naturally, the main dif€erence was, thatirhad been replacedy SF,.

The new circuit breakers were generally of the dead type. The con-struction of the together with their substantial size and strength wasquite similar tohe used for oil breakers as it can be seen in figure

In many cases even the operating mechanisms that had been used for oilcircuit breakers where adapted to operatehe SF, breaker.

The conscious effort made to use some of the ideas from the older tech-nologies is understandable, after all, the industry was accustomed o typeof design and by not deviating radically fromhat idea made it easier to gainacceptance for the new design.

Figure 5.32 Cutaway view of an SF, two pressure type GA circuit breaker.



1.6

; .4

1.2

E

0.:

OXi

11)

0.4

0.2

0

Figure5.33 Pressure-Temperature variation at constant density forF,.

Two pressure circuit breakers were fabricated in either a singleor a threetank version, depending mainly in the assigned voltage rating of the device.Smallerhigh pressure reservoirs were installed next to he low pressure

and they were connected to blast valves that operatedn synchronism with hecontacts. The operating gauge pressures for these circuit breakers were gen-erally around 0.2 Mpa for the low side and 1.7 MPa for he high side(30 psigand 245 psig respectively).

The two pressure circuit breaker design prevailed n the US market untilthe mid-nineteen seventies. At around that time is when the single pressurebreakers began o match the interrupting capabilities f the two pressure circuitbreakers and thus they became a viable alternative.

Cited among he advantages of the two pressure circuit breaker was thattrequired a lower operating energy mechanism when comparedo the one thatis used on single pressure breaker designs. However, in the context of totalenergy requirements, one must take into account the energy that is spent in

compressing the gas for storage and also he additional energy thats requiredto prevent liquefaction f the SF, at low ambient temperatures.

The liquefaction problem representshe main disadvantageof the two pres-sure breaker. As it can be seen in figure 5.33 at 1.7 Mpa the gas will begin to

liquefy at a temperature of approximately3 'C. To prevent liquefaction, andthe consequent drop n the gas density electric heaters re installed in the highpressure reservoir.



Liquefaction of not only lowers the dielectric capabilitiesf the gas butit lead to another problem known as moisture pumping [l81 which mayhappen because of the difference in the condensation point between air and

The problem beginsn the high pressure system whenhe gas liquefies n aregion that s some distance away fromhe high pressure reservoir.If the tem-perature is not sufficiently low to cause condensation of whatever amount of

moisture was present n that region then only he liquefied gas will flow backinto the reservoir leavinghe moisture behind

Since in he mean time, he temperature of the gas in the high pressure res-ervoir is kept above he dew point, then, he warmer gas will flow back intohebreaker attempting to maintain the original pressure. Whatever small amountof moisture s present in the gas contained inhe reservoir it will thenbeported to the region where liquefaction is taking place. As the gas liquefiesagain, then once moret will leave he moisture behind. This process con-

tinue until the pressure-temperature conditions change. However, duringtime, moisture accumulate significantlyat the coldest point of the gas sys-

tem, thus increasing the total concentration and reducing the dielectric capa-bility.

other disadvantages noted are; the high volumes of gas needed, the pro-pensity for higher leak rates due to the higher operating pressures and theadded complexity that results fromhe use of the blast valves.

5.5.6 Single PressureSF6Circuit BreakersSingle pressure breakers have been aroundt least as long as the two pressurebreakers, but initially these breakers where limited to applications requiring

Later investigations and advanced developments provided answershat led

to new designs that had greater interrupting capabilities and around the year1965 high interrupting capacity puffer breakers were introducedn Europe andin the US.

Puffer circuit breakers have been designeds either dead or live as il-

lustrated in figures 5.34 and 5.35.Customarily, single pressure circuit breakersre d e s m i d as belonging to either

the pufkr or the self blast family.ut, in realityyll single pressure circuit breakerscould be thought as being a member of the self blast family becausen either typeof circuit breaker the increasenpressure that place inside of the interrupter iachieved without the aid of external gasompressors.

The most notable difference between these two breaker types is that inpuffer breakershe mechanical energy provided byhe operating mechanism sused to compress the gas, while self blast breakers usehe heat energy that isliberatedfrom he arc to raise he gas pressure.

SF,.

lower interrupting ratings.



Figure 5.34 Dead puffer typecircuitbreakerABB Power T&D.

Figure 5.35 Merlin live tank puffer circuitbreaker.



PufSer Circuit Breakers. The conceptual drawings and he operating sequenceof a typical puffer interrupter s shown in figure 5.36 (a), (b), (c), and (d). A

unique characteristic of puffer interrupterss that all have a piston and cylindercombination which is assembled as an integral part of the moving contactstructure.

Referring to figure 5.36, (a) shows the intermpter in the closed position,where the volume (V) be seen. During an opening operation the main

contacts separate first, followed by the arcing contacts, figure 5.36 (b). Themotion of the contacts decrease the dead volume (V), and thus compress thegas containedwithin that volume.

As the continue to separate the volume s further and atthe instant when the arcing contact leaves the throatf the nozzle the flow of asalong the of the arc is initiated. It is important to recognize hat at highrents the diameterof the arc may be greater the diameter of the nozzle thusleading to the conditionnownasCUITent choking. When thishappens he nozzles

completely blocked and theresno flowof gas. Consequently, the pressure contin-ues to rise due to the continuous changef the volume spaceV and to the heatn-ergy that s extracted from thercby the trappedas.

It is not uncommon to see that when interrupting large speciallythose corresponding to a three phase fault, the opening speed of the circuitbreaker is slowed down considerably due to the thermally generated pressureacting on he underside of the piston assembly.

However, when the currents to be interrupted are low, the diameter of the

arc is small and therefores incapable of blocking the gas flow ands a result alowerpressure is available. For even ower currents, as is the casewhenswitching capacitor banksof just simply a normal load current,t isgenerallynecessary to precompresshe gas before he separation of the contacts. This isusually accomplished y increasing the penetrationf the arcing contact.

The duration of the compression stroke should alwaysbe mefully evalu-ated to ensurehat there is adequate gas flow throughout the rangef minimumto maximum arcing time.h most cases, dependingon the breaker design, he minimum arcing time

is in the range of 6 to 12 milliseconds. Since the maximum arcing time sproximately to heminimumarcing imeplusoneadditionalmajorasymmetrical current loop, which has an approximate durationf 10 millisec-onds, then he range of he maximum arcing times 16 to milliseconds.

What is significant about he arcing time durations that, since interruptiontake place at either of these times, depending only n when a current zero

is reached, then what is necessary is that the appropriate pressuree developedat that proper instant where interruption takes place.

It is rather obvious thatat the maximum arcing time, he volume has gonethrough the maximum volume reduction and has hadhe maximum time expo-



Figure 5.36 Puffer circuit breaker principle (a) breaker closed (b)

main contacts separate, (c) arcing contacts separating gas flow (d) interruption

completed.Legend A=Arc, V= Puffer Volume, P= Puffer 6 & 8 = Arcing Contacts, 9 =

Interrupter Nozzle.



to the heating actionof the arc and thus the gas is expected to behigher. For the arcing time condition, both the compression and the hing of the gas are and therefore the pressureenmted is relatively low.

It follows, from the above discussion, that the critical gas flow conditionfor a puffer interrupter is around the region of the minimum arcing time.However, it also points out that consideration must be given to the openingspeed of the breaker in relation tots opening stroke n order to assure thathe

assumed maximum arcing time is always less than he total travel timeof theinterrupter.

It was mentioned before that when interrupting large currents in a threephase fault, there is tendency for the breaker to slow down and even to stallsomewhere along ts opening stroke. This slowing down generally assureshatthe current zero corresponding to he maximum arcing time is reached beforethe circuit breaker reaches the end ofts opening stroke. However, when inter-rupting a single phase fault thats xot the case. That is, because during a sin-

gle phase fault the energy input from the fault current is lower which repre-sents a lower generated pressure and so the total force that is opposing thedriving mechanism s much lower. Therefore it is quite important to carefullyevaluate the single phase operation o assure that there is a sufficient overlapbetween e he stroke of the puffer (breaker) and the maximum arcing time.

Serf Blast Circuit Breakers. Self blast circuit breakers, ake advantage of thethermal energy released by he arc to heat the gas and to raise its pressure. In

principle the self blast breaker idea is quite similar to the concept of the ex-plosion pot is used by oil circuit breakers. The arc is drawn across a pair ofcontacts thatare located insideof an arcing chamber andhe heated high pres-sure gas s released alongsideof the arc after the moving contact s withdrawnfrom the arcing chamber.

In some designs to enhancehe interrupting performance, n the low ament

range, a puffer assist is added. In other designs a magnetic coil is also in-

cluded [19]. The objectof the coil is to provide a drivingorce that rotates he

arc around the contacts providing additional cooling of the arc as it movesacross the gas. In addition to cooling the arc the magnetic coil helps todecrease the rate of erosion of the arcing contacts and thus it effectively ex-tends the life of the interrupter. In some designs a choice has been made tocombine all of these methods for enhancing the interruption. process and inmost of he cases has proven o bea good choice. A cross section f a selfblast interrupter pole equippedith a magnetic coilss included in figure

5.5.7 Pressure Increaseof SF6Produced by an Electric Arc

The pressure increase produced by an electric arc burning inside of a smallsealed volume (constant volume) filled withF, gas canbe calculated with a



1.Expansion cylinder2. Fixed arcing contact3. Moving arcing contact4. Coil5. InsulatingSpacer6. Fixed main contact7. Moving main contact8. Exhaust volume

Figure5.37 Outlie of a self blast circuit breaker pole.

ream able degree of accuracy using the curve givenn figure 5.38. This curvewas obtained by solving the Beattie-Bridgmanequation, and by assuming a con-

value of 0.8 Joules per -degree C for the heat capacity at constant vIt is of c o m e assumption which will introduce somerrors because the

value of C, increases with tempmture. However theresults be corrected bymultiplying the change by the mtiof the C, o the actualC,. Values ofC,asa functionof temperature are in figure

To calculate the approximate increase in pressure produced by arcing thefollowing procedure maye used.

1. Estimate thearc energy input to the volume. The energy input willbeproximately equal to the productf the average arc voltage times thevalue of the current times therc time duration.For a more accurate calculation the following expression maye used:

0

where:



Q. =Arc Energy inputn joulesE, =Arc VoltageI,,, sinat = Current being interruptedt = arcing time

2. Find the value of the quotient of the arc energy input, divided by the vol-

Find the gas density or a constant volume t normal gasfilling conditions

ume, in cubic centimeters,of the container

using the ideal gas law which says:

’ n per cubic centimeterR x T

where

M= molecular weight ofF, = 146 gP =Absolute pressure n kiloPascal

R =Gas constant= C.C. -kilo Pascalper mole - OKT= Temperature degree Kelvin

4. using the just calculated density extract the factor for the pressure risefrom figure and multiply t by the energy per unit volume obtained nline 1above.

Figure 5.38 Pressure! ncrease or a constantSF, volume produced by arcing.



103

TEMPERATURE "C

Figure 5.39 Coefficientof heat capacity CVor SF6 at constantvolume.

5.5.8 Parameters InfluencingSFs Circuit Breaker Performance

Pressure, nozzle diameter, and rate of change of current were the parameterschosen before as the base for evaluating the recovery capabilities of air blastcircuit breakers. To facilitate the comparisons between the two technologiesthe same parameters will nowe chosen for SF, interrupters and he results areshown graphically n figures 5.40, 5.41and 5.42. [21]

Once again, the significance of the performance relationships which areshown in the above figures, lies not in their absolute values but in thethat they predict. In figure 5.40, for example, it is to see what is intui-tively clear, which is that in order to interrupt larger currents,a larger nozzlediameter is required. It can be seen the effects of the nozzle diameter acurrent magnitude; the smaller he current, the lesser the influence the noz-zle diameter. The curve even suggests that there may be a converging point

for the nozzle diameters, where at a certain level of smaller currents, the re-covery capabilities of the interrupter remainshe same regardlessof the nozzlesize.



Figure 5.40 Interrupting relationship between c w e n t and voltage for various nozzle

diameters.

Figure 5.41 shows the dependencyf the recovery voltage during he ther-

mal recovery periodn relation to he rate of change of current. It s importantto note that he slope of each f the lines are remarkably close considering thatthey represent three independent sources f data extracted from references 7]and [21]. These curves indicatehat the rate of recovery voltage n the thermalregion is proportional to the maximum rate of change of current (at = 0)

raised to the -2.40 power. The 2.40 exponent compares with he 2.0 exponent

obtained with airblast interruptersIn figure 5.42 we find a strong dependencyf the recovery on the gas pres-

sure as evidenced by the equation defininghe relationship which indicateshatthe recovery is proportional to the pressure to the 1.4 power. In com-parison the slope corresponding o the same relationship curves but withir asthe interrupting medium s equal to 1.0.

The results observedfor the dependency on the rate of change of currentand onp re s s u r i c o n f i i what we already know; which is, thatt the same

rent magnitude and at the same pressure, SF, is a better interrupting mediumthan air.



10

1

0.1 1 1 110 100

dvdt (amperes microsec.)

Figure 5.41 Comparison o f interrupting capability of SF, using from inde-

pendent sources.

Figure 5.42 Dependency of the recovery voltage upon pressure forF,.



5.5.9 SFcNitrogen Gas Mixture

Because of the strong dependence of SF6 onressure it hasalways been mvdentto increase theressure in order to impmve the recovery chamcteristics ofhe inter-rupter. However, as it hasbeen before thereare limitations bythe Operating ambient temperature to avoid liquefaction of theas.

To overcome the problem the possibility of mixing Nitrogen(N2)with SF,

has been investigated. A lthough today the issue is only academic when refer-ring to wo pressure breakers since they are not manufactured any longer,t hasbeen demonstrated that he performance of a tw o pressure circuit breaker wasimproved, as shown in figure5.43, when at the same total pressure a mixtureby pressure of 0 YO F6 and50YO 2was used [22],

For single pressure circuit breakers has not been the case, and this isattributed to the fact that suffkient pressure not be sustained due to thehigh flow ratesof lighter gas mixture. In two pressure breakers maintain-

ing the pressure differential high enough is not a problem because the highpressure is maintained in the high pressure reservoir by meansof an externalcompressor.

From the point of view of dielectric withstand no significant differencesfound with mixtures containing a high percentagef N2. For example, with a40% N2 content the dielectric withstands reduced by only about 10%.

0.1

1 1.2 1.4 1.6 2 2.2

PRESSURE (mmgspasab)

Figure 5.43 Intermpting capability for SFGN2mixtures.



5.6 Vacuum Circuit Breakers

Vacuum interrupters take advantage of vacuum because of its exceptional di-electric characteristics and of its diffusion capabilities as an interrupting me-dium. It should be noted that the remarkable dielectric strength of vacuum isdue to the absence of inelastic collisions between the gas molecules which

means that theres not an avalanche m echanism to trigger the dielectric brdown as is the casen gaseous mediums.The pioneering work on the development f vacuum interrupters wascar-

ried out at the California Institute of Technology by R Sorrensen and H.Mendelhall as reported in their paper

Despite the early workt was not until the that thefirst commerciallyviable switching devices where introduced by the Jennings Company, and

when the General Electric Company introduced the medium voltage

power vacuum circuit breaker.Whatprevented he earlier introduction of vacuum nterrupterswheretechnical difficulties that existed n areas such as the degassing of the contactmaterials, which is a process thatis needed to prevent the deterioration of theinitial vacuum due to the release of the gases thatre normally trapped withinthe metals. Another problem was the lackf the proper technologies neededeffectively and reliability weldor braze the external ceramic envelopes to hemetallic endsof the interrupters.

In the last years these problems have been solved and that, coupledthedevelopment of highlysensitive nstrumentationhavesubstantially n-creased the reliability for properly sealing the interrupters to prevent vacuumleaks.

In the there were some attempts made to develop vacuum circuitbreakers for applications at voltages greater than kV. However these de-

were not suitable to compete with SF, circuit breakers and vacuum hasbeen relegated primarily to applicationsn the rangeof to kV.

In theUS vacuum is used most of the ime for indoor applicationsat 5 andkV, and at these or similar voltagest also has the larger share of the maworldwide.

5.6.1 Current Interruption n Vacuum Circuit Breakers

The characteristics of a vacuum, or a low pressure arc, were presented inchapter one. In section the current interrupting process that takes placena vacuum interrupter wille described.

Current interruption, like in all circuit breakers, is initiated by the separa-tion of a pair of contacts. At the time of contact part a molten metal bridgeappears across the contacts. After the rupture of the bridge a diffuse arc col-



umn is formed and he arc is what is called a diffuse mode.This mode is char-acterized by the existence of a number offast moving cathode spots, whereeach spot shares an equal portion ofhe total current. The current that isried by the cathode spot depends on he contact material and or copper elec-trodes a current f about 100 amperes er spot has been observed. he arc willremain in the diffuse column mode until the current exceeds approximately 15k A . As the magnitudeof the current increases a single anode spot appears thus

creating a new source of metal vapors which because of the thermal constantof the anode spot continues to produce vapors even after current zero. Withthe reversal of current, following he passage through zero and because f ionbombardment and a high residual temperature it becomes quite to rees-tablish a cathode spott the place of the former anode. M. B. Schulman et. al.[25] have reported n the sequenceof the arc evolution and have observedhatthe development is sensitive to the methodf initiation.

During normal interruptionof an ac current, near current zero he arc col-umn will be diffuse and will rapidly disappear in the absence of current.Since, during interruption and depending in the current magnitude,he arc mayundergo the transition from the diffuse mode to the constricted mode and bacagain to the diffuse mode just prior to current zero it becomes clear that thelonger the rc is in the diffuse mode he easier it is o interrupt he current

What it is important to realize from the above is the desirability ofmizing the heating of the contacts and maximizing he time during which thearc remains d e s e during the half current cycle. This objective can be ac-complished by designinghe contacts in such way that advantage beof the interaction that exists betweenhe current flowing through the arc andthe magnetic field produced by the current flowing the contacts orthrough a coilhat may be assembled as an integral partof the interrupter [26].

Depending in the method used, the magnetic field may act in a transverseor in the axial direction with respect tohe arc.

Transverse Field To create a transverse or perpendicular field different designsf

spiral contacts, such as those illustmted in figure5.44, have beenused. In the dif-fuse mode the cathode move freely the surface of the cathode elfximdeas if it was a solid disk.. At higher and as the arc becomes coalescent themagnetic field produced by the current flowing through the contactpirals forcesthe arc o move along them as a result of the magnetic forceshat are exertedon the an: column as shown in figure5.45. As the arc mtaks its roots also movealong reducing the likelihood of formingtationary spots and reducing the 10-4heating of the electrodes and thus also reducing the emission of metallic vapors.

When he end of the contact spirals is reached, thearc roots,due to the magneticforce on the arc column are forced tojump he gap and to continue them-tation along thepirals of the contacts.



Figure 5.44 Two types of spiral contacts used on vacuum interrupters.

i

Figure 5.45 Magnetic forcesin a transverse magnetic field.



Figure 5.46 Vacuum arc under the influenceof a transverse magneticield

(a) constricted column 18.8 ka peak), (b)arc showing two parallel d i h e columns.

(Courtesy of Dr.M. B. Schulman, Cutler-Hammer, Horseheads, NY.)



Figure 5.47 vacuum arc under he influenceof a transverse magnetic field

jet column with wedgenstability (26.8 ka peak), (b) 7 kA diffuse arc followingcurrent peakof18.8kA

(Courtesy of Dr.M. B. Schulman, Cutler-Hammer, Horseheads,



The effects ofhe field on he arc are illustrated in figures 5.46 (a), (b) and5.47 (a), and(b) where the photographs f an arc n the diffuse and constrictedmodes are shown.

Axial Field. The axial magnetic field decreases he arc voltage and he powerinput from the arc by applying a m agnetic field that effectively confines thearc column as it can be seen in he photograph of a 101 A peak diffuse arcas

shown in figure.48.

In the absence of the magnetic field, dff is ion causes the arc to expandoutwards from the space between the electrodes. However, when the axial

magnetic field s present the ion trajectory becomes circumferential and a con-fining effect is produced. For a reference on the effects of the axial magneticfields upon he arc column and on the formationf the diffuse arc one can re-fer to the work published byM. B. Schulman et. al. [28].

Axial magnetic fields cane produced by using either, aoil that is located

concentrically outside the envelope of the interrupter and hat is energized bythe current flowing through he circuit breaker [29], or by using specially de-signed contacts such as the one suggested by Yanabu et al and which isshown in figure 5.49. Observing at figure, it can be seen the action ofmagnetic force on the arc column as the result of the interaction ofhe mag-netic field set p by the current flowing throughhe of the coil electrodeand the contact.

Figure5.48 High Current(101 kA peak) diffuse arc inanaxialmagnetic field(Courtesy of Dr. M. B. Schulman,Cutler-Hammer, Horseheads,



Figure 5.49 Direction of the force on the arc producedy anaxial magnetic field.

5.6.2 Vacuum Interrupter Construction

Vacuum interrupters are manufactured by either of two methods. The differ-ences between methods re mainly the procedures used to braze and o evacu-ate the interrupters.

In one of the methods, which s the one commonlyknown as the pinch-offmethod, the interrupters are evacuated individually in a pumping stand afterthey are completely assembled. An evacuation pipe is located a t one end of

the interrupter, generally adjacent to the fixed contact and after the requiredvacuum is obtained the tubes sealed by compression welding.

With the secondmethod the interrupters are concurrentlybrazedandevacuated in specially designed ovens. The advantage of this method is thatevacuation takes place at higher temperatures and therefore there s a greaterdegree of vacuum purity n the assembly.

The interrupter,as shown n figure 5.50 consists of a ceramic insulatingn-velope that is sealed at both ends by metallic steel) plates brazed o

the ceramic body so that a thigh vacuum container s created. The operatingambient pressure inside of the evacuated chamber of a vacuum interrupterisgenerally etween nd IO-' torr.

Attached to one of the end plates is the stationary contact, while at theother end the moving contact is attached by means of metallic bellows. Thebellows used maybe either seamless or welded, however the seamless varietyis usually the preferred type.

A metal vapor condensation shield is located surrounding the contacts ei-

ther inside of the ceramic cylinder, or in series between two sections of the



Figure 5.50 Vacuum interrupter construction.

snsur;iting container. The p q o s e of the shield is to provide a surface where the

metal vapor condenses thus protecting the inside walls of the insulating cylindero

that do not become conductivey virtue of the condensed metal vapor.A second shield s used to protect the bellows fromhe condensing vapor o

avoid the possibility of mechanical damage. In some designs there is ashield that s located at the junction of the stationary contact and he end plateof the interrupter. The purposeof th is shield is to reduce he dielectric stressesin h i s region.

5.6.3 Vacuum Interrupter Contact MaterialsSeemingly contradicting requirements are imposed upon the possible choicesof materials that re to be used for contacts in a vacuum interrupter and there-



fore the choice of the contact material ends up being a compromise betweenthe requirements of the interrupter andhe properties of the materials that arefinally chosen l] .

Among the most desirable properties ofhe contact materialare the follow-ing:

1.

4.

5.

7.

8.

9.

A material thathas a vapor pressure thats neither too low noro high. A

low vapor pressure means that the interrupter is more likely to chop thecurrent since theres not enough vaporo maintain the arc at low values ofcurrent.A high vapor pressure, n the other hand, is not very conducive for inter-rupting high currents because there wouldtill be a significant amount fvapor remainingat current zero, thus making interruption difficult.A material that has a good electrical conductivity is desired in order tominimize the losses during continuous operationf the interrupter.

A high thermal conductivity is also desirable in order to reduce the tem-perature of the contacts and for obtaining rapid cooling of the electrodesfollowing the interruption of the current.Good dielectric properties are needed to assure rapid recovery capabHigh current interruption capabilities.A material that have a low weld strength is needed because contacts invacuum will invariable weld due o the pre-arcing that occurs when clos-ing or to the localized heating of the micro contact areas when the short

circuit current flows through the closed contacts. facilitate the openingof the contacts easily fractured weldsre a basic necessity.Mechanical strength s needed in the material mainly o withstand the im-pact forces, specially during a closing operation.,Materials with low gas content and easef outgassing desirable sincethe contacts must be substantially gas free to avoid the release of anygases from the contacts during interruption and thus to prevent loweringthe quality of the vacuum ambient.

prevent the new cathode from becoming a good supplier of electronmaterial with low thermionic emission characteristicss desirable.

From the above given ist we can appreciate that therere no pure elementmaterials that can meet ll of these requirements. R ehc to ry materials such stungsten offer good dielectric strength, their weldsre brittle and thus areto break. However, they are good thermionic emitters, they have a low vaporpressure and consequently their chopping current levels high and their inter-

rupting capability s low.

quirements. Nevertheless its greatest disadvantage s that due to its ductility it

has a tendency to form very strong welds which are the esult of diffusion

In the other side of the spectrum copper appears to meet most of the re-



" A x i a l Field

10

ELECTRODE DIAMETER (mm)

Figure 5.51 Comparison of interruption capability for vacuum interrupters as func-tion ofelectrode diameter and magnetic field type.

welding. This type of welding occurs, specially inside of a vacuum atmos-

phere, when wo clean surfaces re pushed together and heated..Since anacceptable compromise material can note found among he pure

elements the attention has been directed to investigatehe use of sintered met-als or other alloys. A number of binary and ternary alloys have been studied,but from all of those. that have been considered two alloys, one a Cu-Bi(copper-bismuth) and the other a Cu-Cr (copper-chrome) alloy have prevailedand todayare the most commonly used.

In the Cu-Bi alloy copper is the primary constituent material and the sec-

ondary material s bismuth the content of which is generally up to a maximumof 2%. For the Cu-Cr alloy there aredifferent formulations bu t a typical com-position is a to 40% Cr combination.

In general Cu-Bi contacts exhibit a weld about 7 times lower

Cu-Cr but they have a higher chopping current level thanhat of Thetypical chopping level for Cu-Bi contacts is in the range of to 15 ampereswith a median value ofamperes, while or Cu-Cr is only between1 to 4 am-peres with median value f 2.7.

Other differences in performance betweenhe materials are the higher rateof erosion that is observed in Cu-Bi contacts and the decrease in dielectric



withstand capability that results by he cumulative process of the interruptingduties.

5.6.4 Interrupting Capability of Vacuum Interrupters

From the above discussions it is evident that the interrupting capability of avacuum interrupter depends more thann anything elseon the material and hesize of the contacts and in the type of magnetic field produced around thecontacts [32]. Larger electrodesn an axial field, s shown in figure 5.51, havedemonstrated that they have a better interrupting capability.

Another very important characteristic, relatedo the interrupting, or recov.ery capability, of vacuum interrupters is their apparent insensitivity to highrates of recovery voltage [33]. In [7] it is shown that within a frequency rangeof 60 to800 Hz., for a given frequency the TRV has only a weak effect onhecurrent magnitude. Furthermore it is widely recognized that the transient voltage recovery capabilityof vacuum interrupters s inherently superior to that fgas blast interrupters.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

ANSI IEEE C37.100-1981, Definitionor power Switchgear.Practical Applicationof Arc Physics n Circuit Breakers. oflation Methods and Application Guide, Electra (France) No. 118: 65-79,May 1988.

ANSI C37.06-1979 Preferred Ratings and Related Capabilitiesor ac HighVoltage Circuit Breakers Rated on a Symmetrical Current Basis.International Electrotechnical Commission (IEC), International Standard

IEC 56.G. Frind, Time Constant of Flat Arcs Cooled by Thermal Conduction,IEEETransactionsonPowerApparatusandSystems,Vol.84,No.12:1125-1131,Dec. 965.Em il Alm, Acta Polytechnica, L47 (1949) Electrical Engineering Series

Vol. 2, No. 6, UDC 621.316.5.064.2, Sweden Royal Academy of Engi-neering Sciences.D. Benenson,G.Frind, R E.Kissinger, H. T. Nagamatsu, H. 0.

Noeske, R. E. Sheer Jr., Fundamental Investigations of Arc Interruptionin Gas Flows, Electric Power Research Institute, EPRI-EC-1455,1980.Current Interruption n High Voltage Networks, ea. K Ragaller) PlenumPress, New York, 1978.RoyWilkins,E.A.Cretin,HighVoltageOilCircuitBreakers,Mac.

Graw Hill, 1930.10. F. Kesselring, Theoretische Grundlagenzur Berechnung der Schaltgertite,

Walter de Gruyter& Co., Berlin, 1968.



11. H. Moissan, P. C. LeBeau, Royal Acad. 130: 984-988,1900.12.H.J.Lingal, A. P. Strom, T.E.Browne, Jr., An Investigation of the

Arc Quenching Behavior of Sulfurhexaflouride, AIEE Trans. 72, Pt. 111:

242-246, 1953.13. Allied Signal, Accudri SF,, Technical Bulletin 97-0103.4M.595M 199514. Elizabeth Cook, Lifetime Commitments: Why Policy-Makers Can’t Af-

ford to Overlook Fully Flourinated Compounds, Issues & Ideas, World

Resources Institute, Feb. 1995.15. EPA, ElectricalTransmissionndDistribution ystems, ulfurhex-

aflouride ndAtmosphericEffectsofGreenHouse Gas Emissions,Conference Final Report,Aug. 9-10,1995.

16.Malkom K. W. KO, ienDakSze,Wei-ChyungWang,George Shia,

AaronGoldman,FrankJ.Murcray,DavidJ.Murcray, P. Rin-sland,AtmosphericSulfurhexaflouride:Sources, Sinks andGreenhouseWarming,JournalofGeophysicalResearch,Vol.98,No.D6:10499-

10507, June 20, 1993.17. G. Mauthe, K. Petterson, R. Probst, H.Brtiutigam,D.Ktining, L. Nie-

mayer, B. M. Pryor, CIGRE 23.10 Report.18. Isunero Ushio, Isad Shimura, Shotairo Tominaga, Practical Problems of

SF, Gas CircuitBreakers, PA&SVol.AS-90, NOS: 2166-2174,Sept/Oct. 1971.

19. G. Bernard, A. Girard, P. Malkin, An SF, Auto Expansion Breaker: TheCorrelation Between Magnetic Arc Control and Critical Current, IEEETrans. onPowerDel., Vol. 5, No. 1: 196-201, Jan. 1990.

20. R. Garzon, Increase of Pressure in a Vessel Produced by an Electric A rcin SF,, Internal Engineering Report 3032-3.002-E7, 1973.

21.R.D.Garzon, Rate of Change of Voltage and Current as Function ofPressure and Nozzle Area in Breakers Using SF, in the Gas and LiquidPhases, IEEE Transactions of Power Apparatus and Systems, Vol PAS-95,

22. R. D. Garzon, The Effects of SF,-N2 M ixture Upon he Recovery VoltageCapability of a Synchronous Intermpter, IEEE Transactions n Power Ap-paratus and Systems, Vol PAS-95, No. 1: 140-144, Jan./Feb. 1976.

23.Wang Enh i, Lin Xin, Xu Jianyuan, Investigation of the Properties ofS F d 2Mixture as an Arc Quenching Medium in Circuit Breakers, Proc.of the 10th.Int.Conf. on G a s Disch. and heir Appl., Vol. 1: 98-101,Swansea, UK, Sept. 1992.

24.R.W.Sorensen, H. E. Mendenhall, Vacuum Switching Experiments atthe California Instituteof Technology, Transactions AIEE 45,: 1102-1 105,1926.

25.M.BruceSchulman,PaulSlade,SequentialmodesofDrawnVacuumArcs Between Butt Contacts for Currents in the 1 kA to 16 kA Range,

NO.5:1681-1688,Sep./Oct.1976.



IEEE Trans. on Components, Packaging, and Manufacturing Technology-Part Vol. 18NO. 2: 417-422, June 1995.

26. R. Gebel, D. Falkenberg, Behavior of Switching Arc in Vacuum Inter-rupters Radial Field and Axial Field Contacts, ITG- Fachber (West Ger-many), Vol. 108: 253-259, 1989.

27.M.BruceSchulman,SeparationofSpiralContactsand the Motion ofVacuum Arcs at High AC Currents, IEEE Trans. on Plasma Scien, Vol.

21, No. 5:484-488, Oct. 1993.28. M.B. Schulman, Paul G. Slade, J.V.R. Heberlein, Effect of an Axial

MagneticFieldUpon heDevelopmentof heVacuumArcBetweenOpeningElectricContacts,IEEETrans. on Components,Hybrids,andManufacturing Technology, Vol. 16, No. 2, 180-189, March 1993.

29.H. Schellekens, K. Lenstra, .Hilderink, .erHennere, . Kamans,Axial Magnetic Field Type Vacuum Circuit Breakers Based on ExteriorCoilsandHorseShoes,Proc. X I 1 th, onDiel.Disch.and

Elec.Insulation,Cat.No.86CH2194-9:241-244,Shoresh,Israel,Sept.1986.

30. S. Yanabu,EKaneko,H. Okumura, ndT.Aiyoshi,NovelElectrodeStructure of Vacuum Interrupter andts Practical Application, IEEE Trans.Power App & Syst. Vol. PAS 100: 1966-1974, March-A prill981.

31.P.Slade,ContactMaterials for Vacuum nterrupters, EEETrans. OnParts, Hybrids, and Packaging, Vol. PHP-10, No. 1, March 1974.

32. Toshiba, Technical Bulletin, KSI-E1052-2,1983-6.

33. R. K. Smith, Test Show Ability of Vacuum Circuit Breaker to InterruptFast Transient Recovery Voltage Ratesf Rise of Transformer SecondaryFaults, IEEE Trans. on Power Del., Vol. 10, No, 1:266-273, Jan. 1995.

34. Allan Greenwood, Vacuum Switchgear, IEE Power Series 18, Short RunPress Ltd. Exeter England, 1994.



6

MECHANICAL DESIGN OF CIRCUIT BREAKERS

6.0 Introduction

The two most basic functions of a circuit breaker are to open and close theircontacts on command. This at fust sight implies a rather simple and trivialtask; however, many of the characteristics involved in he process of opening,closing and maintaininghe contacts closed can e quite demanding.

It is interesting to note that accordingo a CIGRE report[l] more than 90% of circuit breaker failuresare attributed to mechanical causes. These find-

ings confirm the fact that circuit breakers are primarily mechanical devicesthat are called upon o perform an electric function.The majority of the time circuit breakers remain closed and simply actas

electrical conductors, but in many occasions they do indeed perform their n-tended protective functions and when this happens, fromhe combined electri-cal and mechanical point of view, undoubtedlyhe contact structure s proba-bly one of the most essential and critical components. A second and equallyimportant component is the operating mechanism employed to produce the

motion of the contacts.These two components are closely linked to each other and in more ways

than one they determinehe success or failure of an interrupting device. Giventhe importance of these two components chapter will be dedicated to thediscussion of subjects relating o these components.

The most commonly used designs of operating mechanisms will be de-scribed in general terms, concentrating primarily n describing the operationalsequences, rather than dealing with specific detailsf how to design a particu-

lar type of a mechanism.The subject of electrical contacts will be treated in more detail so that a

better understanding is gained on area of design which tends to re sccu rfrequently, not only when dealing withhe development of new circuit break-

but when evaluating circuit breaker performancer special applications.

6.1 Contact Theory

Circuit breaker contacts must first be able to their assigned continuous

current rating, without overheating,r deteriorating and must doo withinrea-sonable limits f power consumption.

195



In addition, during short circuit conditions, they muste able to largecurrents for some specified periods f time, and again they must do withoutdeteriorating or arcing.

meet these requirements t is indispensable that among other thingsheresistance of the contacts be kept as low as possible, that the contact area bemaximized, that he materials are properly selected or the application at hand,that proper contact orce be applied, that the optimum number of contacts beselected, that he contact cross section andhe contact mass are properly sized,and that the minimum operating speed, both during closing and during open-ing, are sufficient to limit erosionf the contacts.

6.1.1 Contact Resistance

The resistance f a clean, ideal contact, where any influence dueo oxide filmsis neglected and wheret is assumed that a perfect pointf contact is made at aspot of radius ( r ) , s given by he following equation:

where:

R = Contact resistance

Resistivity of contact material

r =Radius of contact spot

However, in actual practice, is not the case and he real area of contactis never as simple as it has been assumed above. It should be recognized that

no matter how carefidly the contact surfaces are prepared the microscopic in-terface between two separable contacts invariably willbe a highly rough sur-face, having a physical contact area that is limited to only a few extremelysmall spots. Furthermore, whenever wo surfaces touch they will do o at twomicropoints where even underhe lightest contact pressure, dueo their smallsize of these points,will cause them to undergo a plastic deformation that cosequently changes he characteristicsof the original contact point.

It is clear then that contact force and actual contact area are two impoparameters that greatly influence he value of contact resistance. Another vari-able that also must be taken into consideration is the effects of thin films,mainly oxides, hat are deposited along the contact surfaces.

6.1.1.l ContactForce

When a force is applied a m s s the mating surfacesof a contact the small mi-croscopic points where the surfaces are actually touching are plastically de-formed and as a result of deformation additional pointsof contact are es-



tablished. The increase in the number of contact points serves to effectivelydecrease the value of the contact resistance.

The contact forceF, exerted by a pair of mating contacts be approxi-mated by the following equation l]

F = k x H x A ,

where:

H =Material hardness

A, = Contact area

k = Constant between .1 and

The constant of proportionality k is first needed to account for the surfacefinishes of the contacts and secondly because, in reality, the hardness is notconstant since there are highly localized stressest the micropoints of contact.

6.1.1.2 ContactArea

Even though it can not be determined very accurately, the knowledge of theapproximate areas of contact is essential for the proper understanding and de-sign of electrical contacts, It is found that contact resistance is a function ofthe density o f the pointsof contact as well as of the total area of true contactwithin the envelope of the two engaging full contactsurfaces. In awell-distributed area the current will diffuse toill all the available conducting zone,

but in practical contacts his area is greatly limited becauset is not possible ohave such degree f precision on the alignment, nor s itpossible to attain andmaintain such igh degree of smoothness.

In the earlier discussion, dealing with contact pressure,t was implied thatthe contact area s determined solely by the material hardness and byhe forcethat is pushing the contacts together.

Since the original simplified equation or the contact resistance was givenin terms of resistivity of the material and the radius of the contact point it is

then possible to substitute the term for the contact radius with the expressionfor the contact force, noting that:

A , =7ca

When is done and he term RP epresenting the film resistance is addedthe final expression for the total contact resistanceRT hen becomes:



140

120

C

=L 100

80

2 60

40

20

Ld

0

10 100 1,000 10,000 100,000

Load Duration in hours

Figure Resistance run-away conditionas a function of im e of carrying load cur-rents for copper contacts that ire immersed in oil a various temperatures (ref. 4

IEEE)

6.1.2 Insulating Film Coatings on ContactsA pure metal to metal contact surface can onlye achieved in a vacuum, or inan inert gas atmosphere. In air the contact surfaces oxidize and they be-come coated with a oxide film. According to Holm a layer of to

is formed on copper in a few seconds and almost instantly on alumi-

num, while t takes about wo days to form on a silverr silver plated surface.If the formed oxides are insulating, as is the case of CuO in copper con-

tacts, then due to their build-uphe contact spots will gradually reducen sizethus decreasing the contact area and increasing the contact resistance.process, as observed by Williamson and by Lemelson [4] and as shown infigure 6.1 is not very noticeable in ts early stages butn its later stages t willsuddenly get into aunaway condition.

The formationof a sulfide coat on a silver contact surfacelso produces anincrease in contact resistance. situation develop on circuit break-ers after the contacts have been subjected to arcing and when there is no



1000

100

10

Temperature Rise C)

Figure 6.2 Change in contact resistance as a function of temperature rise at the con-tacts. Silver plated contacts exposed to sulfides resulting as by-products of arcing inSF, .

scraping, or wiping motion between the contacts. However in references [5]

and [6] it is pointed out that the sulfide film on a silver surface is easily r e

moved by slight friction and that it may even become decomposed by heat.The later has been demonstrated experimentally and the results are shown infigure 6.2. It can be seen in figure that the resistance is reduced as a func-tion of the temperature rise which in was reached by passing00 amperesthrough the interrupter.

6.1.3 ContactFretting

Fretting is described as an accelerated formf oxidation that takes place a m s s

the contact surfaces and that is caused by any continuous cyclical motion ofthe contacts. Initially the junction points of the contact spots will seize andwill eventually shear, however, shear action does not increasehe contact



resistance becausehe particles are of pure metal.As the cycle repeats andhemetal fatigue progresses thenhe metal layersare softened and separate allow-ing the oxide layer o grow until contacts lost.

The increaseof contact resistance under these conditions has been obseasbeing a strong functionf current, contact force and plating material.

avoid this problem it is important then, when designing a contact interface to consider using silver plating speciallyn aluminum bars.

6.1.4 Temperature at he Point of Contact

Because of the analogy that exists between he electric and the thermal fieldsand if he assumption is made that there is no heat loss by radiationn the closeproximity of a contact the following relationship, relating the voltage dropmeasured across the contact andts temperature, canbe established.

where:

8 = Temperatureh = Thermal conductivity f contact material= Electric resistivity of contact material

V, =Voltage drop across contact

This equation however s valid onlywithin a certain limited range f tem-peratures. At higher temperatures the materials, at the contact interfaces, willbegin to soften and thus undergo a plastic deformation. At even highertemperatures the melting point of the material will be reached. In table 6.1below the softening and the melting temperatures together with their corre-sponding voltage dropre tabulated.

The significance of the above is that now we can determine for a specificmaterial the maximum currents at which either softening or melting of the

contacts would occur and consequentlyhe proper design to avoid he meltingand weldingof the contacts can e made.

The equations or the maximum oftening and melting currents are:

Softening current I , =-- nd

Melting current I,,, =P



TABLE 6.1

Softening and Melting T emperaturesfor Contact M aterials

6.1.5 Short Time Heating of C opper

The m aximum softening and m elting currents as defined by the above equa-

tions are applicable to the point of contact and are useful primarilyfor deter-mining the contact pressure needs. However when dealing with he condition

where the contacts are required to a large current for a short period oftime it is useful to define a relationship between time, current and temperaturefor different materials.

Below is given a general derivation for a general expression that beused for determining the temperature rise in a contact.

it will be that for very short times all the heat produced bythe current is stored in the contact and s therefore effectiven producing a isein temperature.

Then the heat generated by a the current (i) flowing into a contact of (R)

Ohms uring a dt interval is:

Q =R i2dt (Joules) and

The heat required to raisehe contact temperature y d(degreesC is:

Q = SVd0 (Joules)

Since it was assumed that there s no heat dissipation then t is possible to

write

R i 2 d t = S V d 0



1200

1000

G

200

0

0 2 6 8 10

TIME INTEGRALOF DENSITY

SQUARED [(A/d)2sec.]xE16

Figure 6.3 Short time heating of copper as function of the time integral f the current

squared(8 = j J z d t )0

where:

i = current in am perest = tim e in seconds

S = specific heat f material injoules per m3 pere= temperature in OC

R = contact resistance n OhmsV= contact volume nm'

It be shown that substituting the resistance f the contact with he spe-cific resistivityof the material as a function of a standard ambient temperatuwhich is assumed to be 20 "C, and by integrating the function the followingequation is obtained:



by substituting the current densityt is possible to re-write the equationas:

The graph shown in figure represents a general curve for the approximateheating of a copper contact of uniform section, with a current that startsto flow when the initial contact temperature is 20 "C. Similar curves can be

generated for other contact materials by usinghe proper constant for the newmaterial under consideration.

Since it is known the softening, or melting temperatures for a given mate-rial then it is possible to determine, fromhe graph the integral of the currentdensity and furthermore sincehe current is constant we have:

J 2 d t = J 2 t

and therefore

J 2 d tt = -

J2

6.1.6 Electromagnetic Forceson Contacts

As it has been described before, we know that theres a current constriction tthe point of contact. We also know that this constriction is responsible for the

contact resistance and consequently for the heat being generated at the con-tacts, but in addition to th is the current constriction is also the source of elec-tromagnetic forces acting uponhe contact structures. In figure 6.4 the currentpath of the current at the mating of the contact surface is shown and as it canbe seenfrom he figure,as the current constricts into a transfer point, theres acomponent of the current that flows n opposite directions and thushe net re-sult is a repelling orce trying to force the contacts.

According to Holm2] the repulsion forces given by

FR = 10-712In; NewtonB

where:

B = the contact areaa = actual areaof contact point

It is difficult however to properly useh is formula because of the difficulty in

defining the actual contact area. For practical purposes some expressions thatyield adequate results have been proposed. One such expression is given byGreenwood [7] as:



I.'

1

)-

\

I I

Figure 6.4 Current constriction athe actual point f contact.

2

FR = 0.112 (3 lb. per finger

Another practical relationship which was obtained experimentallyith ainch Cu-Cr butt type contactn vacuum is shown n figure 6.5 and is given bythe expression:

c

n!0

10,000

1,000

100

10

10 100

CURRENT in

1000

Figure 6.5 Measuredblow-out force for in. Cu-Cr butt contacts in vacuum.



1.51

FR = 0.885 ($) lb. per finger

where:

I= thepeakcur ren t inkAn = the number of contacts

The differences on the force requirements obtained depending on whichexpression is used serve to point out he uncertainty of the variables involvedand the probabilistic natureof the forces. In general the application of higherforces would yield a higher confidence level and a higher probability of withstanding the repulsion forces.

6.1.6.1Force on Butt Contacts

The repulsion forces acting on butt contacts, suchs those used n vacuum in-

terrupters, canbe calculated using any f the expressions given n the previousparagraph, considering the number of contacts n as being to Themagnitude of the repulsionforcesmust be counteractedby themechanism and therefore a proper determinationf the force magnitude is es-sential for the design and applicationf the operating mechanism.

6.1.6.2 Force on C ircular Cluster Contact

When a contact structure, such as the one shown in figure 6.6 (a), and (b) is

used, it be shown that in addition to the repelling force there is anotherforce. additional force is due to the attraction between a set of two op-posite fingers where current s flowing in the same direction on each contact(fig. 6.6 (a)). The attraction or blow-in force for a circular contact configura-tion can e calculated using the following equation:

2

FA = 0.102(n - 1)(L) Newtonn d

6.1.6.3Force on a Non-circular Non-sym metrical Cluster Contact

The following method can e used to calculate the blow-in force for each ofany of the parallel contacts n the arrangement illustrated n figure 6.6 (c). Theforces are calculated assuming that the current divides equally among eachcontact and althoughhis is not totally accuratehe results are close enoughoprovide an indication of the suitabilityof the design for an specific application.

l I 2P i - 5 = 0.102- (-) COSCC Newton

a n



Figure 6.6 Diagram forces relationships for circular and non-symmetricdcontacts.

l I 24-5 0.102- (-) cosp Newton

h n

cosy Newtonc n

F 4 - 5 = 0.1021(L)n cos8 Newton

6.1.6.4 TotalForce On Contacts

The total force acting on a contact therefore becomes:

FT=Fs+FA-FR

where:

FT = Total force per contact segment

Fs = Contact spring force



FA=Blow-in, or attraction force er contact segmentFR=Blow-out, or repulsion force per contact segment

In a properly designed contactt would be expected thatFAS''

6.1.7 Contact Erosion

Contact erosion is the unavoidable consequence of current interruption and iscaused primarilyby the vaporization of the cathode and he anode electrodes.According to W. Wilson [9] the heating leading to the vaporization at theelectrodes is the result of the accompanying voltage drops. He derived anequation for the vaporization rate in cubic centimetersf contact material lostper kA of current. In figure 6.7 the results of tests reported in reference [9]

have been reproduced. The graph presents theate of erosion in air for variouscontact materials. The data is plotted in what is called "their order of excel-lence," that is from best to worst, best being the material exhibiting the least

amount of erosion."" uI

Figure 6.7 Contact material erosion due to arcing (data from ref. 8).



................. ............... ............. ..............................

0.01

Figure6.8 Measured materialsrate of erosion due to arcing in SF,.

In figure 6.8 the rate of contact erosion in SF, based on a set of personalunpublished data is plotted. It is interesting to note that is a reasonabledegree of agreement between these values and those giveny Wilson. As cal-culated andas "vaporized loss by test."

It seems therefore thathe following equation be used to obtain reason-able estimates for the rate of erosion of the contacts.

1000(Ec +E A )

m=

where:

R = erosion rate,cc per kA - ec.Ec = cathode drop, oltsEA = anode drop, volts

J = heat equivalent, 4.18oules per calorieH = heat of vaporization, calories er gramp = density of contact material, gramser cc



6.2 Mechanical Operating Characteristics

Opening and closing velocities, as well as stroke, or travel distance are themost important operating characteristicsf a circuit breaker. They re dictatedprimarily by the requirements imposed by the contacts. Opening and closingvelocities are important to the contacts in order to avoid contact erosion aswell as contact welding. Circuit breaker stroke s primarily related to the abil-

ity of the circuit breaker o withstand the required operating dielectric stresseswhich are directly related to the total contact gap.

6.2.1 Circuit Breaker Opening Requirements

Two asic requirements or the total opening operation f a circuit breaker rethe opening speed and the total travel distance of the contacts. The openingspeed requirements are dictated by the need to assure that the parting of thecontacts is done as rapidly as possi%le for two reasons; first, o limit contact

erosion and second by the need to controlhe total fault duration whichs dic-tated by the system coordination requirements. The otal travel distance is notnecessarily the distance needed to interrupt the current but rather the gap sneeded to withstand the normal dielectric stresses and the lighting impulsewaves that may appear across the contacts of a breaker that is connected to asystem while n the open position.

The need for canying the continuous current andor withstanding a periodof arcing, makes it necessary to use two of contacts in parallel. One, the

primary contact, which s always madeof a high conductivity material such scopper and the other, he arcing contact, madeof synthetic arc resistance ma-terials like copper or silver tungstenor molybdenum which have a much lowerconductivity than those used or the primary contacts.

By having a parallel setof contacts when one of these contacts opens, dueto the differences in the resistance and the inductance of the electrical paths,there is a finite time that s required to attain total current commutation, that isfrom the primary or main contact branch to thatf the arcing contact.

The significance of the commutation time can be appreciated when oneconsiders that in the worst case commutationay not takeplace until the nextcurrent zero is reached and that during this time he arc is eroding the coppermain contacts. Since arc erosion of the contacts not only limits he life of thecontacts but it can also lead to dielectric failures by creating n ionized con-ducting path between the contacts and thus limitinghe interrupting capabilityof the circuit breaker. It is also important o realize that commutation mustbecompleted before the arcing contacts separate, otherwise, the arc is likely to

remain at the main contacts.The commutation time be calculated by solving the electrical equiva-

lent circuit or the contact arrangements given in figure 6.9.



I

time

Figure 6.9 Equivalent circuit for determining the required arc commutation time be-

tween main and arcing contacts.

It can be shown that:

I = I, + I , = I,,,sin(at ++) and

where:

I = systemcurrentI , = main contact current

R = arcing contact resistance

L = oop inductance of arcing contact

I* = contact current

Commutationis completed att = tl when I2= I

The commutation time is then obtained by solving the above equation forthe tim et,:



From the above it isobserved that o have a successful commutation ofrent it requires thatE, >IR.

While the opening operation continues and as the contact gap increases acritical contact positions reached. This new position represents the minimumcontact opening where interruption may be accomplished at the next currentzero. The remainderof the travel s needed only for dielectric and decelerationpurposes.

Under no-load conditions and when measured over the majority of thetravel distance, the opening and closing velocity of a circuit breakers is con-stant, as it is shown in 6.10 (a) and @) where actual measurements ofhe cir-

cuit breaker speeds are shown.Although M e re n t type circuit breakers have different speed and travel r

quirements, the characteristic shapeof the opening travel curve re very muchsimilar in all cases. The average speed is usually calculated by measuring theslope of the travel curve overhe region defined by the pointf initial contactpart to a point representing approximately three fourthsf the total travel dis-tance.

The opening speedor vacuum circuit breakerss generally specified n therange of 1 to 2 meters per second, for SF, circuit breakers the range is in theorder of to 6 meters per second.

. . . . . . . . . . . . ... . . . . .

CONTACTS SEPARATION . \ . . .: . . i . i .

< )

. . . . ..

(U

" - .""." ~""

~ "..

. . . . . . . . . . . . . . .. . .

-Figure 6.10 (a) Typical distancevs. time measurementofthe opening operation ofcircuitbreaker.



* ."" ". -."A": . . ". . Iv

.I".. _ . ... "- .." ". . . ..

CONTACTSSEPARATION 1. ..

.. .. .

. . . . . . . . . . .

. . CONTACT TRAVEL - . . . . . . .

. . . . . . . .

. . . . ..

. . .. CLOSINGOIL CURRENT

Figure 6.10 Typical distancevs. time measurement the closingoperation of a

circuitbreaker.

For vacuum circuit breakers theres another, often overlooked, requisiteorthe initial opening velocityof the contacts which corresponds the assumedpoint of the critical gap distance. Considering that the typical recovery volof vacuum is about 20 to 30 kV per millimeter then or those applicationsat 25or kV it is extremely important that t an assumed 2 milliseconds minimumarcing time the actual gap shoulde about millimeters which translates intovelocity of 3 meters per second and even though this is a relatively modest

velocity in comparison to other circuit breakers the fact that butt contacts areused in vacuum intermpters means thatwe need a higher initial accelerationsince there is no contact motion prior to the actual beginning of the contactseparation. In lieu of contact wipe or contact penetration, the overtravel, orcontact spring wipe, that is provided for compensation of contact erosion andto create a hammer or impact blow needed to break the contact weld can beutilized as a convenient source of kinetic energy to deliver a highate of initial

acceleration to the contacts. However becausef the mechanics at the time o

impact it is usually recommended that the mass of the mechanism movingparts be to at least twice he mass of the moving contact.



T Open Gap Voltage Withstand

Figure 6.11Prestrike during closing relationship between closing velocitynd system

voltage.

6.2.2 Closing Speed Requirements

During a closing operation and s the contacts approach each other, a points

reached where he gap equals the minimum flashover distance and thereforenelectric arc is initiated. As the distance between the contacts continues to di-

minish the arc gradually shortens until finally he contacts engage and he arcdisappears. Therefore as we have seen not only when opening but also duringclosing an arc can appear across aair of contacts.

Depending upon the voltage, the interrupting medium and the design ofeach particular circuit breaker, he contact flashover characteristics vary verywidely. However, let assum e two characteristic slopesas shown in figure6.1

(s1opes a and b ) . Now, hen these slopes are superimposed onhe plot ofthe absolute values of a sinusoidal wave that represent the system voltage, it

shows, that depending on the instantaneous relation between the contact gap



and the system voltagehe arc is initiated at the point of intersectionf the twocurves.

The elapsed time between the flashover point andhe time where the con-tacts engage represents he total arcing time whichs shownas and t corre-sponding to slopes a and b respectively. It be seen in this figure that

the arcing time decreasess the slopeof the flashover characteristics increasewhich suggests what shoulde obvious that increasinghe closing velocity de-

creases the arc duration.Increasing the closing velocity not only decreaseshe arcing time but i t

decreases the magnitude of the current at the instant of contact engagement.Assuming that the electromagnetic repulsion forces re a function of he peakcurrent squared then as show by P. Barkan [ g ] the work done by the mecha-nism against the electromagnetic repulsion is:

The dimensionless solutionf equation is given as:”- 1 sin(?)

21dm2s 20s

where:

E= work done against electromagnetic force

k 0.112 = constant of proportionalityor electromagnetic force

I , =peak currentat contact touch

S= contact travel

V = contact velocity

Q= radians for 60 Hz urrents

The benefits of a high closing are then a reductionn the mechanismenergy requirements and a reductionf the contact erosion.

6.3 Operating Mechanisms

The primary function of a circuit breaker mechanisms to provide the meansfor opening and closing he contacts. At first this seems to be a rather simpleand straightforward requirement but, when one considers hat most circuit

breakers once they have been placed in serviceill remain m the closed position for long periodsof time and yetn the few occasions when they re calledupon to open or close they must do so reliably, without any added delays, or



sluggishness; then, t is realized that the demands on the mechanismsre not assimple as first thought.

It is important then to pay special attention to such things as the type ofgrease used, the maximum stresses at the latch points and bearings, heness of the whole system and most of all to the energy output of the mecha-

Just as there are different types of circuit breakers, so are there different

types of operating mechanisms, but, what is common to all is that they storepotential energy in some elastic medium which is charged from a low powersource over a longer periodf time.

From the energy storage point f view the mechanisms thatare used in to-day’s circuit breakersall in the spring, pneumaticor hydraulic categories, andfrom the mechanical operation point of view they either of the cam or of thefour bar linkage type.

6.3.1 versus LinkageCams are generally used n conjunction with spring stored energy mechanismsand these cam-spring driven mechanisms are mostly used to operate mediumvoltage vacuum interrupters.

Cam drives are flexible, in the sense that they be tailored to provide awide variety of motions, they are small and compact. However, the cam is

subjected to very high stresses at the points of contact, and furthermore thefollower must be properly constrained, so that it faithfully follows the

cam’s contour, either by a spring which raiseshe stress level onhe cam, or bya grooved slot where the backlash may cause problems at high speeds. Forthose interested in further information forhe design of a spring operatedq-follower system Barkan’s paperlo ] is highly recommended.

To

Contacts

Figure6.12 Schematic diagram of a fourbar linkage



Linkages, in most cases, have some decided advantages over cams;or onethey are more forgiving in terms of fabrication accuracy because small varia-tions on their lengths do not significantly influence their motion. In generalthe analysis of a four bar linkage system s a relatively simpler task andt isnot that difficult to etennine the force t any point n the system.

Since the mechanism must deliver s much work as it receives, then whenfriction is neglected, the force at any point multiplied by the velocity in the

same direction of the force at that point must be equal to the force at someother point times he velocity at that same point, or in other words the forcesare inversely proportionalo their velocities.

Four bar linkages such as the one shown schematically in figure 6.12 arepractically synonymous with pneumatic and hydraulic energy storage mecha-NSmS.

6.3.2 Weld Break and Contact Bounce

Contact bounce and contact welding are two conditions which are usuallyuniquely associated with vacuum circuit breakers.

We already know that vacuum contacts weld upon closing and that there-fore it is necessary to break the weld before the contacts be opened.Breaking the weld is accomplished by an impact, hammer blow force whicis applied, preferably directly to the contacts. To provide force it is acommon practice to make use of the kinetic energy that is acquired by theopening linkagesas they travel over the compressed length ofhe wipe springs.

To assure that he proper impact force will e available, the velocity at thetime of contatt separation be used a guideline.

M , T + MX1 -k,<

dt

4 +M*

where:

M , = Contact’s mass

M, =Mechanism’s mass (moving parts)

= displacement

Contact bounce can have serious effects on the voltage transients geneduring closing and t may cause damage to equipment connected to he circuitbreaker, although protective measures be taken to reduce he magnitude of

the transients it is also advisable to design the circuit breaker so that thebounce is at least reduced,f not eliminated.



Figure 6.13 Simplified drawing of a stored energy mechanism.

Suppression of bounce requires that a close interaction be established be-tween the energy dissipation t the contact interface andhe energy storage ofthe supporting structure of the interrupter and its contacts. If means are pro-

vided for the rapid dissipation of the stored energy n the supporting structurethe effect of the bounce canbe minimized.

6.3.3 Spring Mechanisms

A simplified drawing depicting a typical spring type operating mechanism isshow in figure This type of mechanism is commonly found in somemedium voltage outdoor and in practically all of the medium voltage indoortype circuit breakers. However, it is not uncommon to find these mechanisms

also on outdoor circuit breakers p to kV and in a few cases they even havebeen used on kV rated circuit breakers.its name suggests the energy f mechanism is stored on the closing

springs. The stored energy is available for closing the circuit breaker uponcommand followinghe release of a closing latch.

The spring mechanism, in its simplest form, consists of a charging motorand a charging ratchet, a closing cam, closing springs, opening springs and atoggle linkage. The charging motor and ratchet assembly provide automatic

recharging of the closing springs immediately following the closing contactsequence.



The charged springs are held in position by the closing latch which pre-vents the closing from rotating. To release the spring energy either anelectrically operated solenoid closing coil,or a manual closing lever is oper-ated. Following the activation of the closing solenoid a secondary closinglatch is released while the primary latch rotates downward due to the forcebeing exerted by he charged closing springs and thus he rotation of the clos-ing cam which is connected to the operating rods is allowed. As the cam ro-

tates it straightens the toggle linkage (refer to figure 6.13), which in ro-tates the main operating shaft thus driving the contacts that are connected tothe shaft by means f insulating rods.

The straightening of the toggle links loads the trip latch as they go over-center. The trip latch then holds the circuit breaker in the closed position. In

addition to closing the contacts the closing springs supply enough energy tocharge the opening springs.

Opening of the contacts be initiated either electrically or manually;

however, the m anual operation is generally provided only for maintenancepurposes. When the tripping command is given the trip latch is released free-ing the trip roller carrier. Theorce produced by the over center toggle linkagrotates the trip roller carrier forward, and s the first toggle link rotates aboutits pivot it releases the supporthat provided to the second and third links. Topening springs which are connected to the main operating shaft provide thenecessary energy to openhe contacts of the circuitbreaker.

Figure 6.14 Illustration of a pneumatic mechanism.



IFigure 6.15Toggle linkagesarrangement to satisfy trip free requirements.

6.3.3 Pneumatic Mechanisms

Pneumatic mechanisms are a logical choice for air blast circuit breakers andthat is so because pressurized air is already used for insulating and interrupt-ing; however, pneumatic mechanismsre not limited o airblast breakers, theyalso have been used to operate oil andF, circuit breakers.

Those mechanisms, which are used with air blast circuit breakers, usuallyopen and close pneumatically and in some cases there is only a pneum aticrather than solid link connection betweenhe mechanism and he contacts.

Other pneumatic mechanisms, such as the one that is illustrated in figure6.14, use anairpiston to drivehe closing linkage and to charge set of open-ing springs. This mechanism, which has been used in connection with oil andSF6 circuit breakers, has a separate ir reservoir where sufficientair is storedat high pressure for t least 5 operations without needf recharging in betweenoperations.



Figure 6.16 Photograph of a hydraulic mechanism used by ABB T&D Co.

To close the circuit breaker, high pressureir is applied to the underside ofthe piston by opening a three way valve, the piston moves upwardsransmit-

ting the closing force through a toggle arrangement, figure 6.15, thats used to

provide the trip free capability, to the linkage which is connected to the con-tacts by means of an insulating push-rod. In addition to closing the contactsthe mechanism charges a set of opening springs and once the contacts areclosed a trip latchs engaged to hold the breakern the closed position.

Opening is achieved by energizing a trip solenoid whichn releases the

trip latch thus allowing the dischargef the opening springs which forces thecontacts to the open position..

Another variation of a pneumatic m echanism is one where the pneumaticforce is used to do both, the closing and the opening operation. The directionbeing controlled by the activation of either of the independent opening orclosing three way valves.

6.3.4 Hydraulic Mechanisms

Hydraulic mechanisms are in reality only a variation of the pneum atictor, the energy, in most cases, is stored in a nitrogen gas accumulator and theincompressible hydraulic fluid becomes a fluid operating link thats interposed



18 J

OPEN position

Figure 6.17 Functional operating diagram of themechanism shownin figure 6.16.



between the accumulator and a linkage system that is no different than thatused in conjunction with pneumatic mechanisms.

In a variation of the energy’s storage method, he nitrogen accumulator isreplaced by a disk spring assembly whichcts as a mechanical accumulator.Amechanism of type is shown in figure 6.16. It offers significant advan-tages; it is smaller, there is no chance for gas leaks from he accumulator, andthe effects of the ambient temperature upon the stored energyre eliminated.

The operationof this mechanism can e described as follows; (note hat thenumerical references to components correspond to those shownn figure 6.17).

A supply of hydraulic oil is filtered and stored n a low pressure reservoir(12), from where it s compressed by the oil pump (1 ). The high pressureoilis then stored in reservoir ( 5 ) . The piston (3) which is located inside of thehigh pressure storage is connected the spring column (1). The springs aresupported by the tie bolts (2). A control link (15) checks on the charge of hespring column and activates he auxiliary switch contacts (16) that control he

pump’s motor(10) as required to maintain the appropriate pressure.With the circuit breaker in the closed position the operating piston (7),

which is connected to the conventional circuit breaker linkages (S), has highpressure applied to othof its faces. To open the breaker the opening solenoid(17a) is energized causing the changeover valve to switch positions and con-nect the underside of he operating piston o low pressure (6), thus causinghepiston to move to the open position. Closing as the reverse of the opening andis initiated by energizing the closing solenoid (1%) and by admitting high

pressure to the underside of the operating piston. Item (4) is an storage cylin-der, (9) is a mechanical interlock (13) s an oil drain valve and (14)s a pres-sure release valve.

REFERENCES

1. CIGRE SC13, High Voltage Circuit Breaker Reliability Data For Use InSystem Reliability Studies, CIGRE Publication, Paris France 1991.

2. R. HolmElectricalContacts,Almquist&

WiksellsAkademiskaHand-bkker, Stockholm, Sweden, 1946.J. B. P. Williamson, Deterioration process n electrical connectors, Proc.4th. Int. Electr. Contact Phenomena, Swansea, Wales, 1968.

4. K. Lemelson, About the Failure of Closed Heavy Current Contact Piecesin Insulating Oil at High Temperatures,IEEETrans. PHP, Vol. 9, March1973.

5. G. Windred, Electrical Contacts, McMillan andCo. Ltd., London, 1940.

6.S.

C. Killian, A New OutdoorAir

Switch And A New ConceptOf

Con-tact Performance,AIEETrans. Vol. 67: 1382-1389,1948.



7. A. Greenwood,VacuumSwitchgear,IEEPowerSeries18,London, UK,

8. W. R. Wilson, High-Current Arc Erosion of Electrical Contact Materials,

9. T. H. Lee, Physics and Engineering of High Power Devices, The Massa-chusetts InstituteofTechnology: 48 5489 , 1975.

10. P. B a r b , R. V.McGanity,ASpring-Actuated Follower System,

Design TheoryandExperimentalResults, of ASME, Journal ofEngineering for Industry, Vol. 87, Series B, No. 279-286, 1965.

1994.

AIEE Trans. Part PAS Vol. 74: 657-664,August 1955.



7

A COMPARISON OF HIGH VOLTAGE CIRCUITBREAKERSTANDARDS

7.0 Introduction

The p r e f d atings, the performance parameters and the testing requirements ofcircuit breakersrewell established subjects thatre covered bya number of appli-cable These standards,both nationally and internationally, have evolvedover a long period of time and volution process undoubtedly continue b ecause of he necessity to reflect the changing needs ofhe industry.

Futurechanges in the standardscan be expected,notonlybecause ofchanges in the circuit breaker technology but, due to better, faster, more ad-vanced and more accurate relaying and instrumentation packages and also because of the much broader use of digital equipment that is now replacing oldelectromagnetic-analog devices.

In chapter we will review the most influential current internationalstandards, their historical background, their organization and procedures, aswell as their specific requirements andhe reasons for these choices. W e will

discuss these current standards covering ratings and performance parameters;comparing their differences and providing clarifications to the intent and theorigin of the ratings.

The section dealing with power testingf circuit breakers deserves specialattention and therefore the next chapter will be dedicated completely to this

subject.

7.1 Recognized Standards Organizations

The two most recognized and influential circuit breaker standardsn the worldare supported by the American National Standards Institute (ANSI) and theInternational Electrotechnical Commission (IEC).

ANSI is the choice document in the US, and in places around the worldwhere the US has had a strong influence in the development of their electricindustry.IECstandards are invokedby all othercountriesoutside of thesphere ofUS influence, and today he majority of circuit breakers thatare be-

ing sold worldwide re being built o meet the IEC standardsThe format of the two standards is significantly different, and o some ex-tent there are noticeable differences in the technical requirements, but thesedifferences are not so radical to the point of being non-reconcilable. Thesedif-



ferences have moreo do with ocal established practices than with fundamen-tal theory and f there is something tobe said about experience, then bothdards have demonstrated o be more than adequate for covering the needs ofthe industry. The most significant differences will be discussed in the appro-priate sectionsaswe review the generic requirements.

A current issue that commands areat deal of attention is the issue ofhar-monization. This has become increasingly more important not only because oagreements reachedby the World Trade Organization andhe globalization oftrade, but, also because todayll of the basic development of high voltageF,circuit breakers s being done overseas andll of the US suppliers of h is typeof equipment are owned by Europeanr Japanese multinational corporations.

7.1.1 ANSI/IEEE/NEMA

ANSI is a mem ber’s federation hat acts as a coordinating body or a l l volun-teer standards writing organizations in the US with the aim of developing

global that reflect US interests. provides the criteria and theprocess for approving a consensus.

High voltage circuit breakers and switchgear standards are, for the mostpart, developed throughhe combined and separate effort of E E and NEMkBoth of these groupsare co-secretariats of the Accredited Standards Commit-tee C37 (ASC C37) which serves as the administrator, or the clearing housethrough which he proposed standards re submitted toANSI or publication as

an AmericanNationalStandard.It is theAccreditedStandardsCommittee

whoassigns he ndustry’swell known and readily identifable series C37numbers to the ANSI/ IEEE/NEM A circuit breakers

The membership of the technical committees, subcommittees and workingof the IEEE consists of voluntary individuals that represent users,

manufacturers and interested groups,n a reasonably well balanced proportion.The membership ofNEMA consists solely of equipment manufacturers.

To help visualize he inter-relationships that exists between theserganiza-tions and the flowf the standard documents, a basic organizational diagrams

included in figure 7.1.The first report on standardization rules n the electrical field be traced

to 1899 when, it was prepared by the American Institute of Electrical Engi-neers, (AIEE), ow IEEE. In the years that followed, other organizations, suchas the Electric Power Club, which n 1926 merged with he Associated Manu-

facturers of Electrical Supplies to become whats now NEMA, became inter-ested in the process of standardization. In today’s scheme, NEMA is respon-sible for the development of those standards that are related to Ratings, Con-

struction and Conformance Testing. IEEE responsibilitys to develop all othertechnical standards related o Requirements, Performance, and Testing includ-ing design, production, and field testing.



Board of Standards

Review................................... .....................................

IASCc 3 7

IIEEE

r

...........Standards Board ..............EMA

'PESSwgr.Std.

T&D

Swgr.SGS

HVCB1 Subcomm. I 1 '";" IWorking

Groups

Working

Groups

Figure 7.1 Organizational Chart of ANSI/IEEE and NEMA Groups Dealing with

High VoltageCircuit Breaker Standards.



The mainANSI high voltage circuit breaker standardsre (NEMA) C37.06AC High-Voltage Circuit Breakers Ratedn a Symmetrical Current Basis- Prferred Ratings and Related Required Capabilities, ANSVIEEE C37.04 IEEEStandard Rating Structure for AC High-Voltage Circuit Breakers Rated on aSymmetrical Current Basis, and ANSVIEEE C37.09 IEEE Standard Test Pro-cedure for AC High-Voltage Circuit Breakers Rated on a Symmetrical CurreBasis.

7.1.2 IEC

IEC, as an organization,dates back to the early 1900’s.tsmembership ofnational delegates from member which in the form f na-tional committees. Presently there re forty-nine (49) national committees and on(1) associate membercountry. Each national committee, or member nation havesingle voteThis is one important point to emphasiie, becauset shows IECasbeinga truly in tm tiona l organization where the decisions approved at the country

levels and each country member has only one vote. This represents the majorphilosophical difference between IEC and ANSI, since ANSIs only a ~ t i dr-ganization where each individualas one vote.

The United States National Committee (USNC) to the IEC serves as the

sponsored US delegation. Within this delegation there are a number oftechnical advisory (TAGS) that are called upon to support the indi-viduals who act in the role of technical adviser (TA). It shouldbe noted thatbothIEEEand NEMA areheavily nvolvedandhaveastrong nfluence

the USNC delegations because, most of its members are also mem bersof either or both E E E and N E M A .

IEC high voltage circuit breaker standards are prepared by the TechnicalCommittee 17 (TC 17). More specifically, the scope of TC 17 is to prepareinternational standards regarding specifications or circuit breakers, switches,contactors, disconnectors,busbarandany witchgearassemblies.Subcommittee 17 A (SC 17A) is responsible for the current applicable

dards lEC 56 (1987) and IEC 694 (1980) Com mon clauses for high-voltageswitchgear and controlgeartandards.The organizational relationships IEC are illustrated in the accompa-

nying figure 7.2.

7.2 Circuit BreakerRatings

The ratingso f a circuit breaker, given byhe applicable standards, are consid-ered to be the minimum designated limits of performancehat a re expected to

be met by the device. These limits are applicable within specified operatingconditions. In addition of including the fundamental voltage and current pa-rameters, they list other additional requirements which, re derived from the



IEC CouncilPresidents of 49

National Committees

Committee of ActionReps from 12NC’s

Group B Group Croup A

Tech. Corn.TC 17

Sub-Comm.SC17A

ITech. W.

Figure 7.2 OrganizationalChart of IEC Groups Dealing with High Voltage Circuit

BreakerStandards.

above listed basic parameters, and whichn the ANSI documents are idenW1edas related required capabilities.

C37.06 contains a number of tables where a listf Preferred Ratingsis included. These ratings are just that, “preferred”, because they re the onesmore commonly specified by users and are simply those which havebeen se-



lected byNEMA strictly for convenience, and n order to have what t could besaid to be an off he shelf product. The fact that there is a listing of preferredratings does not excludehe possibility of offering other specific ratingss re -quired, provided that all the technical performance requirements and the re-lated capabilities specifiedn the appropriate standard(C37.04) re met.

7.2.1 Normal Operating Conditions

ANSI considers as normal or usual operating condition ambient temperatureswhich do not exceed plus0 “C nd which are not below minus0°C, nd alti-tudes which do not exceed 3300 ft or m. ANSI does not differentiateservice conditions between indoorr outdoor applications.

IEC does differentiate for indoor, or outdoor applications. It specifies thealtitude limit at 1000 m and he maximum ambient temperature as plus 40 “Cfor both applications; however t additionally specifies that the average f themaximum temperature over a4 hour period does not exceed5 “C.

For the lower temperature limits there are two options givenor each appli-cation. For indoor, there s a minus5 “C imit for a class “minus5 indoor” anda minus 25 “C or a class “minus25 indoor.” For outdoor applications theresa class “minus25” and a class “minus 0.” Limits for icing and wind velocityare also recognizedy IECbut are not byANSI.

7.2.2 Rated Power Frequency

This seemingly simple ratinghat relates only to the frequencyf the ac power

system has a significant influence when relatedo other circuitbreaker ratings.The power frequency rating s a significant factor during current interruption,because for many typesof circuit breakers, the rate of change of current at thezero crossing s a more meaningfbl parameter than the given rms.r peakrent values. In all cases, when evaluating the interrupting performance of acircuit breaker, it should be remembered that, a 60 Hz current is generallymore difficult o interrupt a 50 Hz current of he same magnitude

7.2 Voltage Related Ratings

7.2.1 Maximum Operating Voltage

Whether called rated maximum operating voltage (ANSI) or rated voltagerating sets the upper limit of the system voltage for which the

breaker application s intended.The selection of the maximum operating voltage magnitudes s based pri-

marily on current local practices. Originally, ANSI arrived at the maximumoperating voltage rating by increasing the normal operating voltage by ap-proximately 5 % for all breakers below 362kV and by 10% for all breakers



Table 7.1ANSI and IEC Rated Operating Voltages

(Voltages shown in kV)

For voltages 72.5 kV and below

IEC 72.52 36 247.52.2.6

ANSI 72.58.385.85.55 8.25 4.76

For voltages above 72.5 kV

IEc 765 52620045704 12 100 3 5

1 5ANSI 8005062 242 1694 12

above 362 kV. However, the practice of using the nominal operating voltagehas been discontinued, primarily, becausehe use of only the maximum ratedvoltage has become common practice by other related standards, [l], [2], forapparatus that are used in conjunction with circuit breakers..

As it can be seen below, n Table 7.1, the preferred values that re specifiedin each of the two standards are relatively close to each other. The voltageratings of 100, 300 and 420 kV are offered in the IEC due to the practices inEurope, while these ratingsre not common n the US and therefore are not of-fered by ANSI. Furthermore, t is expected that n the next revision of he pre-ferred ratings for outdoor oilless breakers, which are listed in ANSI C37.06,the values for 121,169 and 242 kVwill be changed to harmonize with he IEC

values.

7.2.2 Rated Voltage Range Factor

The rated voltage range factor, as defined by ANSI, is the ratio of the ratedmaximum voltage, to the lower limit of the range of operating voltage, inwhich the requiredsymmetricalandasymm etrical nterruptingcapabilitiesvary in inverse proportion to the operating voltage. This is a rating that isunique to ANSI, and it is a over from earlier standardshat were basedonolder technologiessuch as oil, andair magnetic circuit breakers; where, s weknow, a reduced voltage results n and increase in he a m e n t interrupting ca-pability. With modem echnologies, namely vacuum and F, this is no longer



applicable. This fact was recognized for outdoor oils circuit breakers, andherated range factor for type of circuit breakers was eliminated moretwenty years ago.

Another reason thatas been given or the adoptionof the K factor ratingsthe convenience of grouping a certain range of voltages under a common de-nominator, which in case was chosen to be a constant MVA. Where,

MVA is to: rated maximum voltage rated symmetrical current.Currently there is a motion that is working its way through the standards

committees seeking to delete requirement as it applies to indoor circuitbreakers.

7.2.3 Rated Dielectric Strength

The minimum rated dielectric strength capabilityf a circuit breaker is speci-fied, by the standards, n the form of a series of tests that mustbe performed,and which simulate power frequency overvoltages, surge voltages caused by

lighting, and overvoltages resulting from switching operations.

7.2.3.1Low requency dielectric

The withstand capability is one of the earliest established rating parametersfa circuit breaker.AIEE as early as 1919, specified a one minute,0Hz., dry ests and selected a valueof 2.25 times rated voltage plus2000 voltsas the basis of rating. In addition to the dry test, a 10 second wet test,which still is required, was included. The voltage magnitude chosen for

test was2 times rated voltage plus 1000 volts.In the IEC standards, the low frequency withstand voltages, for circuit

breakers rated72.5 kV and below, match the one minutery test values speci-fied by ANSI; However, at voltages above 72.5 kV he IEC requirements aresignificantly lower than the ANSI ones. For circuit breakers rated42 kV andbelow IEC does not have a0 second wet test requirements ANSI does.

Since, in general any power frequency. overvoltages that may occursn anelectric systemare much lower than the low frequency,60 Hz)withstand val-ues that are required by the standards, we only conclude that these highermargins where adopted n lieu of switching surge tests which, t the time werenot specified and consequently, were not performed.

7.2.3.2 Lighting impuIse withstand

These requirements are imposed in recognition to the fact that overvoltagesproduced in an electric system by lighting strokes are one of the primarycauses for system outages and for dielectric failures of the equipment. Themagnitude and the waveformf the voltage surge, t some point on a line, dpends on the insulation levelf the line and on the distance between the poof origin of the stroke and the point on theine which is under consideration.



This suggests that it is not only difficult to establish a defmite upper limitorthese overvoltages, but thatt would be impractical to expect that high voltageequipment, including circuit breakers, should e designed so that they are ca-pable of withstanding the upper limits f the overvoltages produced by lightingstrokes. Therefore, and in contrast to the 60 Hz equirements; where the mar-

are very conservative and are well above the normal frequency overvolt-ages that may be expected, the specified impulse levels are lower the

levels that be expected in the electrical system n the of a lightingstroke, whether it is a direct stroke to the station, o r the most likely event,which is a stroke tohe transmission line that is feeding the station.

The objective of specifying an impulse withstand level, even though is

lower that what be seen by the system, is to define the upper capabilitylimit for a circuit breaker and to definehe level of voltage coordination thatmust be provided.

The Basic Impulse Level, (BIL), that is specified, in reality only reflects

the insulation coordination practices usedn the design of electric systems, andwhich are influenced primarily, by he insulation limits and he protection re-quirements of power transformers and other apparatus in the system. Eco-nomic considerations also play an important role in the selection because, inmost cases the circuit breaker must rely onlyn the protection that is offeredby surge protection thats located at a remote location fromhe circuit breakerand close to the transformers and because as is common practice surge ar-resters at the line terminals of the circuit breaker are generally omitted.n ref-

erence C. Wagner et. al., have desm%ed a study they conducted to de-termine the insulation level tobe recommended for a 500 kV circuit breaker.The study, as reported, was done by equatinghe savings obtained by increas-ing the insulation levels, against the additional cost of installing additionalsurge arresters, f the insulation levels were toe lowered. The findings, in the

particular case that was studied, suggested that a 1300V level was needed forthe 500 kV transformers in the installation, however, a level of 1550 kV waschosen as the most economic solution for the associated circuit breakers and

disconnectswitches.Otherstudieshaveshown hat in somecases, for asimilar installation, an kV BIL wouldbe needed. The findingsof the twoevaluations then, further suggested, thatwo values may be required, but hav-ing two designs to meet the different levels s not the optimum solution frommanufacturing point of view and consequently only the higher value wasadopted as the standard rated value.

specifies only oneBIL value for each circuit breaker rating withheexception of breakers rated and kV where two B E values are given.

The lower value is intended for applications on a grounded wye distributionsystem equipped with surge arresters. IEC, in contrast, specifies two BIL rat-



ings for all voltage classes, except or 52 72.5 kV where only onevalue is

given, and for 245 kV where three values are specified.The comparative values for circuit breakers rated 72.5 kV and above are

given in Table 7.2. From t h i s table it be seen that up to the 169/170 kVrating the ANSI values are directly comparable to the higher value given foreach voltage class by IEC, at 245 kV, for all practical purposes, the 900 V

Table 7.2BIL Comparison

ANSI and IEC

ANSI

2 psec.

72.5.8~~ ~ ~~~~~

121 71050.55

145 838 650.5

169 968 750.45

242 1160 900 3.7

362 I 3.58 I 1300 I 1680

1 I3 1 1800 I 2320

2050640

3 psec.ChoppedWave

402

632

748

862

1040

~ 1500

~ 2070

IE CI RatedBIL Voltage

325

245 1050

650

17050

550

145 650

450

12350

72.5

95050 I1175

1050

362



required by ANSI should be adequate for the two lower values ofIEC. As isseen at 362 and 550kV he ANSI values are significantly higher than he IECvalues.

The preceding comparison shows the variability of the requirements andreaffirms the fact that the values, as chosen, are generally adequate wheneverproper coordination proceduresare followed a fact that is corroborated by thereliable operating history ofhe equipment. It can also be observed thatup to

169 kV atings the per unit ratio between the impulse voltage and the maxi-mum voltage of the breaker is basically a constant of approximately 4.5 p.u.but as the rated voltageof the breaker increases theBlL level is decreased. C.Wagner [4] attributes decrease to the grounding practices, since at thesevoltage levels all systems are effectively grounded, and also to the types ofsurge arresters used.

7.2.3.3Chopped wav e withstand

This dielectric requirement s specified onlyin and it has been a part ofthese standards since 1960. This requirement was added in recognition of thefact, that the voltage at the terminals of the surge arresterhas a characteristi-cally flat top appearance, but t some distance fromhe arrester, the voltage issomewhat higher. This characteristic had already been taken into account bytransformer standards where a p e c chopped wave requirement was speci-fied.

An additional reason or establishing the chopped wave requirement waso

eliminate, primarily for economic reasons, the need for surge arresters at theline side of the breaker and thus, to allowhe use of rod gaps and o rely on heusual arresters which are used at the transformer terminals. The 3 p e c ratingis given as 1.15 times the correspondingBIL. This value happens to be thesame as that of the transformers and it assumes that the separation distancebetween the and the circuitbreaker terminalss similar as that betweenthe transformer and the arrester. The 2 p e c peak wave is given as 1.29 timesthe correspondingBIL of the breaker. The higher voltage is intended to ac-

count for the additional separation from he breaker terminals o the arrestersin comparison to he transformer arrester combination.

Basic Lightning ImpulseTests. The tests are made underdry conditions usingboth, a positive and negative impulse wave. The lighting impulse waveis defined as a 1.2 x 50 microseconds wave. The waveform and the points

used for defining the wave [5], are shown in figure 7.3 (a). The 1.2 p e c valuerepresents the front time of the wave and s defined as 1.67 times the time in-terval tf that encompasses the 30 and 90 % points of the voltage magnitude,

when these two pointsre joined by a straight line. The0 p e c represents thetail of the wave and s defined as the point where he voltage has declined to



V

V

:: :

Figure 7.3 Standard WaveForm for Chopped Wave Tests.



half its value. The time is measured from a point= 0 which is defined by theintercept of the straight line betweenhe 30 and90%values and the horizontal

that represents time.In figure 7.3 (b) a chopped wave s illustrated, the front of the wave s de-

fined in the same manner as above but, the time showns &, which representsthe chopping time, is defined s the time from the waverigin to the pointofthe chopping initiation.

Until recently,ANSI has been specifying a3 3 test method, which meantthat the impulse waves applied three consecutive times to each test point,f aflashover occurred during the three initial tests, then three additional tests hato be performed where no flashovers were allowed.

The test methodhas now been changed and the new requirement defines a3 x 9 test matrix. With th is procedure, a group of three testsare performed, ifone flashover occurs then, a second set f nine tests, where no flashoversareallowed, must be performed. The reason for the change in the test procedure

was based on the desire to increase the confidence level on the withstand ca-

pability of the design.IEC specifies a 15 method which requires that a group of iifteen con-

secutive testsbe made and where only two flashovers, m s s self restoring in-sulation, are allowed during the series. However,EC has given the option oremploying the 3 9 method, which be accepted subjected to the agree-ment between he user and the manufacturer.

Conjidence Level Comparison. To evaluate which test yieldeda higherdence level, the approacho the statistical comparison ofhe methods that wasused is given below.

The Binomial Distribution is used to evaluate the required confidence lev-els, th is distribution considers a number of independent experiments, n, eachwith a given probability f success,p , and probabilityof failure, (1 -p). I f x is

the number of successes and (n - is the number of failures, then the prob-ability of a random variable,X, eing greater or equal to somes:

and the confidence level s given by

c =l -PrFor the2 x 15 method the confidence levels calculated as follows:

n = l5 tests2 z 0 = 13 successes minimum



p = 90 %probability of success on any test

h 2 X ( 1 3 ) = C5 15!

(o.gyC(1- o.9)’5-xx=13 (15-

PrT X(13) = 0.816 and

C = 1- 13)= 1-0.816 = 0.184

This ndicates that theres an 18.4% confidence level thathe breaker has a 90%probability of withstanding the lighting impulse.

x. = 3 successesminimum


h=(3)= (3--4!x!1 (O.gY(1- 0.9)?”’ = 0.729

and C = 1-Prrx (3) = 1- 0.729 = 0.271

So there is a 27.1% confidence level thathebreaker has a 90%probability ofwithstanding the lighting impulse.

Next, the confidence levels calculated if one failure occursn the first testsand none n the remaining 9

n = 12 tests

2 x. = 11 successesminimum


h 2 X ( 1 1 ) = c2 12!

(0.9)x(1- 0.9)”-‘ = 0.658(12

and C = 1-h= (3) = 1- 0.658 = 0.34.2

In th is case there is a 34.2 % confidence level that the breaker has a 90 %

probability of withstanding the lighting impulse.

7.2.3.4IECBias

In recognition of he fact that a lightingtroke may reach the circuit breakertany time and of the effects producedy the impulse wave uponhe power fi-e-

wave, IEC requires hat for all circuit breakers rated 300V and above



an impulse bias test e performed. This ests is made by applying a powerie-quency test voltage to ground which when measuredn relation to the peak fthe impulse wave is no less than times the rated voltage of the circuitbreaker. No equivalent requirementhas been established by

7.2.3.5 Switching Impuke W~hs t a nd

This requirement is applicable to circuit breakers rated, byNSI at kV, or

above and by IEC at kV and above. The reason why these requirementsspecified only at these values s because, at the lower voltage ratings, he

peak valueof the specified power fiequency withstand voltage exceeds the.0

per unit voltage surge which is the value that has been selected as themum uncontrolled switching surge that maye encountered on a system.

At the voltage levels, where the switching impulse voltage is specified,there are two levels listed and with the exception of the across the terminalsrating for a kV circuit breaker,all the other values re lower than the3.0

p.u. value mentioned earlier. The lower voltages have been justified by argu-ing that, when the circuit breaker s closed, the breaker s inherently protectedby the source side arresters which re always used. However, with the breakerin the open position, there are line side arresters, the breaker would eunprotected and the specified levels would be inadequate, but nevertheless, itis also recognized that acircuit breakers at these voltage levels use someform of surge control. This practice is reflected in the factors, that are listedon ANSI C37.06, or circuit breakers specifically designed to control line

closing switching surge maximum voltages,

7.2.4 Rated Transient Recovery Voltage

Recalling what was said earlier, in chapter the transient recovery voltagesthat be encountered in a system can be rather complex and difficult toculate without the aid of adigital computer. Furthermore, we learned that theTRV s strongly dependent upon the type of fault being interrupted, the con-figuration of the system and the characteristics of its components, i.e.

formers, reactors,capacitors, cables, etc.We hadalso made a distinction or terminal faults, short ine faults and for

the initialTRV ondition. The standards recognize all of these situations andconsequently have specific requirements for each one of these conditions. Asstated earlier, for standardization and testing purposes, the prospective,or in-herent TRV could be simplified so it could be described in terms of simplerwaveform envelopes. It is important to emphasize that what follows always rfers to the inherent that means the recovery voltage produced by the

system alone discounting any modificationsor any other distortions that maybe produced by the interaction of the circuit breaker.



7.2.4.1 Terminal aurts

Based on the above, ANSI had adopted two basic waveforms to simulate themost likely envelopes ofhe TRV for a terminal fault condition under whatbe considered as generic conditions.

For breakers rated 2.5 kV and below the waveforms mathematically de-fined as:

&v = 1- cosine

where:

= - andT2

T2= Specified rated timeo voltage peak

For breakers rated121 kV and above he wave form is approximately defined

as the envelope of the combined Exponential-Cosinefunctions, and the expo-nential portion 6] s given by he following equation

where:

Rated Max. VoltageL =

%ITQ I

and whereEl,G,R , I , Tl, and T2are the values givenin ANSI C37.06.



250

TIME psec

Figure 7.4 Comparisonof ANSI (1- cos) andIEC b o Parameter TRV for 72.5 kVRated CircuitBreaker.

IEC defines TRV envelopes, one thats applicable to circuit break-ers rated 72.5 kV and below and whichs defined by what is known as the TwoParameter Method, and the second whichs applicable to circuit breakers ratedabove 72.5 kV and which is known as theFourParameterMethod. In figures7.4 and 7.5 the TRV envelopess defined by and IEC are compared, andthe curves shown correspond to a 72.5 kV and 145 kV rated circuit breakersrespectively.

The method employed byEC assumes that he TRV wave may e definedby means of an envelope whichs made of three line segments, however whenthe TRV approaches the l-Cosine, or the damped oscillation shape the enve-lope resolves itself intowo segments.

The following procedure for drawing the aforementioned segments as de-scribed in [7] be used.: Thefirst line segment s drawn from the origin andin a position that is tangential to the TRV without crossing th is curve at anypoint. The second segment is a horizontal line which is tangent the highestpoint on the TRVave. The third segments a line that oins the previouswo



300

250

200

150

100

50

0

0 10000 300 400

TIME psec

Figure 7.5 TRV Comparison o f and IEC Four parameter envelopes

for a 145 kV Circuit breaker.

segments, and which is tangent to the TRV wave without crossingwave at any point. However, when the pointf contact of the irst line and the

highest peak are comparatively close to each other, the third line segment isomitted and thewo parameter representation obtained.

When all three segments re the four parameters defined are:

U, = Intersection point of first and line segments corresponds to

t, = time to reachu1 n microseconds.= Intersection point of second and thirdine segments corresponds to

tz= time to reach in microseconds.

the first reference voltage n kilovolts.

the second reference voltageTRV peak) in kilovolts.

The two parameter line is defined by:



U, = corresponds to reference voltage"RV peak) in kilovolts.t3= time to reach inmicroseconds.

Not only the adopted waveforms different but, there a number of addi-tional dif€erences between the transient recovery voltages that specified byANSI and EC; or example, the "RV characteristics d e s m i d by inde-pendent of the interrupting current ratings of the breaker, while the ANSI char

istics vary in relation to the interrupting rating. ANSI specifies different rates ofchange, (dv/dt), for the TRV. These rates also dependent on the interruptingcurrent ratings but,EC establishes onlya constant equivalent rate thats applicableto ratings. Finally, ANSI that be ungrounded, whichsa conservative approach because it gives a 1.5 factor that is as the multiplierfor the first phase to clear the fault. E C , in the other hand, a distinctionbetween grounded and ungrounded applications and accountsor these differencesby the usage of the multiplier 1.3, or 1.5 respectively for thefirstphase to clear.

However, even in thosecases

where the same 1.5 multiplier is used by bothdards, the con-esponding ANSI values higher because, based on the recommen-dations made by the AEIC study committee [S ] the amplitude factors used. are

to 1.54 for circuit breakers rated 72.5 kV and below and 1.43 for circuitbreakers rated above 72.5 V. In contrast, the amplitude factor used byEC is 1.4for all voltage ratjngs.

There is a relatively signifcant time differences observed between theand the time to crest requirements or circuit breakers rated below

72.5 kV. This difference be explained by considering he European prac-tice where most of the circuits at these voltages are fed y cables, and conse-quently, due to the inherent cable's capacitance, the natural frequency of theresponse is lowered.

The influence of capacitance, n the source side of the breaker, produces aslower rate of rise of the "RV and therefore, what amounts to a delay time bfore the rise in the recovery voltage is observed. ANSI does not specify a de-lay for the conditions of a l- cosine envelope because in case the initial

rate is equal to zero, E C owever, due to the use of the two parameter enve-lope, does specify a time delay which ranges; from a maximumof 16 micro-seconds, down o 8 microseconds. For circuit breakers rated above 121V thedelay time specifled by is a constant with a value of 2 microseconds,ANSI specifies different delay times for different voltage ratings and thesetimes range from a lowof 2.9 microseconds at 121 kV to a maximum of 7.9microseconds for kV circuit breakers.

In both standards, consideration is given to the condition where currents

significantly lower thanhe rated fault currents, and usuallyn the approximaterange of 10 to 60 YO. re interrupted. This condition is assvumed to occur whenthe fault occurs in the secondary side of the transformer, in such cases the



Table 7.3Breakers Rated72.5 kV and below

Multiplying Factors or Reduced Fault Currents

ANSI IEC

FaultEz

Current %U,

Current %

Fault

60

1.070.40 1.130

1.070.67 1.07

10 1.070.40 1.17

Table 7.4

Breakers Rated 121 V and aboveMultiplying Factors or Reduced Fault Currents

ANSIFault

Current

60

30

10 1.17.20

60 I1.07

10.09

t3 I

1oo I121 kV145kV

to245 kV362 kV

550 kV

800 kV

-currents are reduced due to the higher reactance of the transformer. Thesehigher reactance in change the natural frequencies of the inherent "RV.

To account for these changes and IEC have selected three levels of re-duced currents where the specifiedeak voltage is higher and theime to reachthe peak is reduced. The factors by which, the voltages are increased and the

times are reduced, are tabulated in Tables 7.3 and 7.4. The net result is thatthe peak "RV values given by are higher than the ones required byEC

but, the time required to reach theeak is always higher according toEC.



Table 7.5

Comparison of SLF Parameters

ANSI

All = 450ll = 450

IEC

Surge Impedance

Z Ohms

Amplitude Factord

All = 1.6 All = 1.6

Time delay

(seconds (245 = 0.5

(170 kV = 0.2

7.2.4.2 ShortLineFaults

An important difFerence between and IEC for short line faults is that E Crequires capability only on circuit breakers rated 52 kV and above, andwhich are designed or direct connection o overhead l i e s . requires the

same capability for all outdoor circuit breakers. Other differences found be-tween the two standards re compared in Table 7.5.

7.3 Current Related Ratings

7.3.1 Rated Continuous Current

The continuous current ratings that which sethe limits for the circuit breaker

temperature rise. These imits are chosen so that a emperature run awaycondition, as it was described in the previous chapter, s avoided when sinto consideration the type of material used in the contacts, or conductingjoints, and secondly, that the temperature of the conductingparts, which arein contact with insulating materials, do not exceedhe softening temperature ofsuch material. The temperature limits are given in terms of both, the totaltemperature and the temperature rise overhe maximum allowable rated ambi-ent operating temperature. The temperature rise values given to simplify he

testing of the circuit breaker because as long as the ambient temperature isbetween the range of 10 to0 no correction factors need toe applied.



The preferred continuous ratings specifed by are, 600, 1200, 1600,2000, or 2000 Amperes. The corresponding IEC ratings are based on the R10series of preferred numbers and they are; 630, 800, 1250,600,2000, 3150, or4000 Amperes.

The choices of the numerical values for the continuous currents, made byeach of the two standards, are not that different, buthe real significance of theratings is the associated maximum allowable temperature limits that have

established. These temperature limits, as they are currently specified, by eachof the standards, are shown in Tables 7.6 and 7.7.

The maximum temperature limitsor contacts and or conducting oints arechosen based on the knowledge of the relationships that exist or a given ma-terial between the changen resistance due to the formationf oxide films, andthe prevailing temperaturet the point of contact. For insulating materials hetemperature limits are well known and are directly related to the mechanicalcharacteristics of the material. These imits follow the guidelines that are

given by other standards, such s AS", tha t have classified the materials inreadily identifiable

additional requirement givens the maximum allowable temperature fcircuit breaker parts thatay be handled byan operator, these parts re limitedto amaximum total tempemture of50 "C and for those points that can e ac-cessible to personnel the limit is 70 "C. But, in any case external surfacesarelimited to 100 "C.

7.3.1.1B efm re d Number SeriesPreferred numbers are series of numbers that are selected for standardizationpurposes in preference of any other numbers. Their use leads to simplifiedpractices and lead to reduced numberf variations.

The preferred numbers are independent of any measurement system andtherefore they are dimensionless. The numbers are rounded values of the fol-lowing five geometriceries of numbers:ONflo, loNao, and lomo,where N is an integer n the series0, 1,2 , etc. The designations used or the

five series are R5, R10, R20, R40 and R80 respectably, where R stands forRenard of Charles Renard, the originator of series, and the number indi-cates the root of ten on whichhe series is based.

The R10 series s frequently used to establish current ratings.This particu-lar series gives10 numbers that re approximately 25% apart. These numbersare: 1.0, 1.25, 1.60, 2.0, 3.15, 4.00,5.00, 6.30, and 8.00, this series be ex-panded by using multiples of 10.

7.3.2 Rated Short Circuit CurrentThe short circuit current rating as specified by both standards corresponds tothe maximum value of the symmetrical current that can be interrupted by a



Table 7.6

IEC 294 Temperature Limits

, 1 InSF, 1 f jjinPlatedIn Air

InOil00

EXTERNAL TERMINALS oCONDUCTORS

Bare

Silver, Nickel, Tin plated 6505

INSULATING MATERIALS

ClassY

14080lass H115 155 Class

9030 Class

8020 Class

100 Class A



Table 7.7

ANSI C37.04 Temperature Limits



1.4

1.3

1.2

1.1

1

1 1.5 2.5 3.5

Contact Parting Time (cycles)

Figure 7.6 Factor S, Ratio ofSymmetric o Asymmetric Interrupting C apabilities.

circuit breaker. The values f these currentsfor outdoor breakersare based onthe R10 series.

Associated with the current value, which is the basis of therating, there are a numberof related capabilities,as they are referred byANSI,

or as definite ratingsas specified by IEC. The terminology usedfor these ca-

pabilities may differ, ut the significanceof the parameters is the same in both

standards. What it is important is to realize that both standardsuse the sameassumed X/Rvalue of 17as the time constant which defmes most ofhe of therelated requirements or all transient current conditions.

7.3.3 Asymmetrical Currents

In most cases, as described in chapter 2, the ac short circuit current has an

additional dc component, whichs generally referredas the “per cent dc com-ponent of the short circuit current.” The magnitude of this component is a

function of he time constant ofhe circuit, (X/R=

17), and of he elapsed timebetween the initiation of the fault and the separation of the circuit breakercontacts.



ArcExtinction i

*: - Dep*

Rela +

, Contactarting - Time Time+ Arcingpening+ -

Time

- nterruptingime _1Figure 7.7 Interrupting TimeRelationships

ANSI establishes an asymmetry factor, S which, when multiplied by thesymmetrical current, defineshe asymmetrical value of he current. The actualvalue of the factorS can be calculated using the following relationship

S= 1 + 2

-where theYO c equals the dc componentf the short circuit current t the timeof contact separation

ANSI however, to simplify he process, andas illustrated in figure 7.6, hasspecified definite values for S, and where they are referred to the rated inter-rupting time of the circuit breaker. The specified factors are: 1.4, 1.3, 1.2, 1.1

and 1.0 for rated interrupting times of 1, 2, 3, 5, and 8 cycles respectively.IEC establishes the magnitude of the dc component based on a time intervalconsisting of the sum of the actual contact opening time plus one half cyclfrated frequency. In essence, there s no difference betweenhe requirements ofthe two standards, except that n the ANSI way the factors are not exact num -bers but approximations for certain rangesf asymmetry.

Intem pting Erne. The interrupting time of the circuit breaker of the

mation of one-half cycle of relay time, plushe contact opening time, whichs thetime that t the breaker,from he instant the tripoil is energizeduntil he timewhere the contacts separate, plus the aximum arcing time of the circuit breaker.



Note that he contact parting times the summation of the relay time plushe open-ing time. These relationshipsre illustratedin figure

7.3.4 Close and Latch,or Peak Closing Current

The peak asymmetrical closing current is also referred as the close and latchcurrent or the peak making current. This current rating is established for thepurpose of defining the mechanical capability of the circuit breaker, its con-

tacts and its mechanism to withstand the maximum electromagnetic forcesgenerated by current. The magnitude of current is expressed in termsof multiples of the rated symmetrical short circuit current. NSI used to spec-

a factor, but recentlyt was reduced o for the sakeof mathematicalaccuracy. IEC specifies a multiplier. The difference between the two val-ues is due to the differencen the rated power frequencies.

The equation thats used to calculate the peak current is:

z = I (l - o s a t ) +[ "Iwhere:

= Peak of making current

Z = Symmetrical short circuit current

t = 0.4194 cycles= elapsed time to current peak

X 17

Ra 14for Hz) r- for Hz)c=--=-

7.3.5 Short Time Current

The purpose of this requirement is to assure that the short time heating capa-bility of the conducting parts of the circuit breaker are not exceeded. By

defrnition the short time current rating, or related capability,s the rms. valueof the current that the circuit breaker must carry,n the closed position, for a

prescribed lengthof time.The magnitudeof the current is equal to the rated symm etrical short circuit

current that s assigned to that particular circuit breaker andhe required lengthof time is specified as seconds by ANSI and as 1 second byIEC. Neverthe-less, IEC ecommends a value f seconds if longer than1 second periodsarerequired. The 1 second spec ifcation of IEC corresponds to their allowabletripping delay whichn the IEC document is referred toas the rated durationofshort circuit.



Even though ANSI requires a three (3) seconds withstand, the maximumallowable tripping delay is specified as two (2) seconds for indoor circuitbreakers and for outdoor circuit breakers that have aating of 72.5 kV or less,and for circuit breakers with voltage ratingst, or above 121kV, he time re-quirement is one (1) second, which also happens to be the tim e specified byIEC.

Recognizing, that the duration of he short time current does not have toeany greater than he maximum delay time that s permitted on a system,ANSIis in the process of adoptinghe shorter time requirements.

7.3.6 Rated Operating Duty Cycle

The rated operating duty cycle as is referred in ANSI, is known as the ratedoperating sequence by IEC.

ANSI specifies the standard operating duty s a sequence consistingof thefollowing operations;CO-l5 +CO, that is, a close operation follow edy an

immediate opening and then,&er a 15 second delay another close-open op-

eration.IEC offers two alternatives, one is 0 -3 m i n . 4 0 - 3 min.-CO and the

second alternative is the same duty cycle prescribed by ANSI. However, forcircuit breakers that are rated for rapid reclosing duties the time between theopening and close operations reduced to seconds.

ANSI normally refers to the reclosing duty as being an o p e n 4 sec.-close-open cycle which implies that theres no time delay between the open

and the closing operation. However in C37.06 the rated reclosing time, whichcorresponds to he mechanical resetting time f the mechanism, is given as 20or 30 cycles which corresponds o 0.33 or 0.5 seconds respectively dependingupon the rating of the circuit breaker, he reference to0 seconds is supposed tomean that no intentional external time delays cane included.

When referring o reclosing duties ANSI allows a current deratingactorRto b e applied, but IEC does not make any derating allowance. In reality thederating factorwas only applicable to older interrupting technologies and w

modem circuit breakers theres no need for any derating.Have you ever noticed the lights to go out, come back instantly, go out again

and then after a period of time, (15 seconds), come on again andf they gooutagain there is a long period before powers restored.? Well, what you have wit-nessed is a circuit breaker duty cycle including the fast reclosing option.

7.3.7 Service Capability

The service capabilitys a requirement that s only foundin ANSI, what it de-

fines is a minimum acceptable numberf times that a circuit breaker mustn-terrupt its rated short circuit current without having to replace its contacts.This capability is expressed in terms of the accumulated interrupted current



and for older technologies (oil and magnetic circuit breakers) a value of00

% is specified, for modem SF, and vacuum circuit breakershe required accu-mulated value s 800YO f the maximum rated short circuit current.

7.4 Additional Switching Duties

Aside from switching short circuit currents we know that circuit breakers mualso execute other types of switching operations. The requirements for theseoperations are defined in the corresponding standards but,as is the case with

most of all other requirements, there are some noteworthy differences whichare discussed in the sections that follow.

7.4.1 Capacitance Switching

All of the conditions that are considered o be related to capacitance switchingduties are listed separately below.

7.4.1.1 Single Bank

ANSI mandates thatall circuit breakers must e designed to meet he require-ments for the “General-Purpose Circuit Breaker” classification as is listed inANSI C37.06; meanwhile IEC lists duty as an optional rating, it does notassign any interrupting current values and only makeshe recommendation thatthese values e selected using he R 10 series.

The values of currents for switching single capacitor banks hat are appli-cable to general purpose breakers are pre-assigned by ANSI and they havebeen selected using he R10 series

7.4.1.2 Back Back Capacitor Banks

Again is an optional rating according tohe IEC standard and n line withtheir practice, regarding single capacitor banks,he interrupting current valuesare not specified. handles the back-to-back capacitor bank rating by de-fining a group of “Definite-Pwpose Circuit Breakers”. This is an optional rat-ing; and it is not expected that every circuit breaker would meet require-ment, unless it is specifically designed for purpose. In reality, however,most circuit breakers involving modem technologies are normally designedwith requirement in mind, and most manufacturers produce only one ver-sion rather than havingwo different designs.

7.4.2 Line Charging

Line charging requirements are included by NSI as part of their capacitanceswitching specifications and therefore all circuit breakers indoor and outdoor

have certain specific assigned requirements. In IEC standards the rating forline charging is only applicable to circuit breakers which are intended forswitching overhead lines and rated at 72.5 kV or above. There is also differ-



ences on the magnitude ofhe specified currents. At 72.5 kV and p o 170 kVIEC values re approximately to 50 YO ower than the ANSI values for gen-eral purpose circuit breakers, and about0 % lower than those .for definite pupose breakers. At the 245 and 362kV circuit breakers levelshe LEC specifiedline charging current ratings are approximately 25 % higher than the ANSIcurrents for general purpose breakers butre about 30 % and % lower thanANSI values for the respective breakers. At 550 and 800kV the required line

charging currents re the same in both standards.

7.4.3 Cable Charging

Cable charging is considered to be only a special case of capacitance switch-ing, and therefore is included, by ANSI with all of the other requirements orcapacitanceswitching.Since ECdoesnotmakecapacitanceswitchingamandatory requirement, then neithers cable switchingn their standard

7.4.4 Reignitions, Restrikes, and Overvoltages

These requirements serveo define what s considered to be an acceptable per-formance of the circuit breaker during capacitive current switching. The limi-tations that have been specifiedn the standards are aimedo assure that thefects of restrikes and the potential or voltage escalation, which have been de-scribed earlier, re maintained within safe limits.

The maximum overvoltage factor specified because it is recognized thatcertain typesof circuit breakers, most notably oil types,re prone to have re-

What rating does is to allow restrikes, only when appropriatemeans have been implemented withinhe circuit breakero limit the overvolt-ages to the maximum value givenin the respective standard. The m ost com-mon method used for voltage control is the inclusion which canbe the use ofshunt resistors. to control the magnitude f the overvoltage.

REFERENCES

1.

2.

3.

4.

ANSIC84.1-1982,VoltageRatings for ElectricPowerSystemsand

Equipment (60Hz).ANSIC92.2-1974PreferredVoltageRatings for Alternating-CurrentElectrical Systems and Equipment Operating at Voltage above 230 kilo-volts Nominal.L. Wagner, J. M.Clayton,C.L.Rudasill,and F. S. Young, InsulationLevels for VEPCO500-kVSubstationEquipment, EEETransactionsPower Apparatus and Systems, Vol. PAS. 83: 236-241, March, 1964.L. Wagner, A. R Hileman,LightningSurgeVoltages in Substations

Caused by Line Flashovers, AIEE Conference PaperP 61-452, Presentedat the Winter General Meeting, ew York Jan./Feb. 1961.



5. ANSI/IF!EE C37.04-1979, Rating Structure for AC High-Voltage Circuit

6. IEC-56 1987, High Voltage Alternating Current Circuit Breakers.7.TransientRecoveryVoltages on PowerSystems (asRelated to Circuit

BreakerPerformance),Association of EdisonnluminatingCompanies,New York, 1963.

Breakers Ratedon a Symmetrical Current Basis.

8. NEMA -1995 Alternating Current High Voltage Circuit Breakers.

9. IEEE 1-1969, General Principles for Temperature Limits in the Rating of

10. IEEE 4-1978 Standard Techniquesor High Voltage Testing.Electric Equipment.



SHORT CIRCUIT TESTING

8.0 IntroductionSince a circuit breaker representshe last line of defense for the whole electricsystem it is imperative to have a high degree of confidencen its performance,and confidence level only be attained by years of operating experi-ence, or by extensive testing under conditions hat simulate those hat a re en-countered in he field applications. Short circuit testing, whethert involves anindividual interrupter, or a complete circuit breaker, s one of the most essen-

tial and complex taskshat isperformed during he development process.Short circuit testings also important fromhe technical engineering designpoint of view because, n spite of all the knowledge gained about circuit inter-ruption and today’s ability to model the process, the modeling itself is based,to a great extent, on the experimental findings and consequently, testing be-comes the fundamental tool that s used for the development of circuit break-

Short circuit testing has always presented a challenge, because what mus

be replicated is the interaction of a mechanical device, he circuit breaker, andthe electric system , where as we already know, the switching conditionsvary quite widely depending upon the system configuration and therefore

the type of conditions which muste demonstrated by tests.Another challenge has always been the development of appropriate test

methods that overcome the potential lack of sufficient available power atthe test facility. One has to consider that the test laboratory should basicallybe able to supply the same short circuit capacitys that of the system for which

the circuit breaker that s being tested s designed, however, is not alwayspossible, speciallyat the upper end of the power ratings.

Presently it is ossible to test a three phase circuit breaker upo 145 kV anda maximum symmetrical current of 1.5 kA. Anything above these levelsmust be tested on a single phase basis, unless it is tested directly out of an

electrical network, a condition whichs highly unlikely.In the early days of the industry, however, practicallyall testing was done

in the field using the actual networks to supply the required power. Even to-

day, in some cases, direct field tests are still being performed. However,n themajority of he cases the testing is done at any a number of dedicated testingstations available n many countries around the world. The majority of these

ers.

257



test stations use their own power generators which have been specially de-signed for short circuit testing.

With a three phase capacity of MVA, N. V. KEMA, Amhem, TheNetherlands, is the largest test station in the w orld. n aerial viewof this testlaboratory is shown in figure Its subsidiary, KEMA-Powertest locatedinChalfont, Pennsylvania hashe largest capability MVA) in the US.

In additiono the above named laboratories therere a number of other im-

portant testing facilities aroundhe world; inNorthAmerica the newest labora-tory with a three phase capacityf MVA is LAPEM which is located in

Jrapuato, M exico, view of this modem and well equipped facilitys shown infigure

Additionally, in the US there is PSM in Pennsylvania, andin Canada thereis IREQ in Quebec, and Power-Tech in Vancouver. In Europe there Sie-mens and AEG in Germany, CESI in Italy, Electricity de Francen Fonteney,and ABB in both Sweden and Switzerland, MOSKVA in Russia, WSE in

Czechoslovakia, Warszava in Poland, and in Japan Mitsubishi and Toshibalaboratories.

Figure 8.1 Aerial view the N. V. KEh4A laboratory ocated in Arnhem, The

Netherlands B.V. KEMA,Amhem, The Netherlands).



Figure 8.2 Aerial view of the LAPEM laboratory located in Irapuato, Gto., Mexico

(Courtesy of LAPEM, Irapuato, Gto. Mexico).

8.1 Test Methodology

A great deal of latitudes given as far as short circuit current design testingfa circuit breaker is concerned. Because of the levels of power needed, thecomplexity of the tests and the very high costs involved with these test,latitude is not only a convenience but a necessity.

reduce the otherwise required amount of testing, it is perm issible toanalyze results from similar design tests and o use engineering judgment toevaluate these results. However, this judgment must be technically sound,supported by good data and backed-up by a strong knowledge abouthe char-acteristicsof the circuit breaker in question.

long as fllfficient evidence s gathered, andas long as it is satisfactorilydemonstrated that the most severe testing conditions re met, one certifythe interrupter’s performance by combining, n any order, the listed operating

test duties. The tests do not necessarily have to e performed in any particular se-quence, and that they do not even have toe done using the same intenupter, slong as the total accumulated current dutys reached with one individual unit.



MS TTB

Figure 8.3 Elementary schematicdiagram the basic circuit used for short circuitcurrent tests.

Since, even at the largest of the test stations, it is not always possible tosupply the full amount of power needed, alternate, appropriate methods thatrequire less powerut, that yield equivalent results hado be developed. Otherchallenges that had to be overcome had todo principally with signal isolation,surge protection, reliable control, and high speed instrumentation.

The difficulties, that be encountered during high current tests, onlybe appreciated when one considershe need for an absolutely proper sequentialcoordination of a number of events, and where a failure within the sequence

may result in aborting the test or what is worse in a disastrous failure, andwhere, during an individual test, there is only a single chance to capture thetest record.

In recent years, with the advent of high speed, high resolution, digital in-

shumentation and data acquisition systemshe quality and reliability tests,test records, andin general of the total accompanying documentation as beengreatly enhanced. Today it is possible to measure, store and later replay, athigh resolution levelshe complete test sequence, provides a very powerf

tool for the analysisof the events taking lace during current interruption.A t y p i c a l short circuit current test set-up, is illustrated by the schematic

diagram of figure The test circuit consists of a power source (G) whichbe either, a specially designed short circuit generator, suchs the ones that

are illustrated n figures 8.4 and 8.5, or the system's electric network itself.For the protection of the generator, or power source, a high capacity back-

up circuit breaker PUB) is used for interrupting the test current,n the eventthat the circuit breaker being tested (TB) would fail to interrupt the current.

Back-up circuit breakers,n the majorityof cases, areof the airblast type. anda single pole of the air blast back-up circuit breakers manufactured by AEGand whichare used a t LAPEM is shown in figure 8.6 .



Figure 8.4 Generator hall with 2,250MVA and MVA short-circuit test

tors. (Courtesy ICEMA-Powertest, Chalfont, PA,USA.)

I

Figure 8.5 View of LAPEM'S Test Generator (Courtesy of LAPEM, Irapuato, Gto.

Mexico)



‘

Figure 8.6 Single pole of an air-blast high current nterrupting capacity circuitbreaker used for back-up protection of test generator at LAPEM. (Courtesy of LA-

PEM, Irapuato, Gto. Mexico.)

In series with the back-up circuit breaker there is a high speed makingswitch ( MS ) which is normally a synchronized switch capable f independentpole operation and of precise control for closing the contacts at an specificpoint on the current wave. This typef operation allows precise controlor the

initiation of the test current and consequently gives the desired asymmetryneeded to meet the specific conditions f the test.

Current imitingreactors (L) are connected in serieswith the makingswitch, their mission is to limit the magnitude of the test current tots requiredvalue. These reactors be combined in a number of different connectionschemes to provide with a wide rangef impedance values.

Specially designed test transformersT), such as those shown in figure .7,

that have a wide rangef variable ratios are connected betweenhe test circuit

breaker and the power source. These transformers re used primarily to allowfor flexibility for testing at different voltage levels, and n addition to provideisolation betweenhe test generator and he test piece.



Figure 8.7 Test transformers installed at the LAPEM laboratory. Two transformers

per phase are used n this installation. (Courtesy W E M , rapuato, Gto., Mexico.)

Across the test breaker terminals a bankf TRV haping capacitors (C) isconnected and o measure these voltagest least one set f voltage dividers(V)

which are usually of a capacitive type are used. n most cases he short circuitcurrent flowing through the test device is measured by means of a shunthowever, measurements using current transformers are also often made, spe-cially for the currentat the upstream sidef the test piece.

When one is interested in investigating the interrupting phenomena at theprecise instant of a current zero, the use of coaxial shunts is highly recom-mended for distortion free measurements; however, caution shoulde taken tolimit the current flowing through the shunt since these type hunts normally

have only limited current carrying capabilities.The typical test set-uphat has just been described used for practically all

direct tests as the primary current source, and even or those situations wherethe available power s insufficient and where alternateest methods have beendeveloped, the primary sourceof current still is similar to he one that has beenjust described.

8.1.1 Direct Tests

A direct test method is one where a three phase circuit breakers tested, on athree phase system, and t a short circuitMVA level equal to ts full rating. Inother words his is a test where a three phase circuit breakers tested on a three



phase circuit at full current and at full voltage. It should be obvious that test-ing a circuit breaker underhe same conditions at which it is going to be ap-plied is the ultimate demonstration for its capability and naturally, wheneverpossible, th is should be the preferred methodof test.

8.1.2 IndirectTests

Indirect tests are those which permit he use of alternate test methodso dem-

onstrate the capabilities of a circuit breaker for applications in three phasegrounded or ungrounded systems.

The methods most commonly employed are:

1. Singlephase ests2. Two part ests

Unit ests4. Syntheticests

8.1.3 SinglePhase Tests

When one only in terms ofan individual interruptert is realized that,asfar as the interrupter is concerned, it makes no difference whether a threephase or a single phase power sources used, as long as the current and he re-covery voltage requirements for the test are fulfilled, therefore, the use of asingle phase test procedure s totally acceptable; however, when the finalplication of the interrupter is considered, and since n most cases it turns up tobe in a three phase circuit breaker, thenhe neutral shift of the source voltageand someof the potential mechanical interactions hat may occur between hepoles have toe taken into consideration.

In the first place, in a three phase applicationt the instant of current zero,and in the phase where he current interruption s about to take place,he inter-rupter itself does notknow that there are another two phases lagging slightlbehind on time. If the first phase which sees a current zero fails to interrupt,then the next sequential phase will attempto clear the circuit. basically,gives the circuit breaker an additionalwo chances to interrupt he current andtherefore, as it has been shown byW. Wilson [ l] , there is a higher probabilityof successNly interrupting a three phase current thanf interrupting a singlephase fault.

The oscillogram which is shown in figure 8.8 depicts the condition whereinterruption is attempted sequentiallyat each current zero.As it can be seen inthe figure, the first attempt is made on phase B (shownas B,), since no inter-ruption is accomplished the second attempt is made at the next current zero

which is in phase A (shown as AI), again the interrupter is not successful ontry and finally on the the current is interrupted in phase C. Theother two phases are seen to interrupthe current simultaneously t A-B.



Figure 8.8 Oscillogram of a three phase asymmetrical current interruption test, It l-lustrates the sequential attempts hat are made by the interrupter to clear the current at

each successive currentero of each of the threephases.



266 Chapter 8

B

Figure 8.9 Source voltage shift following the interruption of current by one phase

(first phase to clear).

During the first few microseconds following the interruption of the current

it only m atters that the groper transient recovery voltage be applied across theinterrupter, and since in a three phase circuit, as the high frequency oscillations

of the load side TFRV die down, and before the other two phases interrupt the

current, the source side power frequency recovery voltage is reduced to 87 YO

of the line to line voltage, due to the neutral shift, as seen in the vector diagram

for the power source voltage w hich is shown in figure 8.9 .

After the currents in all the phases are interrupted, the voltage in eachphase becom es equal to the line to neutral voltage, which corresponds to 58 %

of the line to line voltage. How ever, this reduction to58%

of the voltage oc-curs approximately four milliseconds after the current is interrupted by the first

phase, this is a relatively long time, especially with today’s interrupters, andwhen taken into consideration that it is long enough, after the interrupter has

withstood the maximum peak of the TRV then it is justifiable to expect that the

interrupter has regained its full dielectric capability and thus, in most cases it

becomes only academ ic the fact that the voltage reduction takes place.

Aside from the purely electrical considerations that have been given above,

the possible influence of the electromechanical forces produced by the currentsand of the gas exhaust from adjacent poles, should be carefully evaluated. It is

also important to carefully balance the energy output of the operating mecha-nism, to compensate for the reduced operating force needed to operate a singlepole, so that the proper contact speeds are attained. This last recomm endationis specially important when testing puffer type circuit breakers.



Short Circuit Testing 267

Naturally, these concerns about the possible pole interaction do not apply

to those circuit breakers which have independently operated poles.

8.1.4 Unit Tests

Unit tests can be considered to be simply a variation of the single phase testmethod which has been used almost exclusively for the extra high voltage classof circuit breakers where, as it should be recalled, it is common practice to in-

stall several identical interrupters in series on each pole mainly fo r the purposeof increasing the overall voltage capability of the circu it breaker.

This test method dem onstrates the in te m pting capability of a single inter-rupter from a multiple interrupter pole, provided that they are identical inter-

rupters. The test is performed at full rated short circuit current and a t a voltage

level that is equivalent to the ratio of the number of interrupters, used in thepole assembly, to the full rated voltage of the complete pole and where the

distributed voltage is properly ad justed to compensate for the uneven voltagedistribution that normally exists across each series interrupter unit and which is

due to the influence of stray capacitance, adjacent poles, and the proxim ity and

location of the ground planes. However, in any case the test voltage must be at

least equal to the highest stressed unit in the com plete breaker.When this test method is used the frequency of the TRV oes not change

but, due to the lower voltage peak of the individual in terrupter, the rate of riseof the recovery voltage is proportionally lower. This characteristic responseholds not only for terminal fault tests but also for short line fault tests.

Whenever the unit test method is utilized and, as it w as the case with the

previous test method, care must be taken to properly scale the mechanical op-

erating parameters to ensure the validity of the tests.

8.1.5 Two Part Tests

A two part test consists of two essentially independent tests. The first test, isone where the interrupter is tested at full rated voltage and at a reduced cur-rent. In the second test the maximum current is applied at a reduced voltage.

The idea behind this method is to test for the dielectric recovery regionwith the first portion of the test md then to complement the results by explor-

ing the thermal recovery region by means of the second test. Application ofthis test method has always been limited to the extra high voltage interrupters,

where, as it should be recalled, the TRV is represented by a w aveform that is

composed of an exponential and a (1-cos) function. When the two part testsare performed, the first portion of the test is made at full rated current and witha TRV that is equal to the exponential portion of the waveform. The second

part of the test is m ade at a reduced current but at full voltage and w ith a TRV

equal to the (1-cosine) component and where the requirements for the voltage

peak and for the time to reach this peak are verified.



This test method is often difficult to correlate with actual operating condi-tions and therefore it is somewhat difficult to justify. This is a test that was

frequently used prior to the development of the synthetic test methods de-scribed below and consequently, today, th is approach should only be usedwhen all other testing alternatives are not suitable.

8.1.6 SyntheticTests

Synthetic tests are essentially a two part test that is done all at once. Thetest is performed by combining a moderate voltage source which supplies thefull primary short circuit current with a second, high voltage, low current,power source which injects a high frequency, high voltage, pulse t a precisetime near the natural current zerof the primary high current.

Effectively, what has been accomplisheds to reproduce the conditions thatclosely simuIate those that prevail in the interrupter during the high currentarcing and the high voltage recovery periods.As ong as the sources, voltageand current, are not appreciably modified, or distorted byhe arc voltage then,the energy input into the interrupter, duringhe high current arcingime regionis no different than he energy input obtained from a ull rated system current,and voltage, because as we well h o w , the energy input to the interrupter isonly a functionf the arc voltage and notf the system voltage.

The behavior of the intem pter in the two classical regions of interest,namely the thermal and the dielectric regions, are evaluated by he high volt-age that is superimposed by the injected voltagdcurrent which when properlytimed embraces the transition point where the peak of the extinction voltagejust appears and the point where the peak of the recovery voltage is reachedthus covering the required thermal and dielectric recovery regions.

In general synthetic tests are performed on a single phase basis, and eventhough schemes have been developed that enable the tests to be made on athree phase circuit, it is only the largest laboratories thatare capable of doingso. In the majority of the facilities the high voltage source for these tests isonly available on a single phase basis; because in most cases some ofhe samepower limitations that existedor three phase direct testtill exist.

As it has been said before the synthetic test method utilizeswo independ-ent sources, one, a current source, which provides the high current, and whichfor all practical purposes is the same source that is normally used for directtests, and a second, a voltage source, whichn most cases consists of a capaci-tor bank that is charged to a certain high voltage that is dependent upon therating of the circuit breaker thats being tested.

There are a numberf synthetic test schemeshat have been developed, butin reality they all are only a variation f the basic voltage, r current injectionschemes. In actual practice, what is used most often byll testing laboratoriesis the parallel current injection technique.



4 TG

CTRV

E :

Source........

....””..

Figure 8.10 Schematic diagram of a parallel current injection synthdc test circuit.

8.1.6.1 CurrentInjection Method

The current injection method is illustrated in the schematic diagram of theequivalent circuitas shown in figure 8.10. method is characterized by heinjection of a pulseof current that s supplied by he high voltage source.

The high current source, as mentioned before, is composed of a short cir-cuit generator, a back-up circuit breakeror the protection f the test generator,a set of current limiting reactors, a high speed making switch andn additionalcomponent, an isolation circuit breaker (B) hose purpose, as its name im-plies, is to effectively separate, or isolate, the current circuit from the high

voltage circuit.The high voltage section f the circuit is made-up of a high voltage source

(VS) consisting of a capacitor bank that is charged to a predetermined highvoltage level. Connected in series with he capacitor is one side of a triggeredspark gap (TG) the other side ofhe trigger gap is connected to a group f fie-quency tuning reactors. Connectedn series with these reactors theres a shortline fault (SLF) TRV haping network, which consist of a combination ofca-pacitors and reactors that in most instances are connected in a classical p i (n)circuit configuration. Generally t is required thatat least five of these sectionsbe connected in series in order to accurately represent he TRV of a short linefault. SLF network however, is only used when the tests that are beingperformed simulate a shorti e ault condition

It is recommended that the frequency for he injected current e kept withinthe range of 300 to 1000 Hz. These limits depend primarily on the character-istics of the arc voltage. What is important is that the period of the injectedcurrent be at least four times longer thanhe transition period where a signifi-cant change in the arc voltage s observed. The magnitude of the injected cur-rent shouldbe adjusted that therate of change of the injected current (di/dt);and the rate of change of the corresponding rated power frequency current



(di/dt), are equal at their respective current zeroes. The timing for the initia-tion of the current pulse is controlled so that the time during whichhe arc isfed only by the injected current is not more than one quarter of the period ofthe injected frequency.

Parallel CurrentInjection. The schematic diagram of the circuit that wasshown in figure 8.10 represents the equivalent circuitconfguration that is used

for the parallel injection method, and in figures.11

and 8.12, the relationshipbetween the power frequency andhe injected currents shown.The test s initiated by closing the making switchMS), which initiates he

flow of the current i,, from the high current source (CS) through the isolatingbreaker and the test breaker (TB). As the current approaches its zerocrossing the spark gap s triggered and at time t,, (see figure 8.12),he injectedcurrent i2begins to flow. The current i, + i2 flows through the test breakeruntil the timet2 is reached. This is the time when the main current il goes to

zero and when he isolation breaker, separateshe two power sources. At timet3 the injected current is interrupted and the high voltage supplied by the higvoltage source provides the desired TRV which subsequently appears acrossthe terminalsof the circuit breaker thats being tested.

d,

.4 t

-”

d, _ _

2 4 8

TIME milliseconds

Figure 8.11 Relationship between primary currentand injected current in a

synthetic test parallel current injection scheme.



A

Figure 8.12 Expanded view of the parallel current injection near current zero.

Figure 8.13 Schematic diagram of a typical series current injection synthetic test

circuit.

Series Current Injection. The series current injection circuit shown sche-matically in figure 8.13 while in figures 8.14 and 8.15 the algebraic summationof the injected currents is shown. The notable difference between the seriescurrent injection and the parallel injection methods is that the high voltage

source, for the series injection version of the test, is connected series withthe high current source voltage.



At the initiation of the test the making switch is closed and at time t,

the spark gap is triggered thus, allowing the current i, to flow through theisolation breaker but in the opposite direction to that of the current il fromthe high current source. At time tl, when the currents il and iz are equal andopposite, the current in the isolating breaker is interrupted and during the tinterval from , to t3 the current that s flowing through the test breaker s equalto i3. current corresponds to the summation of the currents il+i2hat is

produced by the series combination of the high current and the high voltagesources. Following the interruption of the current i3at time t3 the resulting"RV upplied by he high voltage source appearsm s s he breaker terminals.

8.1.6.2 Voltage Injection M ethod

The voltage injection method,in principle, is the same as the parallel currentinjection. The only difference is that the output of the high voltage source isinjected across the open contactsf the test breaker following the interruptio

of the short circuit current which,as explained before, is supplied by the highcurrent source. The high voltages injected immediately after the current zerand near the peak of the recovery voltage that is produced by the power fre-quency current source.A capacitor is connected, in parallel a m s s the contactsof the isolation breaker, in order to effectively apply the recovery voltage ofthe current source to the test breaker. This test methods not very popular be-cause it requires a very accurate timing for the voltage injection. This timingbecomes a criticalparameter which n most cases is rather difficult to control

.oo

0 . 9 0

0.80

0.70

0 . 6 0

0.50

0.40

0.30

0.20

0.10

0 . 0 0

0 2 4 6 8 10

TIME mill iseconds

Figure 8.14 Relationship between primary current and injected current in a syn

thetic test seriescurrent injection scheme.



Figure Expanded viewof the series current injection near current zero.

8.1.6.3. Advantages and Disadvantageso Synthetic Tests

As is the case with any of the other test methods, there are a number of advan-

tages and disadvantages hat are associated with synthetic tests. The principaladvantage that should be m entioned is that these tests re of a nondestructivenature and therefore they are ideal for development test purposes, where theultimate limitsof the device can e explored without destroyinghe test model.Also the synthetic test method is the most adequate, andn some cases he onlyway of performing short line fault tests.

The disadvantage of synthetic tests s that these tests are primarily asingle oop est,whichexplains why they are considered to be anon-

destructive test, and although a reignition circuit be used to force a longerarcing time,or a second loop of current,t still is very difficult o do a fast re-closing with extended arcing times. Another disadvantage is that method



is not suitable for testing interrupters which have an impedance connected nparallel with the interrupter contacts in which caset is likely that the fu ll re-covery voltage can not be attained due to the power limitations the highvoltage source.

8.2 Test Measurements and Procedures

Test procedures and instrumentation, naturally vary in accordance, not onlywith the test method, ut also with the purpose of the tests that are being per-formed. The test instrumentation, can be significantly different, for example,when doing interrupter development tests, than when doing verificationr cir-cuit breaker performance tests.

In the investigative portion f the tests it is likely that special attention wbe paid to the phenom ena occurring at, or very near, current zero, where a

higher degreeof resolution is needed. In these tests, what s of interest is what

takes place around current zero n a time region which s normally in the mi-crosecond range, while for verification tests,r complete circuit breaker inter-ruption tests, and withhe exception the "RV, the time of interest is in themillisecond range.

For development tests in most casest is important to have accurate meas-urements of arc voltage, interrupter pressure, post arc current and other verydefinite measurements, depending solelyn the type of information that is be-ing sought. While, because the tests that re made to demonstrate the capabil-

ity of the circuit breaker, and the requirements that have been to demon-strate its compliance with he existing standards, are well definedn the appli-cable standards [4] and because in most cases it s assumed that a signifi-cant number of development tests have already been performedhe needed in-

strumentation is what may be considered as conventional, consisting meas-urements phase currents, phase voltages andV .

Since the proceduresfor development tests are rather specialized and spe-cific in nature according to he circumstances, or to the objectives the tests

being performed it is then difficult to provide f m uidelines for the instru-mentation tobe used and for the tests procedures to e followed; however, hetechniques that are to be described, and which re used for the design verifica-tion tests also be used for other purposes, suchs exploratory tests.

8.2.1 Measured Parametersand Test Set-Up

It goes without saying that the fundamental parameters of phase currents andcorresponding phase voltages must e measured. In addition to these parame-

ters it is advisable, specially when testing vacuum circuit breakers, to makemeasurements of the amplified arc voltage. This measurement be veryhelpful in determining the precise instant where contact part occurs. It also



serves to determine the stability of the arc and the effectiveness of the inter-rupter at the transition point f the current regions.

Another valuable and important measurement, that sometimess neglected,is the measurement of the breaker contact travel which, when everything goesright on a test may not be needed but, for those times when there s a failurethis measurement would help to answer questions such as: Was the circuitbreaker fully open?, didt stall?, was t fully closed?.

The test current s generally measured using a low resistance shunt, andnsome occasions, or even better accuracy and response, a coaxial shunts used.The voltage measurements are usually made with a capacitive compensatedvoltage divider, and the measurement is preferably made on a differentialmode to avoid distortions due to possible ground shifts.

8.2.1.1 Grounding

The one essential requirement s that grounding of the circuit should be either

at the source, or at the test breaker but nott both places.

8.2.1.2 Control Voltage

Although in the standards it is indicated that the rated control voltage of thedevice should be used, one could take exception to because in most casesit is very convenient to use a higher control voltage, perhapss much as 20%

over the rated value,as the means for minimizing the variation in the time th

it takes to open the contacts.n addition to the higher control voltage t is ad-visable to use a dc supply whenever possible, rather than anc supply. isdone with he sole objective f minimizing the variation of the contact openingtime, which s important becauseof the need for proper controlof the point onthe wave where it is desired to break the circuit so that the proper currentsymmetry is achieved during he test..

8.2.1.3 Close-Open Operations

In some laboratories, performing a close-open operation at high asymmetricalcurrents can be quite harmful to he health of the circuit breaker,this happensbecause when, both, closing against the peak of a fully asymmetrical currentand opening at the point of maximum total current is being attemptedthere is a risk that he peak of the closing current may be substantially higherthan what is required. This risk exists because, even though super excitationsapplied to the test generator, he asymmetrical valueat contact part can not eeasily achieved and when the symmetrical value of the current is raised to

compensate for the rapid dc decrement, the initial peak of the current is alsoproportionally increased. One method that has been used successfully to over-come this difficulty has been he addition of a series reactor which limitshe



peak of the current during the closing operation, but as soon as the circuitbreaker is closed, and beforehe opening is initiated, a switch thats connectedin parallel withhe reactor is closed, effectively shorting out the reactance anthus increasing he current at he time of contact separation.

8.2.1.4Measuring the TRV

Despite the act that in most cases the TRVs measured during he actual inter-

ruption testing measurement is not alwaysa valid one. Because of the in-fluences exerted byhe characteristics of he arc voltage, the post arc Conduc-tivity, and the presence of TRV modifying components suchs capacitors andresistors that may have been installed acrosshe contacts and which will mostdefinitely affect the TRV wave of the circuit. Therefore unless the abovementioned effectsare insignificant and he short circuit current does not havedc. component, comm only obtained test records can not be used and specialprocedures must e utilized to determinehe inherent TRV ofhe test circuit.

The two methods most commonly used are:

1. Currentnjection2.Capacitornjection

Networkmodeling/calculation

Currenf injection. This method consists of injecting a small power frequencycurrent signal into a de-energized circuit and then interruptinghe injectedrent using a switching devicehat has negligible arc voltage, and postrcrent. A device with such characteristics couldbe a fast switching diode, onethat exhibits a reverse recovery timeof less than 100 nanoseconds. When us-ing these type of diodes, it is permissible to havea shorting switch across hediode if there is a possibility that the current carrying capabilities ofhe diodecould be exceeded. The shorting switch will have to e opened shortly beforethe zero current crossing where the TRV measurement is to be made. Themeasurements of the current and voltage waveforms must be made using in-strumentation suitable or high speed recording.

Capacitor injecfion. Here a low energy capacitive discharge is used as thesource for the injected current. In reality method is no m e r e n t theprevious method except hat in case the ac. source or the injected currenthas been replaced by he dc. voltage stored in a charged capacitor. Since thefrequency of the discharged currents proportional to capacitance ofhe sourceand the inductanceof the circuit, then the frequency of the measured voltagedefines the inherent TRV.

For best results he frequency of the discharge current for these measure-ments shouldbe (0.125 of the equivalent natural frequencyf the circuit beingmeasured.



Network modeling/calculation. This method consists of either an analog or adigital modelingof the of the network that s being evaluated. The accuracy f

method, of course, depends uponhe selection of the appropriate represen-tative parameters of the circuit that s being evaluated.

8.2.2 Test Sequences

As it was the case with regard o ratings, so it is for testing; the and IEC

test requirements are not exactlyhe same. Nevertheless, the required tests aresufficiently close in both documents, and with only a little extra effort inchoosing equivalent test parameters, specially for TRV, and by adding a fewextra tests, he requirements of both standards cane concurrently met.

8.2.2.1 LEC Requirements

The short circuit capability, accordingo the IEC standards, s demonstrated bya test series consistingf five test duties. Test duties1,2 and 3 consist of three

opening operations which be demonstrated using the standard duty cyclewhich, as it can be recalled, consists of the following sequence, 0-t-CO-t’-COwhere t is either 3 minutes or 0.3 seconds depending on weather the circuitbreaker is rated for reclosing duty or not. These test duties are perFormed withsymmetrical currents of 10% 30% ((20%) and 60 % ((10%) of theratedshortcircuitcurrent espectively.The TRV requirements ncludeaslightly higher voltage peak and a significantly shorter time durationo reachthe voltage peak. These est duties are performed with the intentf simulating

the interrupting behavior f a circuit breaker n the eventof a fault in the sec-ondary side of a transformer, where, as the current is reduced the TRV be-comes more severe, s it has been verified y Harner et,al. [ 5 ] .

Test duty 4 consists of the prescribed operating duty cycle. The openingoperation is made under symm etrical current conditions, while the m ax im masymmetrical current peak muste attained during he closing operation n or-der to demonstrate the close and latch capability of the circuit breaker. Thesymmetrical current or the opening following he closing is obtained by delay-

ing the trip sufficiently SO that the dc and ac transient components have de-cayed to an asymmetrical valuef less than20%.

Test duty 5 is a test similar to test duty except that both the opening andclosing operations are made withan asymmetrical current. The asymmetry ofthe current is that which corresponds to a time constant of approximately 45

ms and which corresponds ton Xm value of 14 for 50Hz. or 17for 60 Hz.The asymmetrical valueof the current is determined using the actual con-

tact opening time the circuit breaker to establish the elapsed time that is

measured from he time of current initiation o the point of contact separationtime thus, determines he asymmetrical value for the test by followinghe

procedure which was described earlier in chapter.



TABLE 8.1

Test for Demonstrating the Short circuit Ratingof a High Voltage Circuit Breaker

(Three Phase Test)

13 and 1 CO

or20 1 .58V

14 a d COor2 1 .58V

150 - 1 5 s - 0 o r 0 - 1 5 s -

16 or 0-15s- ,.58V

S8VK

CO

CO

Current Interrupted atContact Part

.07o . l3< 50

.4 .6 > 50

I <50I

SI I >50

KSI I >50

RKSI I >50

Smaller of

Smaller of1.15 Ior

~ . 9 t0 .951 I <20



8.2.2.2 ANSI C37.09 TestSequences

The ANSI test requirements are specified in Tables 1 and of the C37.09 teststandard for highvoltagecircuitbreakers.The ncludedTable8.1,corre-sponds to Table 1f the above mentioned test standard, except thathe last twocolumns have been omitted. it can be seen from Table 8.1, a rather exten-sive test series would be required if every test is performed precisely as de-scribed; however,as it will be shown later, when one looks closely to the re-quirements it is found that some f the tests canbe combined, while for others,alternate methods are used dueo limitations ofhe testing facilities.

The first three ANSI sequences are quite similar, if not identical to those

required by IEC 56. They are composed of one opening and one close-openoperation at reduced current values. The current values are specifiedwithin a percentage range that encom passes the specific percentage valuesgiven byIEC 56, however, for test duties 1 and ANSI specifies asymmetricalcurrents. But as it ispresently recognized and agreed upon; that, givenhe in-tent of these tests, which as it was previously stated, are meant to simulate asecondary fault and to demonstrate the breaker capability to withstand higherrates of TRV, he tests should e performed with symmetrical currents.

Test duties 4 and 5 are symmetrical current tests, both are basically thesame in terms of operating sequence, except that test 5 is performed at thelower operating voltage and the higher short circuit currentas defined by therated voltage range factorK. This test, consequently, s only applicable to in-

door circuit breakers since the K factor for all other breakers is equal to 1.Furthermore, as it has been repeatedly said, for those circuit breakers that util-ize modern technologies, the inverse relationship between voltage and currentis no longer applicable, ands it is usually the case, the maximum currentbe interrupted at the maximum voltage. What this implies is that, in mostcases, for newer technology circuit breakers it may be possible to omit testduty 5 , provided that the higher current is used. Figure 8.16 shows a typicaloscillogram of this particular test duty

It is important to note that some alternativesre allowed for test cycle,one being a Close-Open-l5 sec -Close-O pen sequence however, it is advis-able to avoid using sequence simply because t is a little more difficult ocontrol the symmetry of the current during the opening operation that followsimmediately after the closing operation, is so even though, therequirements are applicable only to one opening and one closing operation.Since it is easier to adjusthe time of the opening of the contacts for an initialopen only operation, one should take advantage of condition and subse-

quently control he closing of the contactsso the desired valueof asymmetry isachieved leaving the remaining opening test duty operation to occur after ashort time delay neededor the current to regaints symmetry.



l" 4196.Vlm

0.212 0.335 0.459 0.563 0.706 0.830.954 1.073 1.m1

Figure 8.16 Typical record oscillogram o f a threephase C los e-O pe n4 .3 sec.-

Close-Open operationwith anasymmetrical current.



The above comments are applicable not only to this particular test dutyutalso to all of those tests that with a closing of the test circuit breaker,orthat require multiple operations. Furthermore, it is also possible, as it is donein E C 6, to perform test duty with a time interval of.3 seconds and thusmeeting the test requirements for a fast reclosing duty by essentially havingcombined test duties 4 and 9. However, when done the possible need ofderating the interrupter m ust be considered. Derating was almost always ap-

plied to older style circuit breakers, buthen considering he application ofhenew technologies, is a requirement that does not appear to have muchvalue left.

Test duties6,7 A andm, re also a demonstrationf the standard test dutycycle, except that n test duty 7A, which is intended for circuit breakers ratedabove 121 kV, a second test duty cycle is performed 15 minutes of thefirs. Remembering that these standards were written, primarily with andoil circuit breakers n mind, it is understandable that the possible effectsf the

stored heat and the drop in pressure due to the previous interruption hado beinvestigated. With today’s vacuum orF6 circuit breakers t is known thatdoes not constitute a problem and therefore the only justification for the ex-tended duty cycles is to accumulate the required 800% interrupted currents,and the demonstration of the worst switching condition. The time intervalbetween duty cycles alsos no longer that important andhe tests canbe m adewithin a time frame of only a few minutes as i t may be dictated strictly byconvenience. Again, when performing these test duties it is recommended to

with an opening rather than with a close-open operation, and thats so theasymmetry of the current can be properly controlled. It is also important tonote that the high asymmetrical values, those above that are specified inthe test tables are unrealistically high, for a circuit breaker with normal inter-rupting timesof 3 or 5 cycles and or a system having n X m value of 17, andtherefore the alternative is to test with a lower asymmetry, sayo 45y0, or asis suggested by he standards, to adjust for the total current t the time of con-

part but to reduce the symmetrical rms. of the current and test with the

higher totalrms. asymmetrical current value. Nevertheless, aest made with areduced symmetrical rms. and with a higher dc components valuable becauseit provides at least some indicationf the breaker capabilityor applications insystems wherehe Xm values are higher than 17.

Test duty11 calls for the circuit breaker to e closed against aull value ofa fault current, to current for a time equal to the maximum ratedpermissible trip delay (2 seconds for breakers rated 72.5 kV and below, or 1

second for higher voltage ratings) and then to interrupt theull rated symmetri-

cal short circuit current. This is a very diffcu lt test to perform, mainly, be-cause the thermal ratingf the test generators are almost always exceeded, an



this kind of power is simply not available on a three phase basis at any testlaboratory, for the higher voltage rated circuit breakers.

In some laboratories test may be performed synthetically, using twopower sources, the closing operations done using a high voltage and highrent source; initial closing portion f the test is obviously, no different thanany of the routine closing operations thatre performed as part of the other re-quired test duties. However, immediately after the circuit breakers closed the

high voltage sources removed by means of an isolating switch and then acurrent, that is supplied by a low voltage source, is superimposed upon theclosed circuit breaker contacts. The high currents maintained for the requiredlength of time and afterwards, when the time requirements are met, the highcurrent source is removed and the high voltage high current source is onceagain inserted into the circuit that the current interruption portionmay beperformed.

Prior to he publication of the 1964 edition of the C37.04 and C37.09

the requirements for test duty 11 were promulgated sepm tely, and close and

latch, and a momentary current rating were published. The testing to demonstratethese requirements was done independently and in separate operations. This ap-proach is still taken inmany because of tacit agreementabout thecality of the present requirements. What it must be remembered is that the closeand latch test is made to prove the mechanical capability of the circuit breakthat demonstration is made several times while performing the complete testsequences required for breaker certification. The requirement for carrying theur-

rent for an specific time duration (longerhat required inhe previous test dutis met in the very next test sequence, where the test demonstrates the short timethermal capability of the contacts. It is generally agreed that if capability isbuilt-in into the contact the higher contact temperature at the momentcontact part does not have any negative effect upon the interrupting capabcircuit breaker.

Test duties 13 and 14 are supposedo demonstrate the capability of the cir-

cuit breaker to interrupt a line to ground faultn a grounded system. The tests,

naturally, are performed onsingle phase basis and are only required wheheprevious test duties have been done on a three basis, these tests are not re-quired when allhe testing is done using a single phase source.

What it is important to realize s that these wo tests rather unique becausethey areperformed ata voltage of .58% of the line to line rated voltaand that is in contrast with the .87% of the line to line voltage whichs used

when aU he testing s done witha single phase source. lso the specified fault cur-rent,for test duties 13 and 14,as chosen by applying the inverse voltagecurrerelationship and where the factor 1.15 is simply the ratio between thevoltage V and 0.87, (U0.87= 1.15); however. same criteria is not applied to ttest when madeon single phase basis.



Finally, it is not clear what theTRVvalue should be for test duties and14, however what is generally agreed is that the peak of the recovery voltageshould be between and times the maximum rated lineo line voltage.

It should be expected that in he near future, as the ANSI standards are re-vised and as the harmonization process with the international standards pro-gresses further, these small inconsistencies will probably disappear.

The last two test duties, and are also single phase tests, but are ap-

plicable only o outdoor circuit breakers. The aim f these tests is to prove theshort line fault capability of the circuit breaker. However, in regards to thesetest duties, another inconsistency is found between two of the ANSI docu-ments. states that all outdoor circuit breakers must e capable of inter-rupting a short ine fault, but in it is said that it is not needed to dem-onstrate capability for circuit breakers rated kV and below.

Experience has shown that the short line fault requirements are not con-fined only to the very high voltage circuit breakers and, s a matter of fact, a

number of circuit breaker failures which cane directly attributable to the in-ability to properly handle the shortine TRV have been reported. This fact isnow widely recognized and even though, it is still unofficial, short line faulttesting is being performed in all outdoor circuit breakers, regardless of theirvoltage rating.

8.2.2.3 Most severe switching conditions

The mostsevereswitchingconditionsaregenerally eferred as to the case

where he ntermpter is subjected oamaximumarcing ime, or what itamounts to a conditionof maximum arc energy input. Basically what it is in-tended by testing for the most severe switching conditions is to show that inthe worst case the intermpter in any one of the poles of the circuit breaker iscapable to withstand the maximum arc energy input.

The most unfavorable conditions wille those where he contact separationoccurs duringa minor current loop and wherehe duration of the arcing time sjust short of the minimum arcing time requiredor interruption by that particu-

lar design. As we already know, if the minimum arcing time requirement isnot met then interruption willnly take place afteran additional half cycle fcurrent, which forhe worst case condition will constitute a major current loop.

Since in a three phase system under symm etrical current conditionsherent zeroes occurat a sixty electrical degrees interval then, theres a mil-lisecond window to accommodate the variation in the possible arcing time.What this means is that, with symmetrical currents in a three phase system,fone of the phases fails to clear the fault at its first current zero, this phase

most likely will never see its true maximum arcing time because one of theother phases s likely to intermpt the current before the original phase reachesa repeat current



T

-1

-1.5 lFigure 8.17 Relation of arcing time for symmetrical currents and or M e r e n t contact

parting windows for circuit breakerwith a minimum arcing time of 4 ms.

T3d,

1

0.5

-0.5

-1

Figure 8.18 Relation ofarcing time for symmetrical currents and for different contactparting time windows for circuit breaker with aminimum arcing time of 10ms.

and, even though, the energy input to then t q t e r which failed to clear thecurrent at its first attempt continues to increase, because whenhe in one ofthe phaseshas been intenupted theemaining phases will evolve into a singlephase current whichs then ntempted by the remainingpoles in the to-

talenergy willbe less thanwhat be expected from a fuuysingle phase fault thatas a maximum arcingtime.



rent zero where intdrmption should occurs designated by he letters A, B, andC. This designation matches the identification thats given to the correspond-ing arcing windows. Figure 8.17 represents a circuit breaker that has a mini-mum nominal arcing time of approximately 4 ms. This arcing time is gener-ally a characteristicf vacuum interrupterswith currents greater than5 kk

Figure 8.18 shows a minimum arcing timeof 10 ms. which is representa-tive of a SF, circuit breaker, wherehe range of the minimum arcing times s

generally between 7o 13 ms.One suggested method that cane used to determine the maximum arcing

time is illustrated in figure 8.19. For the first interruption the contact part isadjusted so that it occurs at a current zero of anyf the three phases,n our ex-ample phaseA was selected first. By observing figure 8.19 we can see that:)if the minimum arcing time s less than ms. then interruption will take laceat B. b) if the minimum arcing time is less than 6 ms. then interruption willoccurs at C, and c) if the minimum arcing times less than 8 ms. then the

rent will be interrupted at A. For the second test the contact part is advancedby approximately 2.5 ms. to tz and in way the arcing time window, thatwe had mentioned befores not exceeded andf the test s repeated the point ofinterruption will be the same as in the previous test., any further advances ofthe contact part will then result n a shifting of the corresponding current zerowhere interruption takes place.

1

1

0

d,

-0 .S

-1

-1 lFigure 8.19 Method for obtaining the maximum arcing time for a symmetrical three

phase currenttest.



The same situation, as it was described for the symmetrical currents, doesexist with asymmetrical currents, except that nowhe arcing time window s nolonger a constant2.77 ms. but it depends upon the dc. and ac. components ofeach of the phase currents. Figures and serve to illustrate he shift ofthe interruption point or a circuit breaker that has a minimumrcing time of 4

ms., when the contact parting points displaced by about5 ms. What it shouldbe noticed is that for the conditions shown in the figures the maximum energy

input to the interrupter, shown, in arbitrary per unit values; does not occur onthe fmt phase to clear but rather on one ofhe last phases to interrupt. How-

there are now two phases in series that are interrupting the current thusmaking the interruption an easier task.

Of the procedures given, by he respective testing standards, for obtainingthe worst switching condition, the one described in IEC 56 provides a moreclear and definite approach nd it yields the same results that re being soughtby ANSI.

1 . 5

1

0.5

0

- 0 . 5

- 1

- 1 . 5

-2Figure 8.20 Arcing time variation depending on pointf contactpart for asymmetri-

cal currents. Assumed minimum arcing time 4 milliseconds, a comparative value of

arc energy inputE (arbitrary per unit value) s shown in enclosedbox.



1 .5

1

0.5

0

-0.5

-1

-1.5

-2

P a r t I

*E=331

+E=65

4 E = 2 6 8-igure 8.21 Arcing time variation produced by advancing approximately4.5 the

point contact part from the original position in figure 8.21. Assumed mini-

mum arcing time is still 4 milliseconds, and the sam e com parative value of arc energyinput E (arbitrary per unit value) is shownin enclosed box.

The referenced test procedure callsor the following sequences:

1. for the first operation the point of contact part is set so that the requiredvalue of the total current is obtained

2. for the second test, he initiation of the short circuit currents shifted by60electrical degrees and if inhe first test he first phase to clear dido aftera major current loop, the trip time is advanced by approximatelyelectrical degrees otherwiset is advanced by only 5 degreesfor the third operation the procedure of the second operation may be re-peated and the same criteria about the first phase to clears applicable.

The only objection that perhaps may be raised about the procedure is inconnection with the change thats required of the inception angle of the shortcircuit current, which is required so that the asymmetries of the currents aretransposed between the phases, it seems that it would be simpler to change

only one parameter,he contact parting time, rather than two parameterst thetime, considering that the results for similar contact part conditions for eachindividual phase, woulde the same.



1.5

1

0.5

0

-0.5

-1

-1.5

-2

TIME milliseconds

Figure 8.22 Three phase asymm etrical currents times tl, h, and t3show the changes

in the contact parting time to obtain the required maximum energy conditions Contact

parting times are advanced, or delayed by approximately 4.2 ms (45 electric degrees)for a 3 cycle breaker.

In figure 8.22 it is shown how one may accomplish the required intenup-tion the current after a portion of a minor loop plus a full major loop, by

varying thepoint o f contact part. T h i s is done by controlling he trippingof the circuit breaker so that, for our example which corresponds to a circuitbreaker with a minimum arcing timef 4 ms., the contacts will separate onhephase with the highest asymmetry (phaseA in our example), at a point on theminor loopthat is less than 4 m. rom its next current zero, and whichn the

case that is illustrated corresponds o a current that is close to the peak of theminor loop; interruption then willake place &er the full major loop. For thenext test the trip is advanced by 4.2 ms. and the intenuption will occur onphase C, and finallyfor the last test the trip signal is retarded by4.2 m. ndintenuption then will occurs in phase B. For other circuit breakers havinglonger arcing times a similar procedure cane used to determine he requiredchanges in the contact parting time.

Another point that one may find to be questionable, and that is only be-

cause of the performance characteristics today's technologies, is the needfor satisfying the maximum arcing time requirements while performing testduties 1,2 and 3.



1 . 5

l

0 . 5

0

-0 .5

-1

-1 .5 lFigure 8.23 Graphical representation o f a method for obtaining maximum arcingtimes during single phase symmetricalcurrent tests.

The currents, and he energy input during these est duties is relatively lowand therefore it seems that a better option coulde to do these tests with theminimum arcing time rather than withhe maximum arcing time, sincet theselower current levelst may be expected that he interrupter may quench the arcsooner and with n shorter gap which may note able to withstand he fullre-

covery voltage.A more realistic and n easier demonstration for the maximum arcing time

capabilities of an interrupter is obtained with a single phaseest because, in a

single phase test he arcing time be controlled fairly without having o beconcerned about the interference from the other phases in the event that theinterrupter fails to intermpt at the fifit current zero.

For single phase tests with symmetrical currents (refero figure 8.23), themaximum arcing time can b e obtained by first, adjusting the contact part to

coincide with a current zero crossing (at time,), interruption will occurat ei-ther the first current zero (point 2), for circuit breakers that have minimumarcing times less than eight milliseconds, or at the second current zero(point 3) for circuit breakers with arcing times greater than eigbt millisecond

For the next test, he point contact part Q is advanced by approximately4.5 milliseconds, under these conditions interruption will takelace at the firstcurrent zero (point 1) for breakers with minimum arcing times of less than 4

milliseconds (vacuum circuit breakersor example), at the second current zero

(point 2) for circuit breakers that have a minimum arcing time greater thanmilliseconds but less than 12 milliseconds,r at the third current zero (point)for circuit breakers that have minimum arcing times greaterhan 12 millisec-



' T0 . 5

3ei

0

-1 .5

-2 l

Figure 8.24 Graphical representation of a method for obtaining maximum arcingtimes during single phase asymmetrical current tests.

onds. Applying the above described procedurewill ensure that the interrupterhas been subjected o the longest possible arcing time when interruptingmetrical

For an asymmetrical condition, see figure.24, a similar proceduremay beused as follows:

First, the region where he contacts must part to satisfy the total current rquirements should be chosen, time is then adjusted to coincide with thebeginning of the minor loop showns time t,, if intem ption occurs at the first

current zero (point 1) the time should be retarded by about milliseconds, totime t2, ntemption then will most likely takelace at current zero correspond-ing with point 2. This last test will demonstrate the maximum arcing timeconditions for circuit breakers that have minimum arcing timesn the range of4 to about 12 milliseconds. Since is generally the range of minimum arc-ing time of modem circuit breakers he above test will be sufficient to verifythe maximum energy input condition within reasonable limitsof accuracy formost of oday 's vacuum andF, high voltage circuit breakers.

REFERENCES

1. Walter W. Wilson, Casjen F. Harden, Jmproved Reliability Fromcal Redundancy of Three Phase Operation of High Voltage CircuitBreak-



IEEE Transactions of Power Apparatus and Systems, Vol. PAS-90,No. 2: 670481, MarcWAprill971.

2. ANSYIEEEC37.081-1981, EEEApplicationGuide for SyntheticFaultTesting of AC High-Voltage Circuit Breakers Ratedn a SymmetricalCur-rent Basis.

3.ANSYIEEEC37.09-1979, EEE TestProcedurefor AC High-Voltage Circuit Breakers Rated on a Symmetrical Current Basis.

4. IEC 56, High-voltage alternating-current circuit-breakers.5. R.H. Hamer, J. Rodriguez,“TransientRecoveryVoltagesAssociated

with Power-System Three Phase Transformer Secondary Faults’’ Transac-tions of Power Apparatus and Systems : 1887-1896, NovemberDecember1972,

6, IEEE Tutorial Course, Applicationof Power Circuit Breakers, IEEE PowerEngineering Society,93 E H 0 388-9-PWR: 74-84.

7. W. P. Legros, A. M.Genon,M. M. Morant,P.G.Scarpa, R. Planche,C. Guilloux, Computer Aided Design of Synthetic Test Circuits for HighVoltage Circuit Breakers, IEEE Trans. Power Deliv. Vol. 4, No. 2: 1049-1055, April 1989.

8. G. St.-Jean, M, Landry, Comparison of Waveshape Quality of ArtificialLines Used for Short Line Breaking Tests on High Voltage Circuit Break-

IEEE Trans. Power Deliv. Vol. 4, No. 4: 2109-2113, Oct. 989.9.L.M. Vries, G. C. Damstra, Areignition nstallation with Triggered

Vacuum Gaps for Synthetic Fault Intenuption Testing, IEEE Trans. Power

Deliv. Vol. PWRD-1, No. 2: 75-80, April 1980.



PRACTICAL CIRCUIT BREAKER APPLICATIONS

9.0 Introduction

In an earlier chapter, a circuit breaker was defined as: “a mechanical devicewhich is capable ofmaking,carryingandbreakingcurrents.”Itwasalsolearned that there are a number different types of circuit breakers and anumber of different system conditions where theyay be applied and, that s aconsequence, it is to be expected that unusual conditions, or situations thatdeviate from what s considered to be an standard, or normal condition canbe

encountered, and hat n most nstances hese non-standard conditions willhave a significant effect upon the applicationf a circuit breaker.

The primary aim f this chapter s to provide some simple and practical an-swers to questions relating to non-standard applications, they can be used tofacilitate the evaluation a given circuit breaker for a given application.Naturally, there are so many unique conditions that it will not be possible tocover all the foreseeable applications; but, we will concentrate on those thare most frequently encountered.

Among the most undamentalandoftenaskedquestionsaboutcircuitbreaker applications re those relating to:

a) Overload currents and temperature riseb) High X/R systemsc) Systems with frequencies other than 50 or 60 Hz.d) Size of capacitors banks or capacitor switching operationse)High TRV applications

Highaltitude nstallationsg) Low current, high inductive load current switchingh) Choice between or vacuum.

9.1 Overload Currents and Temperature Rise

The continuous current carrying ratingf a circuit breakers predicated on hepremise hat heambient emperatureand the elevationwhere the circuitbreaker is applied is within the limits that have been set byhe applicable stan-

dards. As ambient temperatures vary widely on a daily and on a seasonal ba-sis, to provide a constant base of reference an ambient temperature ofwas selected as the upper limit. This selection was based on the meteorologi-



cal reports provided by the S Weather Bureau, which indicate that the ambi-ent temperatures in the continental US very seldom exceed this upper limit.The maximum standard altitude as it will be recalled is 1000 meters (3300

feet), over sea level. This elevation is considered to e within the limitsof thestandard operating conditions because the majority of the applications worldwide do not exceed this limit..

The altitude limitations are related to the lower density and therefore les

cooling capability of the air at higher elevations; while, the ambient tempera-ture is directly related to the total temperature of the equipment whichs dic-tated by the limitations that are established basedby the characteristics of thematerials that are employed in the constructionf the circuit breaker.

To evaluate the behavior of the circuit breaker under conditions which adeemed to be different than those consideredas standard; be them largerrents, higher ambient r higher altitudes, he problems reduces to onef estab-lishing the ultimate temperature rise required to dissipate, by convection andradiation losses the watts generatedt specific currents.

For electrical equipment that has only few ferrous materials componentsthe losses are essentially proportional to the squaref the current. However, asthe temperature increases,so does the resistance and if the losses were due tothe conductors alone then the loss curve will rise slightly faster than the sfunction. But in most circuit breakers there is a significant amount of ferrouscomponents and the losses due to eddy currents are approximately proportito the 1.6 power of the current. Considering these to values to be the extremelimits and based primarily on practical experience an exponent valuef 1.8 hasbeen established as a suitable compromise.

When the circuit breaker has reached its ultimate temperature rise for agiven steady state current it is clear that the total losses must be dissipatedsince the equipment s then no longer storing any f the generated heat. Theselosses are divided essentially into radiation and convection losses. The formervaries approximatelyas the difference, raised to the fourth power,f the abso-

lute temperatures, while the latter varies at a much lower power of the tem-perature. The above statements about the losses are given only s general ref-erence and it is not implied, nor is it necessary to calculate these dissipationfactors before solving the problem at hand.

9.1.1 Effects of Solar Radiation

For outdoor applications, in addition to he heating produced by the loadrent and by the ambient air temperature, one must be aware of the possibleadditional heating that may result from the effects of solar radiation. On thebasis of field tests and accumulated operating datat has been determined thatin most cases a maximum temperature riseof approximately 15°C (27'F) maybe expected on the conducting partsf the circuit breaker.



""""T I "" . ~""" "" " _l I l I II l I I l

l

""""

I I I

""IIII

""

I

II

""l l I I Il I I I l

b b i20

Monthly Normal Tempera- (deg. C)

Figure 9.1Altitude correction factors or continuous currents.

When the circuit breaker is operated at a monthly normal maximum ambi-ent temperature above25OC ( 7 7 O F ) derating of he continuous current capabil-ity of the circuit breaker may e necessary. The derating factor tobe used [l]

is given in figure 9.1 as a function of the maximum monthly normal tempera-ture as given by theUS Weather Bureau.

9.1.2 Continuous Overload Capability

There are times when it becomes necessary to operate a circuit breaker withload currents that are higher than those corresponding to the full rating f thecircuit breaker. Operation under these conditions is possible provided that theambient temperature is consistently below the maximum allowable

find the allowable current that can be carried at a given ambient tem-perature the following equations given.



where

= allowable current= rated continuous current

8, = allowable hottest spot total temperature8, = actual ambient temperaturee,= allowable hottest spot total temperature rise at rated continuous

rent

In order to prevent that he maximum temperaturesat any given point andfor any given material are not exceeded when the load current is adjusted tocompensate for the lower ambient temperatures the following rules should e

observed

1. the actual ambient temperature is less than the component withthe highest specified values of allowable temperature limitations shouleused for determining e,,, and 8, ,

2. the actual ambient is greater than the component with the lowestspecified values of allowable temperature limitations should be used fo rdetermining e,,,=. and 8,.

By using these valuesor the alculations it is ssured that he temperaturesof any parts of the circuit breaker would notbe exceeded. However, in manycases it would perfectly safe to exceed these limits without the risk of impaing the performance or the life of the circuit breaker.This is because gen-erally the minimum limits of temperature are at breaker locations that arereadily accessible to the operating personal while the maximum temperaturesare allowed at external locations, that are not accessible to operating perso

these parts are excluded from consideration, higher values of permissiblecurrents will be obtained from the calculations but, these calculated valuesshould be used judiciously and only when the particularities ofhe design arewell known o insure that there is no possibility of damage to adjacent lowertemperature materials.

As an example, let us consider a circuit breaker that hasa continuousrent rating of 1200 . This circuit breakers going tobe applied at an ambienttemperature of The maximum allowable temperature rise is limited to

by its bushings. It is desired to find what is the maximum current capa-bility for breaker under the given conditions.



Using the given equation, the maximum allowable current or th is breakeris determined tobe amperes.

9.1.3 Short Time Overloads

The permissible time durationf overload currents, which re predicated upona specific temperature ceiling, are intimately concerned with the thermal ca-pacities of the components and hence with the rate of temperature

with time. Therefore, to determine what would e a safe overload, in terms ofcurrent or time, we must find the interrelation of the above three factors: thewatts that are generated at specific currents, the ultimate temperature rise re-quired to dissipate by convection andy radiation losses and the naturef the

of the temperature with respect to time, towards the ultimate tempera-ture rise for any particular current.

For simple structures, where circuit breakers maye considered to be one,it is fairly accurate to assume that the temperature increases exponentially to-

wards the ultimate temperature rise. means that the of the tem-perature is progressing in such a way thatt is continuously consuming a fixedproportion of the remaining temperature rise in equal intervals of time. Theexponential temperature rise curve reaches of ts remaining risein an in-

terval time equal tots time constantThe time constant on critical circuit breaker parts generally alls betweento minutes, and value may be specified by he circuit breaker manu-

facturer, butf not, it would be safe to use a valuef minutes.To calculate the time durationf a short-time overload he following equa-

tions shouldbe used.

1.8

Y = -

where e = maximum allowable total temperature8, = actual ambient temperature C



Zi = initial current carried by the breaker during the preceding 4 hoursZ = short time load current in amperesZ = rated current in amperes

= thermal time constant of the circuit breakerf, = permissible time in hours for carrying overload current

The emergency load current capability forcircuit breaker s treated on he

referenced application standard2] by establishing emergency load current carrying capability factors which are based on an ambient temperature of 40(C,for two distinct overload allowable periods; four hour and an eight hour pe-riod, for the numerical values these factors refer toigure 9.2.

According to the rules,t ispermissible to operate 15°C abovehe limits oftotal temperature or the four hour period and0°C for the eight hour period.

The following guidelinesre a direct quote from the referenced standard:Eachcycle ofoperation is separate,andno ime-current ntegration is

permissible to increase the number of periods within a given cycle. However,any combination of separate four hours and eight hours emergency periodsmay be used, but when they total sixteen hours, the circuit breaker shalle in-spected and maintained before being subjected to additional emergency cy

For ambient temperatures other than the0°C maximum specified, the pro-cedures that were previously outlined maye used.

k

BP

W

Total Temp deg.C

Figure 9.2 Overloading factors four and eight hours intervals.



Four-Hour Factor. This factor shall e used for a cycle of operation consistingof separate periods of no longer thanhours each, with no more than four sucoccurrences before maintenance.

Eight-Hour Factor. This factor shall be used for a cycle of operation consist-ing of separate periods of no longer than eight hours each, with no m ore thantwo such occurrences before maintenance.

Each cycle of operations separate, and no time-current integrations permissi-ble to increase the number of periods within given cycle. However, any combi-nation of separate four hours and eight hours emergency periods maye used, butwhen they total sixteen hours, the circuit breaker shalle inspected and maintainedbefore being subjected to additional emergency cycles.”

For ambient temperatures other than the maximum specified, the pro-cedures that were previously outlined maye used.

9.1.4Maximum Continuous Current at High Altitude Applications

Generally, applications at high elevations do not pose much of a problem becausthe interrupters that are used in today’s circuit breakersre sealed devices and con-sequently the contact structure itself is not affected by the high altitude and the

lower air densities. Those parts of the circuit breaker whichre exposed to the out-side atmosphere are not generally the most critical parts and more importantlyas

the altitude increases it isess likely that the ambient temperature would reach theupper limit.In the event that it is desired to calculate the maximum allowa

current at high elevations the appropriate multiplying factor thats plotted in figureas a h c t i o n of the maximum monthly normalempmture as given by theUS

Weather Bureau. it may be seen in the figure, even at meters (10,000it)and at ambient temperatures of the circuit breaker is capable, of Carrying its

full rated continuous current.To determine the short time overload characteristics of a circuit breaker it

is possible to calculate what the overload woulde at sea level, and then m ul-tiply this value by the factor obtained from figure for the corresponding

ambient temperature.

9.2 Interruption of Current from High /R Circuits

As it has been explained before the short circuit ratings assigned by the stan-dards are based on n X/R value of at Hz. or at Hz. naturallyconstitutes only a compromise average value. which is representative of themajority of the applications. But that still leaves a significant number of ap-plications where the Xm of the system is greater than the values adopted by

the standards. When this happens then these questionsarise: What rating doIneed in the circuit breaker? What s a circuit breakerI ave good for?



1.2

1.15

g 1 .1

d

1.05

i l

8

0.95

0.9

+Amb. 25 -\

"Amb. 30

-x- Amb. 35

+ mb. 40

I500 l000 1500000500 3000

ALTITUDE meters

Figure 9.3 Altitude correction factor.

First, it should be remembered that circuit breaker ratings are based onhesymmetrical current values and that these symmetrical current ratings are hevalues that should not e exceeded. However, it is also known that the currentasymmetry is a functionof the time constant, or Xm of the system and there-fore, for a constant contact opening time of a circuit breaker, the totalcurrent at the point contact separation increases s a functionof the increase

of the asymmetry which inurn is the resultof the increase in the time constaof the circuit.

Whenever a fault occurs at a location that is physically close to a largepower generator, there may be a significant ac exponential component of theasymmetric current which decays very rapidly during theirst few cycles afterthe initiation of the fault. However thisac decay is generally considered not tobe significant at locations that are distant from the power generator, where thshort circuit currents fed through wo or more transformations,or those appli-

cations where the reactance of the system is greater than 1.5 times the sub-transient reactanceof the generator.



1.8 l I1.7

1.6

- 1.5

91.4

Y 1.3

1.2

. l

1

0 20 400 8000

ELAPSED TIME milliseconds

-X/R=17+X /R =20 + X R = 2 5 + X R = 3 0

-x-XlR=35 + X R = 4 0 + X R = 4 5 + X R = 5 0

+ R =60 + /R =70 -+-X R =80

Figure 9.4 Factor “S” for asymmetrical currentvalueswith dc decrement only.

For high voltage circuit breakers applications, n practically all cases, it is

possible to ignore the effects of the ac transient component and to consideronly the dc component. Obviously, this introduces some error in the calcula-

tion, especially in the distribution class circuit breakers, where there are per-haps only a few instances where closer attention should be paid to the effectsof the ac transient component. The expected errors however, would be in theconservative side and the results would lead to specify a circuit breaker withigher rating thus assuring a greater margin safety. W e m ust also recognizethat the error is within what can e considered tobe acceptable operating lim-its since a rigorous mathematical analysisf the complete circuit s not feasibleand moreover, where the data available or the values of the components very

rarely would have an accuracy better than 10%. Furthermore, there is some-thing tobe said about operating experience which has shown that thiss a rela-tively conservative and valid approach.



For the discussions that follow, and to answer the two questions posed elier, a plot of the ratioof the total rms. asymmetrical to the symm etrical rms.current at contact separation, plotted, as a function the elapsed time, fromfault inception for different X / R values is given in figure 9.4. The ratio be-tween the two currents is calledhe factor “S” and is to be used as a multiplierto establish the related values between the symmetrical and the asymmetricalcurrents or vice versa.

The first step on the application process is to determine the magnitude ofthe short circuit current which can be calculated using either of the methodsthat were given in chapter 2. Next, it is necessary to calculate the X / R valuefor the circuit.

If the X/Rvalue is equal or less than17 then it ispossible to simply choosea circuit breaker with a symmetrical current nterrupting capability hat is

equal or greater than he calculated short circuit current.If the X/R factor of the circuits greater than 17, thent is necessary to de-

termine the elapsed time, or contact parting time, which according o its defi-nition is equal to a one-half cycle relay time plushe contact opening time ofthe circuit breaker. Once this value has been established, from figure 9.4he Sfactors may be determined for the calculated / R and for the standard value o17.Multiply hecalculatedshortcircuitcurrentby he S factor hatcorre-sponds to the higherXm o obtain the total value of the asymmetricalcur-rent then. divide this value by the factor S corresponding to the X / R of 17.This is the minimum interrupting current rating that s needed for this particu-

lar application.For example, let us assume a 121 kV circuit capable of delivering a short

circuit current of 14,000 amperes and having an /R of 50. It is desired to se-lect, from a table preferred ratings, a 5 cycle circuit breaker with a contactparting time, or elapsed time of50 milliseconds. From figure 9.4 the S valueforanX/Rof50is1.39andforX/Rof17is1.1.

I , = 1 4 , 0 0 0 ~.39= 19,460 amperes

I , = 19,460+ 1.1= 17,69 amperes

were

IT = Total rms current at /R = 50

I, = rms symm etrical current t X/R= 17

The results above indicate that a standard circuit breaker having a preinterrupting ratingof kA or higher should e selected.

Now, contemplating the case where for example, a standard kA, 3 cyclecircuit breaker is available and it is desired to apply this breaker on a system



that has anX/Rvalue of 80. The maximum interrupting capability of this cir-cuit breaker for this application can be determined by simply multiplying therated symmetrical capability by the ratio of the standard circuitfactor to theS factor corresponding to the high X/R system. The elapsed time, or contactparting time, for this circuit breakers ms. and the two S factors, from fig-ure 9.4, are 1.2 and 1.56 respectively. The S factor ratio is then equal to 0.77

and the product f this factor times the symmetrical current ratings 20 .77=

15.4 kA which represents the new rated symmetrical current forhe applicationon a system withan X/R of 80. Ifwe were to assume that circuit breakerwas being installed in close proximityo a source of generation and if were toconsider the effects of the transient ac component then the multiplying factorfor the highX /R condition, as given in figure 9.5 would e approximately 1.42and the ratio between factors is 0.84. Thentempting capability now becomes16.9 kA. Comparing the results we see that the difference s within the rangeof accuracy of the circuit components and that n m y case the error is on the

safe side.

1.5

1.4

1.2

Elapsed Time milliseconds

Figure 9.6 Asymmetrical factor“S” including ac decrement.



When evaluating the applications on systems that have a higherXm thanthat assumed by the standards, he only consideration given o far is o the in-terrupting capability of the circuit breaker however, attention should also begiven to the maximum current peak that can occur sincehis current peak is afunction of the time constant,or X/R of the circuit and care must e taken notto exceed the maximum current peak that has been assigned to the circuitbreaker. In figure 9.6 the multiplying factor or thepeak currents is plotted asa function of the system’s Xm and for the example given above we find thatthe current peak multipliers are 2.6 and 2.775 and the peak currents are 202.6 = 52 kA and in the worst case 16.9kA 2.775 = 46.9 kA for thestandardrating and or the higher X/R,respectively.

9.3 Applications at Higher and Lower Frequencies

In general the question of how a circuit breaker will behave when appliedt a

higher or a lower frequency than thator which it was designedsposed mostlyin applications involving medium voltage and up to 145 kV equipment. Themost common low frequencies consideredare those which are associated withtransit applications the two most popular ones are 25 and 16 and 2/3 Hz. Higfrequency applications are rare and usually they are associated with very spe-cialized applications n the medium voltage class.

Vacuum interrupters have demonstrated that it does not make any differ-ence whether theyare used on 50 or 60 Hz..At lower frequencies, primarily,

because of a lack of applicable data,t is customary to reduce the interruptingcapability of thente m pte r as a function of2 t .

The derating functions expressed as:

where

I, = Original symmetrical interrupting rating circuit breakerI2 = Derated capabilityfi = normal rated frequency50 Hz)X = desired low frequency

For example or an application at 25 Hz. the short circuit current shouldelimited to 0.707 of the rated current t 50 or 60 Hz.

For applications of vacuum circuit breakers at frequencies higher than 60

Hz, experience has shown that the interrupter is not sensitive to the rate ofchange of current (di/dt) however, this does not mean that because of theshorter periods higher currents are allowable. The maximum short circuit



rent is still limitedby the peak valueof the current whichs that of the originalrating at 50 or 60 Hz.

For SF, circuit breakers the situation is different, as it was described inchapter 5 , the interrupters, in most cases, are sensitive o the rate of change ofcurrent. This would imply that at lower frequencies higher currents can e in-tem pted; however, in he case of a puffer circuit breaker, even whenhe inter-rupter can successfully handle the extra input energy causedy the longer pe-

riods, it is unlikely that the speed and travel requirements can e properly ac-commodated by conventional designs. Therefore it is generally advisable thatunless it is specifically sanctioned by the manufacturer, applications at lowfrequencies should be avoided. The same could be said for high frequencyapplications first; because of the derating required which is a function of thedi/dt at the high frequency current and secondly, becausehe TRV would also

increase as a functionof the system frequency and byow we are cognizant ofthe high sensitivityof SF, interrupters toTRVvalues.

9.4 Capacitance Switching Applications

Capacitance switching applications involve not only interrupting capacitivecurrents, a subject which has been dealt in previous chapters, but alsohe en-

ergizing of overhead lines, cables and capacitor banks. Capacitor banks andcable systemsmay be either isolated,or back to back connected.

isolated capacitor bank, or cable, s defined as follows “Cables and

shunt capacitors shall be considered isolated if the maximum rate of change,with respect to time,of the transient inrush current does not exceedhe m axi-mum rate of change of the symmetrical interrupting current capability of thecircuit breaker athe applied voltage.”

This is represented mathematically byhe following expression.

where

= rate of change of inrush current

a= n.f or for Hz

f = power frequency

V,, = ratedmaximum oltage

Vwlicd =maximum applied voltage



C B 2L C 1

Figure 9.7Single line diagram of a typical installation showing capacitor banks con-

nected on a back-to-back fashion.

I = rated short circuit current

The following definition is given for back to back capacitor banks orcable circuits: "Cable circuits and shunt capacitor banks shall be consideredswitched back to backf the highest rate of changef inrush current on closinexceeds that specifiedas themaximum for which the cableor shunt capacitorbank can b e considered isolated." In simpler terms, isolated cable circuit, orcapacitor bank means that only one cable, or one bank is on one bus, whileback to back means that more than one cable.r capacitor bank s connected tothe same bus. typical system illustrating a back to back connection of ca-pacitor banks s shown in figure .7 .

9.4.1 Isolated Cable

Energizing a cable by closing the contacts of a circuit breakeresult in the floof a transient inrush current. The magnitude and the rate f change of inrush

current is, among other factors, principally a function of the applied voltage, thecable geomeby, the cable surge impedance and the length of cable.

Since it is given that the circuit breaker must e able to withstand the mo-mentary short circuit requirements of the systemhe transient inrushcurrent toan solated cable is never a imiting factor n he application of a circuitbreaker.

9.4.2 Back to Back Cables

When switching back to back cableshigh magnitude transient inrush currents thatare accompanied by a high initial ratef change may flow betweenhe cablesbeing switched. The transient inrush current is limited by the surge imped



CB Cable 1

4 7Cable 2-

B 2 ab

Figure 9.8 Back to back cable connected cables: (a) circuit single line diagram, (b )

equivalent circuit for calculation of current.

where

E,,,= peak of applied voltage

= cable surge impedanceL = total circuit inductance between cable terminalsIm = rated peak inrush current

= trapped voltage on cable being switched

9.4.3 Isolated Shunt Capacitor Bank

The magnitude and he frequency the inrush current resulting from energiz-

ing an isolated capacitor bank is a function of the point on the wave of theapplied voltage where the contacts were closed,f the capacitance and induc-tance of the circuit, of the charge on the capacitor at closing time and of nydamping resistance containedn the circuit.The transient inrush current that flows into an isolated capacitor bank is lessthan the available short circuit current at the terminals of the circuit breakerandsince hemomentarycurrentratingof hecircuitbreakerreflects hemaximum short circuit current of the system then, the isolated capacitor bank

inrush isof no consequence for the application ofhe circuit breaker.When switching an isolated capacitor bank the value of the inrush current

and its frequencys given by he following expressions.

and



Figure 9.9 Simplified equivalent circuit for back to back capacitor banks calculatioand where LI andL, re the inductance between capacitor banks and s the busin-ductance.

where

ELL= lime to line system voltage

L, = system line inductanceC =bank's capacitance

9.4.4 Back to Back Capacitor Banks

When energizing capacitor banks in a back to back configuration high magni-tudes and high frequencies f the inrush currents canbe expected. The magni-tude of the current being limited only by the impedancef the capacitor bankand by the inductance between the banks thatrebeing energized.

Atypical single line circuit representing the back to back condition isshown in figure 9.9 and the equations that define this condition are given be-

low.

E ti =



wherec,=c,+ c,

L1 nd L, = Inductance between capacitor banks including inductanceof the banks

Lb= inductance of the buss betweenhe banks

The typical values for the inductance between capacitor banksre given in

reference and are reproduced in Table 9.1 below.

Table 9.1

Rated Maximum

between banksf bus (Wft)oltage&V)

T y p i c a l Inductancenductance per Phase

& above

--------



9.4.5 General Application Guidelines

The first factor that needs toe taken into account when consideringhe appli-cation of a circuit breaker for capacitance switching is the type of circuitbreaker to be used. If, by any chance, the candidate is an oil circuit breaker,then very serious consideration must be given to avoid the possibility of ex-ceeding the given ratingfor the maximum frequency f the inrush current. Oilcircuit breakers are extremely sensitive tohe inrush current and failing to op-erate below the maximum limits could leado catastrophic failures.

With modern circuit breakers the frequency of the inrush currents a lesserconcern for the circuit breaker itself; however, in most cases still constitutesthe limiting factor because of other equipment in the system such as linearcouplm and current transformers and also due to the effects of the inducedvoltages on the control wiring andhe possible rise on ground mat potentials.

The problem with linear couplers and current transformerss related to thesecondary voltage thats induced a m s s the terminals and which be calcu-lated using the following formulas:

For linear couplers

E , = - x L C R x I2

For current transformers

E , = - x C C T R X I X L , . , , ,2

where

E, = secondary voltage across device terminalsLC R & C TR = linear coupleror current transformer ratioJ; = system power frequency

= transient frequency.

I = transient currentL,, = relay’s inductance c0.3 ohms

Just for illustration purposes lets us assume an inrush currentf kA andan inrush frequency f Hz; or a linear coupler and current transformerboth having a ratiof 1000 to then the calculated secondary voltage are:

Linear couplerE,7=- - = 13,330 volts

1000

Current TransformerE, =- -000x x 0.3 = volts



The results from this example indicate that voltage limiters need toe usedto protect either the transformer,r the linear coupler.

The magnitudeof the peak current, as long as it does not exceed he maxi-mum peak of the given close and latch capabilityf the circuit breaker, shouldnot present any problem even ifts greater than the values that are publishedsstandard ratings. Nevertheless before exceeding he rating values it should beverified with the manufacturer of the circuit breaker that there are no design

features thatmay say otherwise.

9.4.5.1 Lim iting Inrush Frequency andCurrent

When'found that t is necessary to limit the magnitude andhe fiequency of theinrush current whats recommended is the useof:

1. Closing resistors or inductors, which are inserted momentarily duringthe capacitor energizing period and thenre subsequently bypassed.

2. Fixed reactors which are permanently connected to the circuit. Thisprocedure reduces the efflciency of the capacitors and increases thelosses of the system

Synchronous switching where the closing of the contacts is synchro-nized that it takes place ator very near the zero voltage effec-tively reducing the inrush current. The poles f the circuit breaker orthis operation must e staggered by2.7 milliseconds.

9.4.5.2 Applica tion of Circuit Breakers Near hunt Capacitor Banks

Special care must e taken to insure that line circuit breakers are not appliedoswitch capacitor banks because for certain types of faults the contributionmade to the fault by the out-rush current from the capacitor banks will exposethe circuit breaker to currents that are, in most cases, greater than those en-countered on back to back switching, which meanshat not even those breakersrated as definite purpose circuit breakeray be suitable for these applications.The solution to this problem may the inclusion of pre-insertion and openingresistors on the circuit breakers or the installation of current lim iting reactorsin series either with the capacitor banks,r with the individual circuit breakers.A single inediagram llustratingacircuitconfiguration eading to largeamounts of out-rush currents s shown in figure .10.

9.5 Reactor Current Switching, HighRV Applications

In general switching of reactor currents s associated with small magnitudeof

currents, high frequency transient recovery voltages, and high overvoltages. Itis thenconceivable hat nsome of thoseapplications nvolving eactorswitching, when current limiting reactorsre connected in a close proximity to



Figure 9.10 Diagram illustration of typical installation where the fault contribution

the capacitor banks may producecurrents that exceed the capabilities o f circuit breaker

CB2.

the circuit breaker, the resulting RV may exceed the limits for which the cir-cuit breakers have been designed and tested.

it should be recalled SF6 circuit breakers are likely toe more limited n

their capability to withstand higher rates of recovery voltage than similarlyratedvacuumcircuitbreakers.Consequentlyfor heseapplications,and if

available, a vacuum circuit breaker coulde a better choice. Nevertheless, ansince we also h o w that at voltages higher than kV the most likely choicewould be an SF, circuit breaker. One solution to reduce he rate of rise of therecovery voltageso that the circuit breaker is not overstressed, specially dthe thermal recovery period which takes place duringhe first 2 microsecondsafter current interruption is to add capacitors, either in parallel to the inter-rupter contacts, or connected from line to groundt the terminalsof the circuitbreaker. For this simplistic approach the size of the capacitor can be calcu-lated by simply assuming that theRV s produced by an equivalent seriesCcircuit and where the initial time delay ( td) , hat is now required has to begreater than microseconds, is given approximately by:

td = 2 in microseconds

where

Z = surge impedance



C,,,= externally added capacitance in microfarads

It is also possible, and this is a more realistic approach, specially whendealing with applications at the higher end of he voltage scale, to utilize cir-cuit breakers that are equipped with opening resistors, or to use metal oxidesurge arresters applied directly tohe circuit breaker.

Opening resistors re connected in parallel tohe main interrupting contacts

of the circuit breaker and constitute an effective method for the reduction ofovervoltages andfor the modification of TRV. The value of the closing resis-tor should be approximately equal tohe ohmic reactance ofhe reactor.

Another approach that be used and which will be discussed in moredetail in he next chapter s the synchronized opening ofhe circuit.

when the circuit breaker s used to switch shunt reactors thatre connectedto the bus its fault current interrupting capability should be determinedn rela-tion to the full requirement of the system but, if the circuit breaker is used to

switch shunt reactors that are connected to transmission lines the fu ll fault ca-pability may not be required although the short time and the momentary ca-pabilities of the circuit breaker should e at least equal to the ratings of the cir-cuit breaker thats providing the primary fault protection or the circuit.

9.5.1 Ferroresonance

As described above the use of capacitors helps to improvehe TRV withstandand therefore he interrupting capability of a circuit breaker,n other instancescapacitors are also placed acrosshe contacts on poles that have multiple inter-rupters with the purpose of equalizing the voltage distribution across he indi-vidual interrupters. In general. for these purposes the larger the capacitancethe greater the improvement; however, the down side to the use of larger ca-pacitors, aside from the cost and added complexity, is the possibility of creat-ing a ferroresonant condition between the capacitor and the potential trans-formers that may be connected to lines thatre de-energized.

This condition is created when there is a transformer connected to he bus,

in which case thereis

a series connection of he capacitors and he transformeras illustrated in figure 9.11. The equivalent circuit, basically represent a volt-age divider (X,,, X , -X,) and when the capacitive reactanceX , equals the in-

ductive reactanceX,,, of the transformer then, at least in theory the voltage atthe bus will become infinite, butn reality the voltage is limited due to he nonlinear impedance of the transformer and its magnitude is determined by theintersection of the capacitive voltage with he transformer’s saturation voltageand whenever the intersection point is below the knee of the saturation cu v e

then the overvoltage could be severe and damage to the transformer takeplace. An expeditious solution for problem is to add a low ohmic resistorconnected across the secondary of he transformer.



a

CB

b

Figure 9.11 Relationship circuit components for ferroresonance: (a) equivalent

circuit and (b) single line diagram.

9.6 High Altitude Dielectric Considerations

The application of a circuit breaker at elevations greater than 1000 meters(3300 ft) dictates that its dielectric capabilitiesbe reduced due tohe lighter airdensity. For sealed intenupters, he withstand capability acrosshe contact gapis not affected and only those insulating structures whichre exposed to the air

atmosphere should be considered for derating. Keeping in mind that what isdesirable, but not always attainable,s that in case a flashover due to exces-sive overvoltages, said flashover should occur externally to the contacts. De-pending on he type of interrupter used and on the designf the circuit breakerit is possible that there s air dependent insulation located n a parallelpath tothe SF, or vacuum insulation acrosshe breaker contacts. This implies that inthese cases he altitude derating must e applied.

In those cases where the design provides fllfficient coordination between

the non-atmospheric and the atmospheric paths t is conceivable that the pos-sibility of applying the equipment either without derating or with a limitedderating may be considered; however approach must be carefidly evalu-ated to assure that the insulation coordination with the rest of the equipmentinvolved on the particular installation is not compromised in any way. Fur-

thermore it will be necessary to provide adequate protection for the circuitbreaker in the form of properly rated arresters located botht the line side andat the bus side and may prove to be uneconomical when compared to a

fully rated circuit breaker.The applicable derating factorK) is given by the following equation ands

shown in figure 9.12.



0 1000 2000 3000000 5000 6000

Elevation above sea level ineters

Figure 9.12 Correction factor for the dielectric withstand of the components a cir-

cuit breaker exposed to air ambient at elevations other than sea level.

K = ewhere

K = derating factor

H = altitude where circuit breakers to be applied in meters

Once a correction factor has been determined,he next is to calculatethe operating voltage rating at the standard conditions. The selection f a cir-

cuit breaker that has a rating equal r greater than hat which has been calcu-lated will generally take caref the power frequency andhe impulse dielectric



withstand requirements at he new altitude. For example, let us chose a circuitbreaker for a 145kV system tobe used at 3000 meters above sea level.

The factorK is equal to e-(3000-100018150)

= 0.782

The maximum operating voltage rating at standard conditions forplication is equal to l / 0.782 145 = 185 kV. The closest higher standard

rating is 242 kV.Now let suppose that the maximum voltagef the system is 121 kV in-stead of 145 kV what we find with these new conditions is that the requiredmaximum operating voltage is 155 kV. Nowwe are faced with a situationwhere it may be possible to use a 145 V circuit breaker protected by properlysized arresters or to opt for selecting a 242kV circuit breaker. If a dead tankcircuit breaker is being considered and the system is grounded then the rec-ommended choice should be the 145V rated circuit breaker.

9.7 Reclosing Duty Derating Factors

The need for derating factors for applications involving rapid reclosing or extended duty cycles depends primarily on the type of circuit breaker thats be-ing used. Being more specific, it should apply only to oil or air magnetic cir-cuit breakers. This s due to the fact that following an interruption these circbreakers need additional timeo cool and to regain their dielectric capabil-ity. SF6 and vacuum circuit breaker generally do not need he extra time and

therefore there should be no need for reducing their interrupting capability;nevertheless, there may be some cases where the manufacturer chooses o doso and consequently when the application of these breakers involves eithermore operations, or shorter time intervals between operations than whats es-tablished as the standard duty cycle its advisable to consult with the manufac-turer.

When a derating factor needs to be applied it can be calculated used therelationship that is given below, and which s applicable within the following

constrains.

1. the duty cycle does not consistsf more than operations2. all operations taking place within a 15 minute periodare to be part of the

3. a period of 15 minutes between operations is considered to be sufficientsame duty cycle.

for the initiation f new duty cycle.

The derating factor isbtained by frrst calculating the totalequiredper cent e-

duction factor D which is found by adding the individual reductions that areobtained by multiplying theconstant d, given in figure 9.13 by the number



2 0 40 60 8 0

Rated Short Circuit Current@A)

Figure 9.13 Reduction factor d, that is used for calculating derating factors for rapid

reclosing and extended duty ycles.

of operations in excessof the two which are required by the standard dutycle (CO + 15 sec + CO) and by the ratio of the time difference for each reclo-sure with a period that is less than 15 seconds. This is expressed mathemati-

cally by he relationship given below:R = 1 0 0 - D

(15 - l ) (15 - , )D = d , ( n - 2 ) + d 1 - + d l-5 15

+""_

where

R = derating factor

D= total reduction factor in percent

dl = multiplier for the calculationof reducing factor



t , = first time interval whichs less than 15 seconds

t2 = second time interval that s less than 15 seconds

n =to ta l number of contact openings

The following example should serve to illustrate the concept. Given a 45

kV, 20 kA circuit breaker that s going tobe applied for a (0+CO+ 10 sec +

CO + 1 minute+CO) duty cycle.The multiplier from figure 9.13 s equal to 3.3 at 20 kA and the reduc-tion factor Ds

(15-0) (15 - 10)D = 3 . 3 ~ ( 4 - 2 ) + 3 . 3 ~ - + 3.3 x

15 15

D = 6.6+3.3 + 1.1

D = l l % a n d R = 1 0 0 - l l = 8 9 % o r 0 . 8 9

The short circuit ratingof the circuit breaker s then reduced to20 x .S9 =17,800 amperes

9.8 Choosing BetweenVacuum or SF6

A fair comparative assessmentof vacuum or SF6 circuit breakers can onlybemade for medium voltage circuit breakers where both types of technologiescan be used interchangeably. The choice between vacuum,or SF, is mostly amatter of preference. The basic performance of both technologies is essen-tially the same because both are designed and tested to meet the same per-formance standards. There may be however some specific applications whereone technologymay be deemed more appropriate.

Vacuum circuit breakers have a very long and relatively maintenance freelife which represents a desirable attribute and a significant advantagefortechnology.Themaindisadvantageforvacuum nterrupters,on he otherhand, is their noticeable propensity for initiating overvoltages which may be

harmful to other equipment. Although for most applications there is no needfor concern, it s recommended that for applications involving the switching oftransformers and/or rotating machinery due considerationbe given to the useof surge arresters, and better yet to the use f surge suppressers which consistof a resistor and a capacitor in series. This components combination notreduces the frequency f the transient voltage but it also reduces the magnitof the voltage. Another advantage of this protection is that it serves to detunethe circuit and preventshe possibility of having a resonant circuit.

For applications where a large numbers operations under load, or faultconditions are requiredor where high rates of rise of recovery voltage are ex-



pected such as in the case o f reactor switching, vacuum circuit breakers maybe the betterchoice.But n heotherhand for applicationsoncapacitorswitching or the switching of transformers SFs circuit breakers may be advan-tageous. In either of the last two mentioned applications the addition of ca-pacitors or surge suppresser will serve to equalize the applicability of bothtechnologies. Another factor that may influence the choice and which at thetime of this writing is still unknown are the possible future environmental re-

strictions and liabilities that ay be imposed on the use of

REFERENCES

1. ANSIAEEE C37.24-1986 IEEE Guide for Evaluating the Effect of SolarRadiation on Outdoor Metal-Enclosed Switchgear.

2. ANSVIEEE C37.010-1979 IEEE Application Guide for AC High-VoltageCircuit Breakers Ratedn a Symmetrical Current Basis.

3. ANSIAEEE C37.04 Rating Structure for AC High-Voltage circuit Break-ers Ratedon a Symmetrical Current Basis.

4. ANSVIEEEC37.012ApplicationGuide for Capacitance Current Switch-ing for AC High-Voltage Circuit Breakers Rated on a Symmetrical Cur-

rent Basis.5. IEEE C37.015 IEEE Application Guide for Shunt Reactor Switch-

6. D. L. Swindler, Application of SF6 and Medium Voltage Circuit Break-ers, IEEE CH 3331-6/93/0000-0120, 1993.

7. V.D.Marco, S. Manganaro,G.Santostino, F. Comago,Performanceof Medium Voltage Circuit Breakers Using Different Quenching Media,IEEE Transactionson Power Delivery, VOL-PWRD-2, 1,Jan 1987.

8. J. H. Brunke, Application of Power Circuit Breakers for Capacitive andSmall Inductive Current Switching, IEEE Tutorial Course, Application fPower Circuit Breakers, 3-EH0-388-9-Pw 43-57, 1993.

9. S. H. Telander, M. R. Willheim, R. B. Stump, Surge limiters for VacuumCircuit Breaker Switchgear, IEEE Transactions on Power Delivery Vol.

10. A. N. Greenwood, D. R. Kurtz, J. C. Sofianek, A guide to the applica-tion o f Vacuum Circuit Breakers, IEEE Transactions PA & S Vol. PAS-90, 1971.

ing.

PWRD-2 NO. 1 Jan . 1987.



I O

SYNCHRONOUS SWITCHING AND CONDITIONMONITORING

10.0 Introduction

Synchronous switching and condition monitoring are two subjects that havegained a great dealf relevance not only becausef their potential or increas-ing reliability and for making a contribution to improve the overall powerquality of the electric systems, but also for economic reasons. These conceptscan be instrumental in minimizing the use of auxiliary components, such as

pre-insertion resistors, in reducing equipment wear and unnecessary mainte-nance and thus reducing the total cost of ownership throughout the full lifetime of the equipment.

Synchronous switching s not a recent, r a new idea and or at least the las t30 years the feasibility of synchronized switchinghas been studied, many con-cepts have been investigated and even some commercial equipmenthas beenbuilt and utilized. [1,2,

Condition monitoring is a relatively newer concept that has came about

primarily because of recent developments of electronic sensors and data ac-quisition equipment that have made this idea not only technically feasible butalso economically attractive. Condition monitoring is an essential componentfor synchronous switching simply becauselot depends on ow accurately theoperating characteristics of the circuit breaker can be controlled. It is wellknow that the operating characteristics be affected by extreme ambienttemperatures, and by other prevailing conditions suchs mechanism operatingenergy evels,controlvoltages,operatingfrequencyof the equipment, its

chronological age, and its maintenance history among others. Collecting in-formation about these variations would provide with a data source from whsuitable correction factors may be selected to compensate for those operatingdeviations which re critical for accurate synchronous operation.

10.1 Synchronous Switching

Opening or closing the contacts of a circuit breaker s normally done in a to-

tally random fashion and consequently, as it has been described before,sient current and voltage disturbances may appear n the electrical system. Atypical way for controlling transient behavior has been to add discretecomponents such as resistors, capacitors, reactors, surge arresters, r combina-



tions of the above to the terminals of the circuit breaker nevertheless,n manycases it would be possible to control these transients without the addition ofexternal components but by operating the circuit breaker in synchronism witheither the current r the voltage oscillations, depending uponhe switching op-eration at hand. This means that for example, the opening of the contactsshould occur at a current zero when interrupting short circuit currents, or thatthe closing of the contacts should take place at voltage zero when energizing

capacitor banks.Operations that can benefit the most by synchronized switch are those in-

volving the switching f unloaded transformers, capacitor banks, and reactors.Energizing transmission lines and openinghe circuit breaker to intermpt shortcircuit currentsare also good candidates or synchronous switching 4].

Ideally and to obtain the greatest benefits, synchronous switching shouledone using circuit breakers that are capablef independent pole operation.In-dependent pole operation is already a standard feature in circuit breakers that

are rated above 5 5 0 kV and it is also used, under special request, for applica-tions as low as 145 kV. However for those designs where all three poles areoperated in unison the implementation of controlled switching concepts willrequire the development of specially designed circuit breakers which are pro-vided with suitable methods for staggeringhe pole operating sequences.

select the operating characteristics f a circuit breaker which have themost direct impact for synchronized switching requires a clear definition andunderstanding of the cause and effect relationship that exists betweenhe me-

chanical operation of the circuit breaker and the behavior of the electrictem. In every case it is indispensable to closely analyze the mechanical andelectrical propertiesof the design, including contact velocity, contact openingand closing time, minimum arcing time for different interrupting duties andcurrent levels and cold gap voltage withstand capability; furthermore and inrelation to the operating times the effectsf control voltage fluctuations, ambi-ent temperature, tolerances of the mechanism’s stored energy and operatingwear must also e considered.

10.1.1 Mechanical Performance Considerations

Consistency in the making and breaking times of the circuit breaker is abso-lutelyessentialforsuccessful mplementation of all types of synchronousswitching. But considering the fact that a circuit breaker is a mechanical de-vice, an even though modem designsre highly reliable, they must stillbe im-proved to minimize the influenceof wear, aging, cold temperatures, and volt-age control sources. Of all these parameters the ones that have the greatest i

fluence are the ambient temperature, the levelf energy stored n the operatingmechanism and the control voltage level.



The effects wear and aging upon the equipment presentlyn service, at

least for now, are unpredictable. There s only scarce and incomplete data andfurthermore there s a need for data obtained from actual field operating expe-riences to fully evaluatehe influence of these conditions upon the life ofcir-cuit breaker. Another condition that is difficult to evaluate and which mayhave a significant influence upon the repeatability the operations is the ef-fects of long periods time when the breaker has not been operated, specialwhen during the idle periods the equipment has been subjected to very lowambient.temperatures.

TABLE 10.1

Typical Operating Characteristicsan EHV Puffer Circuit Breaker

TABLE 10.2

Typical Operating Characteristicsof a Vacuum Circuit Breaker



As an illustration only and for future reference, the operating charactof what may be considered, from a qualitative point f view, typical of a SF,puffer high power circuit breaker are tabulated below in table Similarvalues for vacuum interrupters are givenn table

In addition to he mechanical characteristics, shown in tables andthe following electric attributes representing the minimum arcing timesor dif-ferent typesof switching duties are given.

Puffer SF,

Short Circuit Current Interruption millisecondsCapacitive Bank Current Interruption millisecondsLow Inductive Current (Reactors) Interruption milliseconds

Vacuum

Shortircuiturrent Intem ption milliseconds

Capacitiveankurrentnterruption millisecondsLow InductiveCurrent(Reactors)Interruption 2 milliseconds

For most synchronous switching applicationshe aggregate of all the varia-tions in the operating times, corresponding to the worst condition, both in

closing and in opening, should not exceed a maximum of plus or onemillisecond. For example, observing the operating times that re shown in ta-

ble it is evident that the aggregate time of circuit breaker, as shown

below, far exceeds the maximum openingor closing time allowable deviation.

Maximum deviationn closing time

(min. volts) + temp) + force)= ms

Maximum deviation in opening time

(min. volts) + (min. temp) + force) = ms

It is quite evident that the greatest influence on the operating timessexerted by the control voltage. This parameter, in comparison toll the othersoffers better possibilities for enhancement and consequently it i s where themajor improvement wouldbe expected. Possible solutions are the use a regu-lated power supply, r capacitor discharge systems for the supply ofhe controlcircuits energizing the operating coils. In additionhe solenoid coils should eoptimized to reduce the operating time range.

far mention has only been madef the operating times without mention

of the operating speeds, but, these also are quite importantor they are relatedto the minimum arcing time during opening and to the prestrike time duringclosing. The product of the velocity and the above mentioned parameters de-termine the minimum contact gap requiredor the respective operation.



10.1.2 Contact Gap Voltage W ithstand

The contact gap voltage withstand capability is an arbitrary defmition that re-lates to the change in gap distance in relation to the instantaneous voltagewithstand capability across said gap during a normal closing operation.approximate value can e obtained for an interrupter, but first,o minimize thevariables, it is assumed that the interrupter has not been subjected to an electri

arc immediately preceding the measurement. The voltage withstand thus ob-tained is somewhat in the optimistic side, since during normal service, andmore so during a reclosing operationt will be reasonable to expect that arcinghas occurred. The effects of prior arcing must still be investigated, but it isexpected thatf there is a reduction in capabilities such reduction woulde lessthan 10%of the cold gap withstand.

For SF6 interrupters the contact gap withstands a function of the gas pres-sure andof the geometry of the contacts; higher operating pressures and lower

dielectric stresses in the field acrosshe contacts would produce higher contactgap withstand values. Reported values range from approximately kVper millimeter to 25 kV per millimeter. While for vacuum intermpters thecontact gap capability is primarily a functionof the electrode material andis in the range of 20 to kV per millimeter.

10.1.3 Synchronous Capacitance Switching

For capacitance switching, the primary concerns not as much the intermptionof capacitive currents because, due to the inherent characteristics of vacuumand SF6 circuit breakers the problems associated with restrikes, found withearlier technologies, have been greatly reduced and today, indeed restrikesrea very rare occurrence. On the other hand, failures are often reported whichare the direct result of inrush currents and overvoltages that have propagatedthemselves into lower voltage networks causing damage specially to electroequipment connected to the circuit. A comparison of the voltage transient ora non-synchronous operation is shown in figure 10.1 and in figure 10.2 thevoltage response of a synchronized closing s shown. As it can be seen in theillustrations hehigherfrequencycomponent of the voltage is practicallyeliminated when the contacts are closed at a nominal voltage zero condition.

Energizing a capacitor bank In order to completely eliminate the overvoltagesproduced by the closure of circuit breaker onto capacitor bankt is thatthere be a zero voltage difference across the contacts of the circuit breakw,M~U-

rally is not always possible simply because some deviationiom theoptimumopemting conditions is to be expected. have shown that the overvolt-

ages can be reduced to acceptable limits when the closing of the contactsoccur within one m illisecond either before or after the vo ltage zero point.The significance of this requirement is better appreciated when consid-



-TIM E m l l l s e c o m d s

Figure 10.1 Voltage corresponding to a non-synchronous closing a circuit breaker

into a capacitor bank.

25 30

T 1VIE

Figure 10.2 Joltage corresponding to a circuit breaker synchronous closing into a

capacitor bank.



0 1 2 3 4 5 6 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7

TIME milliseconds

Figure 10.3 Relation between system voltage and interrupter gap withstand capabil-ity.

ered in conjunction with the gap withstand capabilityof the contacts as illus-

trated in figure 10.3 where the absolute valuef a sinusoidal voltage s plottedin conjunction withhe slope of an assumed gap withstand characteristic.As itcan be seen in the figure the point where the prestrike takeslace correspondsto the intersection of the two curves. In the figure to better illustrate the con-cept the several different times for the intersection of the gap withstandme shown. Ideally the rate of change of the gap withstand voltage shoulde atleast 10% higher than theate of change the system voltage to assure propercoordination the two rates..

further investigate the conceptet consider the circuit breaker whosecharacteristics were shown in table 10.1, let us furthermore assume thathe gapwithstand capability of this circuit breaker is 10 kV per millimeter. Using thegiven closing speedof 4.3 meters per second, shown in table 10.1, it is foundthat the corresponding ate of change withstand capability s 43 kV per mi-

crosecond.Assuming that the circuit breaker under consideration s intended for ca-

pacitance switching duty ona kV ungrounded system which. Since, as we

know, an ungrounded system representshe worst case n terms of voltageeakbecause this peak, for the last phase to close, s equal to 1.5 times the peak of

the rated line to line system voltage.



120

100

81)

&

5

>

40

20

0 5 105 20

TIME ms.

Figure 10.4 Rates change of gap withstand capability for the circuit breaker thegiven example and the rates change the system voltage for grounded and un-

grounded applications,

The maximum rateof change of voltage with respect o time at the nstant

of voltage corresponding to the above phases

For a grounded application

dE J2-=-

dt A E = 0.82 72 377 = 22.2 kV per microsecond

For an ungrounded application

dE J5

dt J5= 1.5 - E o = 1.225 72 377 = 33.2 kV per microsecond

Comparing the two rates of change, the ones Corresponding to the system

voltage against the one related to the rate of changef the gap capability, as itshown in figure 10.4, it appears that the interrupter being considered in the eample wouldbe adequate for this application.



Table 10.3

Maximum rate f change of system Voltagek V / u I

Ratedoltageroundedank I Ungroundedank

72 I 22 33

121

254 169 550

16812 362

114 7645

68 45 145567

However, in the event that the example interrupter is considered for an

application at 145 kV, y following he same procedure s before and usinghe

tabulated values shown in table 10.3 for the customary preferred rated systemvoltages it is evident that this interrupter is inadequate for the application.Nonetheless, if the gap withstand capability s increased to 25 kV er millime-ter, assuming the same .3 meters per second speed, then he rate of change ofthewithstandcapabilitybecomesapproximately108kV per microsecond.

suggests that t would be possible to consider the use ofhis intenupter forall grounded capacitor bank applications p to 245 kV and if two of these in-terrupters are connected in seriest would also be possible to meet the 550kV

application requirements. For ungrounded banks a single interrupter is goodonly up to 145 kV, and even with two interrupters in series it would not bepossible to meet the 550V rating.

increase the rateof change of the withstand capability in an SF, circuitbreaker any of the following three options either individually or in combina-tion may be applied.

1. increase the gas operating pressurereduce the electric field stress in the contact region

3. increase the closing velocity f the circuit breaker contacts

Increasing the operating pressure, in many cases, is not a viable solu-tion becauseof the possibility of SF, liquefaction at low temperatures. Never-theless, it should be kept inmind that at voltages above 362 kV the systemsregrounded almost without exception and therefore our imaginary non-idealcir-

cuit breakermay already be acceptable.For vacuum circuit breakers, and since presently they are used almost ex-

clusively at system voltages in the rangef 15 kV to38 kV, the minimum rateof change of the gap withstand does not present any problem. The minimumrate of gap dielectric may be assumed to be in the neighborhood of 30 kV per



millisecond, while the maximum rate of change of voltage for a kV un-grounded bank s 17.5 kVper m illisecond.

The precise points wherehe contacts of each pole m ust close depend upthe systemconnections.When hecapacitorsand the systemneutralsaregrounded then the optimum point to close the contacts is each pole independ-ently at the voltage zerof the corresponding phase.

When the capacitors are connected in an ungrounded system t is possible

to close the first pole at random, since there will be no current flow with onlone pole closed. The second pole and the third pole must then close at theirrespective voltage zero. Another alternative would be to close two poles si-multaneously at a voltage zero and then close the third ole at its correspond-ing voltage zero.

When we of voltage zero what s mean is that the voltage differencea m s s the contacts of the circuit breaker s zero. Because of the possibility oftrapped charges in a capacitor it would then be necessary to monitor the ac

voltage from the source side as wellas the dc voltage in the capacitor at theload sideof the circuit breaker.

Presently, there are a numberof control devices available. As an exampleof one such device, which is marketed under the name ACCUSWITCH, itsblock diagram s shown in figure 10.5.

De-energizingacapacitorbank. Capacitive currents generally require veryshort arcing times which means that the actual contact gaps very short and n

some cases, when he magnitude of the recovery voltage exceedshe dielectriccapability of this small gap,t leads to restrikes.

It was indicated earlier that typical minimum arcing timesor capacitanceswitching with SF6 circuit breakers is about 2.5 milliseconds and 1.5 millisec-ond for vacuum interrupters. Since it is not indispensable that the opening ofthe contacts coincides with he minimum arcing time but rather that t shouldbe longer than that which s considered minimum, a satisfactory choice woulbe to part the contacts at a point thats at least 4 milliseconds prior to the

rent zero. Consequently the controls for this type of application need not bethat sophisticated again all that is needed is that the contacts separateciently aheadof the current zero.

it was said before, with the advent of the new technologiesf high volt-age circuit breakers theres basically no need for synchronous opening of hecapacitor banks and that this mode of operation should only be considered invery special occasions when it is known that there is a real possibility of re-striking.



nS

b



10.1.4 Synchronous Reactor Switching

For reactor switching operations the basic needsre the opposite of thosecon-sidered to be desirable for capacitance switching, that is closing the circuit isnot as important s is opening.

Closing Control. Typically, most high voltage circuit breakers will pre-strikeduring a closing operation ands a result of pre-strike an overvoltage that

generally is less than 1.5 per unit is produced. Since overvoltage, by nomeans, should be considered to be excessive there is no pressing need to con-trol its magnitude and consequently a controlled,r synchronized closing,romthe point of view of switching overvoltages is considered to be unnecessary.Furthermore if the closing is synchronized with a voltage zero conditionwould result in a high asymmetric current which may develop excessive me-chanical stresses within the of the reactor being switchedon. Addition-ally, if in a grounded circuit the closingf the contacts takes place t a voltage

zero it is possible that an excessive zero-sequence current may flow thusais-ing the possibility f the zero-sequence relays being activated.

AT ime axis in milliseconds

Figure 10.6 Representation of voltage appearing across the circuit breaker contacts

when reclosing intoa shunt compensated line.



If synchronous closing is to be considered, it would be prefemble to close thecontacts at voltage, which the relatively since there s anatural tendency for the contacts to doust that, plus the fact that nearhe peakofthe voltage its m e of change is basically zero and therefore there is room for alarger tolerance in the allowable variation of the closing time.

One simple way to accomplish controlled closed would be to reduce theclosing speed of the contacts that the rate of change of the gap withstand

becomes significantly lower than the maximum rate of change of the systemvoltage.

A unique condition that is worth mentioning because of the significantbenefits that can e attained from synchronous closing,s when rapid reclosingof a shunt reactor compensated lines required. Reclosing implies hat currentinterruption has just taken place and since following interruption a trappedcharge may be left on he unloaded line. In case the voltage across he cir-cuit breaker willhow a significant beat, s illustrated in figure 10.6, due tohe

frequency difference between theine and the load sides. At the source side ofthe circuit breaker he voltage oscillates withhe power frequency while t theload side the frequencyf the oscillation may be as low as one-half that of thepower system frequency (60 Hz.) or as high as to approach the power fre-quency, it only depends upon the degreef compensation; the higher he com-pensation, the lower the frequency.Since synchronization should be made at a beat minimum, where the voltageacross the contacts s relatively small, it is evident that the degree of complex-

ity for detecting zero voltage condition across the contacts has greatly in-creased, thus making this task extremely difficult.This is further complicatedby the fact that he variable beat frequency creates a high degreef uncertaintyfor predicting the voltage zero acrosshe circuit breaker contacts.

Opening Control. Synchronized opening of the contacts in an application in-volving the switching of reactors shouldbe considered for the purpose of re-ducing overvoltages that maye generated as the result of current choppingor

reignitions that may occur during a normal opening operation. Onebenefit ofsynchronous opening of reactor circuits, specially those that use reactors forshunt compensation, is that it substantially reduces switching surge overvolt-ages.

Synchronous control for opening reactor circuits is not difficult to achievesince it is only necessary to separatehe contacts at a time which s larger thanthe minimum arcing time required for that operation by that particular circuitbreakerdesign.What is important is that the contact gap be sufficient to

withstand the recovery voltage and thathe contact separationbe close enoughto the minimum arcing time to reduce the possibility of current chopping.



The likelihood of reignitions is greatly reduced by synchronized switchingwhen the three poles of the circuit breakerre gang operated. If the poles are operated independently of each other and each allowing an arcing time of at leastmil-

liseconds then any probability of reignitions is virtually eliminated.It should be noticed that the synchronizing requirements for type of

applications are dependent primarily upon the characteristics of the circuitbreaker, rather than the characteristics or the connections, grounded or un-

grounded, of he system.

10.1.5 Synchronous Transformer Switching

Basically speaking, the switching ofan unloaded transformer is no Werentswitching a reactor, thats the voltages and currents involved in opening and cling the circuit of the transformer generally have the same characteristics of thoseproduced by the switching of reactors. However, for application the most riti-

cal variable is the transformer's inrush current [S] which, in some occasions

reach magnitudes that approach those of the short circuit current. The magnituthe inrush current depends on the transformer’s impedance,n the magnetic ham-teristics of the core of the transformer and on the of its magnetic flux rem-

nants at the instant when the circuit is energized. The severity of the energizingprocess is greater for transformers that havehigh remnants, for those that arcompletely demagnetized. It follows then that forfullsynchronization it is neces-

to detect the remnants level prior to the energizing of the transformer,r alter-nately that all openings of the transformer circuit be made synchronouslyo that the

remnants conditions controlled and be well defmed for the next closingp-eration. If remnants is not considered to be a problem then closing may be donesatisfactorily at voltage peak with a tolerance ofsmuch asX2 milliseconds.

10.1.6 Synchronous Short Circuit Current Switching

Synchronized switching of short circuit currentss a desirable feature from thepoint of viewof reducing contact erosion which translatesn extending the lifeof the circuit breaker. However the benefits need to be weighted against the

possible cost andhe complexity of he task.

Synchronousclosing. The aim of synchronous closing would be to reducecontact erosion by reducinghe arcing time during closing, whichs due to pre-strikes across the contacts. The benefits that maybe achieved by synchronousclosing must be kept into perspective since contact erosion during a closingoperation is significantly ess handuring nterruption.Unless the rate ofchange of the gap dielectric capability s extremely slow the pre-arcing time isbound to be considerably shorter than the intermpting arcing time. Further-more it should be considered that due to the low instantaneous values of cur-



rent at closing the energy input will be significantly lower than that which isseen during interruption.

The optimum switching angle for reducing,r eliminating prestrikes would eat voltage nevertheless, may present a problem because thecurrent peak is reached under these conditions due to thewhich is produced when the current flow, in an inductive circuit, is initiated at avoltage As a consequence of the high current peak the elem -mechanical-

stresses imposed on the circuit breaker the highest. translates into higheroutput energy requirements for the operating mechanism and in general larger

for the circuit breaker. alternate possibility for sem i- syncW smwould be to choose an optimum switching angle, one whichn most caseswould bebetween 30 and 45 electrical degrees.

Synchronous interruption. Theoretically a synchronous interrupter is one thatchanges instantaneously from a perfect conductor to a perfect isolator.

characteristic constitutes a physical impossibilityor a mechanical device suchas a conventional circuit breaker since, a finite gap would have to be devel-oped in essentially zero time.

Nevertheless, it is possible to design a quasi-synchronous interrupter, onethat separates its contacts consistently at a predetermined time just ahead ofcurrent zero. However, in order to achieve this a high operating contact speedis needed that a contact gap, that is large enough to withstand the transientrecovery voltage can be attained during he very short arcing period available.

This approach has been successfully demonstrated in a number of experimentadevices, for at least the last 30 years. A prototype design ofa synchronous cir-cuit breaker was installed and remainedn service for over 15 years. A photo-graph of this prototype breakers shown in figure 10.7.

In spite of the work thatas been done and the knowledge thatas beenthere are still a number of practical problems associated with concept. re-duce the mechanical stresses on the interrupter, due to the high operating forceswhich are required to accelerate the moving contactst would be desirable to reduce

their mass but, these lighter contacts, may no be able to the rated continuoscurrent. regain capability a parallelset of primary contacts mayneed to beprovided, butn doing the operating scheme of the circuit breaker becomes mcomplex. If two sets of contacts, each moving at aH eren t speed, then acomplicated mechanical scheme, orwo independent operating mechanisms wouldhave to be provided for each pole assembly. A further complication to a

lesser extent with puffer circuit breakers, but more efiitely so with selfpres-suregenerating breakers that utilize the arc energy to generate the interruptingres-

sure,

from the fact that by minimizing the arcing time there may not be sufficienergy and stroke for the former and time for the later to generate the requirrupting pressure.



Figure 10.7 ITE synchronous interrupting circuit breaker nstalled by American

Electric Power.

The primary benefit to be derived from synchronous operation s a reduc-tion in size and a decrease in the erosion of the arcing contacts. A relativecomparison of the arc energy input for a non synchronous circuit breaker hav

ing a 12 millisecond arcing time and a synchronous interrupter with only a 1millisecond arcing times illustrated in figures 10.8 and 10.9 respectively.

It is unlikely that synchronous interruption will produce a noticeable im-provement in the in tm p t in g capacity becauseas it was shown in chapter therecoverycapability of an SF6 intenupter is directly related o he rate ofchange of current at he instant of current interruption rather than tohe magni-tude of the current peak. With vacuum interrupters however some improve-ment may be expected because he duration of the coalescent arc mode maye

significantly reduced and furthermore, depending uponhe total magnitude ofthe current being interruptedhe arc may remain in ts diffuse mode or the fullduration of the arcing time period.



Figure 10.8 Arcing time for non-synchronous interruption.

Figure 10.9 Maximum arcing time for a synchronous interruption.



For circuit breakers hat have a characteristically ong arcing ime theopening speed tends to reach levels that are considered o be impractical.get a feel for the opening velocities that are required consideror example a72kV circuit breaker where its contacts move at 3 meters per second and theminimum arcing time s 10 milliseconds, thus the minimum contact gap caneassumed as being approximately 30 millimeters. If arcing is to be limited toonly 1 millisecond then the required opening speed is 30 meters per second.

Considering the mass of the contacts and the linkages involved attainingspeed would be a very difficult task. Since circuit breaker design almost in-variably entail compromises, some reduction on the contact erosion may besacrificed to gain some reduction in the operating speed and the likely me-chanical wear. The decision must be based upon sound evaluation of thetechnical and economic advantages of the concept talking into considerationthe frequency of operations under fault conditions.

10.2 Condition Monitoringof Circuit Breakers

it has been said before,circuitbreakers constitute an important and criticacomponent of the electric system, they re the last line of defense and Conse-quently proper and reliable operations paramount to the quality of power be-ing delivered, to he promotion of customer satisfaction and most f all to thesafety and integrity f the system

sustain the confidence level on critical pieceof equipment compre-

hensivemaintenanceprogramshavebeenestablished.Thesemaintenanceprograms follow established standard guidelines and the recommendations ofthe manufacturer which generally are based on their operating experience.This practice may not only be inefficient but, it is also costly because of thedown time required to perform these procedures. Additionally, it is not un-common that problems develop following maintenance of otherwise satisfac-torily performing equipment. A more logical approachmay be to continuallyevaluate the conditionof those components that through experience have been

identified as being the most likely to fail and those whose failure could pro-voke a severe damage that would disrupt the service.

Historically, mostof the of circuit breaker failures that have been obserin the field canbe attributed to mechanical problems and difficulties related tthe auxiliary control circuits. A number of studies, suchas those made byCI-GRE lo ] provide with an excellent insight into the failure statistics of thecomponentsof a circuit breaker. The report indicates, as shown in figure10.10, that 70% of the major failures in circuit breakers are of a mechanical

nature, 19% are related to auxiliary and control circuits and 11% can be at-tributed to electric problems involving the interrupters or the current paththe circuit breaker.



ElectricalComponents

11%

Controls

19%

Mechanical

Components70%

Figure 10.10 Typesof circuit breaker failures by major components.

A further breakdownof the problems shows, (figure 10.1), hat in the me-chanical failure category approximately 16% nvolves compressors, pumps,motors, etc.,7% involve the energy storage elements, 10%s caused by controlcomponents, 7%are originated by actuators, shock absorbers, etc. and re-sult from failuresf connecting rods and components ofhe power

In the group of electrical controls and auxiliary circuits,ailure to respondto the trip and close commands accountsor 6% of he problems, 5% are due tofaulty operation of auxiliary switches, 6% are caused by contactors, heatersetc. and 9% are attributed to deficiencies ofhe gas density monitors.

For those problems whichre judged to be of electrical nature,he majorityof them; are due to failureof the insulation with respect to ground, 12%are due to the interrupters themselves and 2% are dueo auxiliary interruptersor opening resistorsor grading capacitors.

The just mentioned be used as a guideline for the selection of thoscomponents hatshouldbemonitored.Although hemostdesirableoptionwould be to develop a system that constantly monitors critical components an



Closing &

Opening Others

Resistors 4% Motors

C ~ t a c t O r s 5% Trip & CloseHeaters Coils

6%

Figure 10.11 Percent of reported failures components.

which is able to detect any deterioration that ay occur over ime and to pre-dict, in a pro-active way, impending failures of mechanical components.task however, has proven tobe rather elusive. A number of detection systemshave been investigated, including the use of acoustic signatures [11,12,but at least at this time these systems have not yet been translated into aiableproduct and they till remain mostly in a laboratory environment. Furthermorits reliability may be questioned primarily because of its complexity and itshigh sensitivity, which makes it vulnerableo noise and o the influence of ex-

traneous sources. Simpler schemesmay provide adequate protection, but natu-rally, a final choice should be based on an evaluation of the benefits againstthe complexity and the difficulty of implementing the specifically requiredmonitoring function.

The information thats gathered by he monitoring system does notave tobe limited exclusively to evaluate the condition of the circuit breaker, but italso may be used to enhance the accuracy of the controls for synchronous op-eration, if such operating option is available. It is entirely possible touse the

data to adjust he initiation of the closing or opening operation as to com-pensate for variations in the making or breaking times hat are due to the influ-ence of the parameters thatre being monitored.



10.2.1 Choice of Monitored Param eters

There are a significant numberof parameters that can be chosen for monitor-ing, there are as well a variety of methods each having varying degrees ofcomplexity for executing the monitoring function. The optimum system wouldbe one that selects he most basic and important functions and thus m inimizesthe number of parameters thatare to be monitored and yett maximizes the ef-fectiveness of the evaluationof the system thats being monitored.

In addition to optimizing the numberf monitored parameters, the methodsused to do the monitoring should be kept as simple and straight forward as

possible. It will be desirable, if not essential, from he point of view of avail-ability, cost and operating experience, that commercially available transducersthat are used in any related industry shoulde given preference and used whereat all feasible.

10.2.1. Mechanical parameters

The most likely parameters to be monitored because of their significance andthe simplicity of the monitoring scheme are given below. The list of likelycandidates for monitoring and he possible methods hat can be used are givenonly as a suggestion. Other parameters may e deemed to be as mportant andtherefore they should be added to the list while others may be disregarded.The methodology may vary to suit the conditions of the application. What isimportant is to be able to have an ndication, or warning f something ischanging in the circuit breaker.

1.

2.

3.4.

5.

6.

7.

8.

charging motorscontact travel distance and velocitypoint of contact separation and contact touchspace heaters conditiontrip coils and close coilsmechanism stored energybreaker number of operations

ambient temperatureCharging motors. Encompassed under denomination are all motors thatare used for driving a gas compressor, a hydraulic pump,r for compressing aset of operating springs. From available statistics, motors, compressors andpumps exhibit the biggest sharef failures.

A convenient way to monitor the condition of motor would be to measure thestarting and the running currents. These current valueshen be used o calculatethe torque of the motor, torque value then be to judge the conditionfthe equipment or components that being driven by the motor. For example,n-



usual increases in torquemay indicate added fTiction which maye due to deterio-rating bearings or galling of parts.

If the running time and the frequencyf operation of the motor is compared

to a baseline data established under standard operation it will be possible todetect leaks in a pneumaticr hydraulic system.

Contact travel and veloc ity. This could easily be considered as the most im-

portant function being monitored. It provides dynamic information about theoperating components of the circuit breakers a whole including not only mechanical links but also the interrupter contacts. The information that beextracted from these measurements is always extremely valuable for judgingthe overall statusof the circuit breaker and fortunately is probably one ofthe simplest and easiest function to monitor.

From the measured travel characteristics by comparing the new data to a

base line signature for the specific circuit breakert should be possible to infer

not only deteriorationof linkages, but increased friction that could mean lackproper lubrication and or deteriorationf bearings for example.For puffer circuit breakers the travel characteristics, when usedn conjunc-

tion with the magnitude of the short circuit current being interrupted, wouldserve to indicate the degreef ablation of the interrupter’s nozzle.

The contact, or breaker travel measurement can e easily consummated bymonitoring the displacement or the rotation of he output shaftof the operatingmechanism. The closingor opening velocities then cane obtained by finding

the derivative with respect to time of the displacement measurement. Thismathematical manipulation most likely would e done electronically with theassistance of a central processing unit where all the signals woulde collectedfor evaluation and data storage.

The displacement measurement can be made using either contact o r non-contact transducers. Sliding resistors, linear or rotating resistor potentiomete

travel recorders, etc. are considered to be contact type transducers be-cause there is an actual, physical connection between the transducer and the

component being measured. Non-contact transducers re those such as opticalmotion sensors, proximity sensors, LTV sensors, etc. that do not require aphysical connection between the sensor andhe moving part. In all cases it ishighly desirable that non-contact transducers be used to minimize errors thatmay be caused by the inherent inertia of the moving parts of a mechanicalcontact transducer.

It has been mentionedhat it may be possible to comparehe monitored in-stantaneous velocity to a predefined velocity for a particular circuit breaker

operating under different setsof conditions, such as short circuit interruption,reduced energy from the mechanism, reduced ambient temperatures, etc. The



evaluation of comparison process would then e the criteria that s used tojudge the condition of the circuit breaker.

Contact make and contact break. Contact break and contact make indicationscan be obtained either by direct, or indirect methods. If a direct indication sdesired it can be obtained, when the measurement is made under load condi-tions by monitoring the voltage acrosshe contacts. Additionally for a closing

operation the measurement of the initiation of current flow provides an ap-proximate indication of contact touch. The current indication can be used in

conjunction with the no load travel measurement to compensate for normalprestriking.

Forno oadoperations, or when ndirectmeasurements are made, themethods that are used to determine contact displacement are applicable. Themajor draw back of approach is that it fails to take into account thechanges in the makingr breaking of the contacts that may occur s a result of

possible contact erosion.Space heaters. Their function s simple and yet heir failure may cause signifi-cant problems. Monitoring the integrity f the heater elements s a rather triv-ial task that can be done by simply circulating continuously a very small ur-rent. Another alternative is to use thermostats that are strategically located nclose proximity to he heaters, one disadvantageof method is that heatersare not energized all the time but only when the ambient temperature dropsbelow a certain level and consequently a logic circuit that relates ambient teperature heater temperature shouldbe provided but still it does not providecontinuous monitoring ofhe continuity of the heater element itself.

Trip and close coils. Monitoring of these components is a relatively simpletask, although caution muste used to prevent the creationof parasitic currentpaths that could createissperations and problems with control relays.

In its most simple version he monitoring system would only require eithera continuous or an intermittent high frequency signal o determine the electri-

cal continuityof the coils.Mechanism spring charge. The function referred to s spring charge s in real-ity a measurement of the kinetic energy that is stored n the operating mecha-nism. The measurement can be made by measuring the spring compression,n

terms of its change of length, or the gas pressure of pneumatic accumulatorsused in conjunction with hydraulic mechanisms. The same type of measure-ments are also applicable for springr pneumatic mechanisms.

The displacement measurements made using he same typeof instrumen-tation as that which is used for measuring he travel of the contacts of the cir-

cuit breaker. In general it is only necessary to detect he extreme limitsof the



displacement which corresponds to the limits for the required spring deflectithat covers the specified rangef operating forces.

Operations counter. This is a seemingly elementary piece of information andyet it is very significant specially on those breakers wherehe operating char-acteristics show variations that are related to the accumulated number of op-erations. Thismeasurement is not something new and in most circuit breakers

it has been routinely made if nothing else to keep a tally in order to performmaintenance at the next recommended maintenance interval.

Ambient temperatureMonitoring the ambient temperature may also e consid-ered as an elementary or trivial piece of information; but information isneeded to detect deviations from he historical operating characteristics of hecircuit breaker under similar conditions to those being monitored. The meas-urement can also be useful to compensate for variations in the operating timein synchronous switching applications.

10.2.1.2Electricalparameters

Dielectric failures and interrupter failures represent a high percentage of helisted reasons for circuit breaker problems. Although, many of these failureswould take place without any prior warning, there are some cases where itwould be possible to anticipate n impending failure based on some conditionwhich are generally wellknown and predictable s is the case with high levelsof corona, high leakage currents, high moisture content and low insulating gas

density . While some of these parameters can be monitored with only reason-able efforts, there are others which are difficult to monitor while the circuitbreaker is energized and in service. hat follows is again only a suggested listof significant electrical components that coulde monitored.

Contact erosion and interrupter wear . Monitoring contact erosion and inter-rupter wear has a strong, direct influence upon the required maintenance fre-quency therefore, it is not only desirable, but beneficial to accurately evaluate

the condition of the interrupters rather than to rely on the presently usedmethod of simply adding the interrupted currents until the estimated accumu-lated duty that is given by applicable standards, r by the m anufacturers rec-ommendations, is reached.

Measurements of contact erosion, or interrupter wear can notbe made di-

rectly, but it can be done conveniently by indirect methods using measure-ments of current and arcing time. The interrupted current can me measwedusing conventional instrumentation, such as current transformers, which are

generally available in circuit breaker as an standard component. The arcingtime, depending in the desired degree of sophistication, can be determined byoptical detection of the arc, by measurement of the arc voltage, or simply by



estimating the point of contact separation using the information given by hecontact travel transducer andhe duration of current flow from time until tis interrupted.

The product of the current and the elapsed time from contact separationocurrent extinction gives a parameter to which the interrupter wear canbe re-lated. It is assumed that sufficient ata has been collected during developmenttests relating contact erosion and nozzle ablation to ampere-secondsf arcing,

and therefore by keeping trackf the accumulated ampere seconds an adequateappraisal of the interrupter condition be made.

Gas density. For SF, circuit breakers gas density rather than gas pressures theparameter that should be monitored. To do this, it is possible to use commer-cially available temperature compensated pressure switches or alternatively,the density may be determined by electronically processing separate pressureand temperature signals. These signals cane combined by an algorithm rep-

resenting the well known equation of state for the gas. Any deviation from aconstant density line will indicate that there is a gas leak in the system andunless a massive catastrophic failure occurs, slow leaks can be a i m e d andprotective actions can thene implemented.

The selection of the corresponding constant density line depends on theinitial filling conditionsf the circuit breaker.

Gm moisture. Dielectricfailuresconstituteoneof the highestpercentagemode of failure. A possible cause or these failures s a high moisture contentin the insulating gas, w hich may lead to tracking alonghe surfaces of he insu-lating materials used in the interrupter assembly. prevent moisture relatedproblems monitoring he moisture contentof the gas should e a high priority.

Checking for moisture content n a gas is an easiertask when done duringroutine maintenance, when the circuit breaker is out of service. In most cases

routine inspection is all that should be needed because it is a standardpractice to install insidef the interrupter moisture absorbing materials suchsactivated alumina. Furthermore, for circuit breaker which has been properly

dried and evacuated prior to filling, there is no reason for any significantamount of moisture to migrate inside of the interrupter unless there is a gasleak and the interrupter is then incorrectly refilled. Nevertheless, if it is de-sired to monitor the moisture content while the circuit breakers in service, itis possible to do using commercially available moisture monitoring instru-mentation hatcan be readilyconnected o hegassystemofanycircuitbreaker.

Partial discharges. The type and rate of deterioration of insulation and thetime required or a breakdown f the insulation depends onhe thickness of hematerial, on ts chemical and thermal stability, on the applied stress andn the



ambient temperature and humidity. Ultimate failure is usually caused by cu-

mulative heatingof the discharges,or by tracking across the material surfaces.Although there is no absolute basis for predicting the life of materials in thepresence of electrical discharges, it has been demonstrated that insulationustcan not be reliably operated above the discharge inception voltage. It is thendesirable to detect the onset,s well as the magnitude the discharges.

Measurement these two characteristics is rather difficult even under

carefully controlled conditions, such as in a test laboratory. The problem liesin the low magnitude of the signals involved, since the discharge pulses maybe as low as one microvolt[lS, 151.

An alternate solution wouldbe to use acoustic or optical detectors. Thesemethods can provide a reasonably accurate indication but mostly only as a

qualitative evaluation f the discharges t or near the surfacef the insulation.

Contact temperature. The temperature at or near theain contacts be a goodindicator for a number of possible potential problems with the circuit breaker.Large changes n contact temperature maybe due to bmken contact fingers, exces-sive burning of the contacts, material degtadation, oxide formation, weakcontact springs, or even n improperly or not fully closed circuit breaker.

The temperature of the contacts, or the conducting parts be measuredusing optical methods, or else t can be approximated by measuring the tem-perature of the surrounding gas, of the ambient temperature and of the con-tinuous current that is being carried. Knowing the normal temperature rise ofthe breaker the corresponding temperaturet these particular conditions canecalculated. The results then can be compared with what is expected fromcircuit breaker based on previously obtained development data.

10.2.2 Monitored Signals Management

Considering the operating conditions under whichhe monitoring is done andthe actions taken s a resultof the information being obtainedt would be pos-sible to characterize these signalss being proactiveor reactive.

Roactive signalswould be thosewhichdonotdependon the circuit

breaker being operated before the condition a certain parameters could bedetermined. On the other hand reactive signals are those that can only bemeasured during a dynamic condition such as the opening or closing of thebreaker contacts.

Proactive signals truly fulfill the intent of condition monitoring by givingthe opportunity to take action either by sending an alarmr by preventing theoperation of the circuit breaker when one f the components being monitoredhas failed. One thing that should be avoided is sending information continu-

ously while all the conditions are normal. Decisions should be made at thebreaker level s to what action must e taken.



'7IGH IMOIST.

HIGH

i I IBlock

Alarm

Figure 10.12Logic diagram for actiontobe based on proactive signals.



Contact . ,

Travel

Original

Compare OK

W d t to

U- Compare&c toL d s ,

Max. OK

NO

Compareto Max. -motor

motor

Measure-

omparerunning

OKormalime

-o

NO

Running

frequencyStartkr -

Figure 10.13Logic diagram for actiono be takenbased on reactive signals.

To accomplish his objective it is possible o set up a switching ogicscheme such as those shown in figures 10.11 and 10.12 or the proactive andthe reactive signals respectively.

In the case a reactive signal,f deviation from he maximum acceptablelimits is observed, the action most likely toe taken wouldbe to sendan alarmaccompanied by blocking f the next operation. The actual schemes will vary,depending upon the established local operating philosophies and consequentla number widely diversified schemes may be found. As the technologyfurther develops it may be possible to use neural networks and logic tomake more complex decisions once the networks have been properly trainedand sufficient operational data is available that justifies the program med ac-tions.



REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

J. Beehler, L. D.McConnell,ANewSynchronousCircuitBreaker forMachineProtection, EEETransactionsT-PA&S73:668-672,March-April 1973.L. D. McConnell, R. D. Garzon, The Development of a New Synchro-nous Circuit Breaker, IEEE Transactions T-PA&S 73: 673-681, March-

April 1973.R. D. Garzon, Synchronous Transmission Circuit Breaker Development,Finaleport,nergyesearchndevelopmentdministration,

Controlled Switching A State Of The Art Survey Part 11, Electra 164, TF13.00.1: 39-69, Feb. 96.Aftab Khan, D. S. Johnson, J. R. Meyer, K. B. Hapke, Development ofa new synchronous closing circuit breakeror shunt capacitor bank energi-

zation, Sixty-First Annual International Conference of Doble Clients, Pa-per 5E, March 1994.J.Roguski,Experimentalnvestigationof the dielectricecoverystrengthbetween the separatingcontacts of vacuumcircuitbreakers.IEEE Trans, Power Delivery . Vol2: 1063-1069, 1989.B..Ware,.G.Reckleff,,G.Mauthe,G.chett,ynchronousSwitching of Power Systems, CIGRE Session 1990, Report No. 13-205.G.Moraw,W.Richter,H.Hutegger,J.Wogerbauer,PointonWave

ControlledSwitching ofHighVoltageCircuit-Breakers,CIGRE13-02,1988 Session.H. Karrenbauer, C. Neumann, A Concept for Self-checking and Auto-control of HV Circuit Breakers and its Impact on Maintenance and Reli-ability, CIGRE Symposium, Berlin, Paper 120-03, April 1993.

CONS/2155-1, August 19 76 .

10. Final Report of the Second International Inquire on High Voltage CircuitBreakers Failures and Defects n Service, Working Group06ofstudycommittee 13 (Reliability of High Voltage Circuit Breakers), June 1994.

11. L. E. Lundgaard, M. Runde, B. Skyberg, Acoustic Diagnosis of Gas In-sulated Substations; a Theoretical and Experimental Basis, IEEE Trans.Power Delivery: 1751-1759, Oct. 1990.

12. V.Dem-janenko,H.Naidu,A. M. K. Tangri, R. A. Valtin, D.P.Hess, S. Y. Park, M. Soumekh, A. Soom,D.M.Benenson, S. E.

Wright, A Noninvasive Diagnostic Instrumentor Power Circuit Breakers,IEEE Trans. Power Delivery: 656-663, April 1992.

13.D. P.Hess, S. Y . Park, M. K. Tangri, S . G.Vougioukas,A. Soom,

V.Demjanenko,R. S. Acharya,D.M.Benenson, S. E. Wright,Non-invasiveConditionAssessmentandEventTiming for PowerCircuitBreakers, IEEE Trans. Power Delivery: 353-360, January 1992.



14. T. Tanaka, M. Nagao, T. Okamoto, M. Ieda, H. m e r , W. Kodoll,

15. H. Mason,DischargeDetectionandMeasurements, hoc. IEE, Vol

16. Application o f Diagnostic Techniques for High Voltage Circuit Breakers,

Electra No. 164, TF 15.06.01:85-101, February 1996.

112, No. 7, July 1965.

CIGRE 13.101, 1992.



APPENDIX: CONVERSION TABLES

N

3

2

.2W

W

351



ac decay,ac transient,Accredited Standards Committee

acetylene,acoustic signatures,activated alumina,adjacent poles,AEG,

air blast circuit breakers,air break,air consumption,air magnetic circuit breakers,

(ASC

arc chute,arc horns,arc plates,arc resistance,interrupting capability,

air,oil, SF, interrupters, current

alternating current component,altitude derating,aluminum oxide,ambient temperature,American Institute f Electrical

Engineers, (AIEE),American National Standards

Institute,

amplified arc voltage,

ANSI/ lEEE/NEMA,

chopping,

ANSI,

arc roots,arcs

anode drop,anode voltage,cathode drop.,

cathode region,cathode voltage,characteristics of, conducting plasma,constricted, 9diffuse, 9,

discharge,high pressure,5hysteresis,interruption of,low pressure (vacuum),8resistance,roots,time constant,time lag,

areas ofcontact,asymmetrical,asymmetry factor,S,asymmetry three phase currents,

axial field,

axial insulating nozzle,Ayrton,

back to back capacitor banks,back-up circuit breaker,



balanced three phase,balanced three phase faults,Barkan,base KVA,

Basic Impulse LevelBIL),basic “RV Calculations,Beattie-Bridgman equation,

bellows,bias test,Biot-Savart law,blast valve, blow-in force, Boehne,Browne,Browne’s Combined Theory,bulk oil breaker,butt contacts,by-products neutralization,

C

cable charging, calculation of electrodynamic

forces,calculation of short circuitcurrents,

calculation of transient recoveryvoltages,

California Institute f Technology,

capacitance switching, capacitor banks, capacitor injection, Cassie’s Theory,cathode drop.,cathode region,cathode voltage,CESI,

charge carriers,charges, trapped, charging motors,

chopped wave psec,chopped wave psec,chopped wave withstand,chopping level, chopping numbers, CIGRE,circuit breaker

air blast,air magnetic,bulkoil,deaddesignsindoor,live tank,

minimum oil,outdoor,preferred ratings, pufferself blast standards,sulfurhexafluoride, vacuum,

circuit simplification,close and latch,closing control,closing resistors,closing speed, coil electrode,coil polarity, commutation time,

condition monitoring,conducting plasma,

conductors at right angles,confidence level,conformance testing, constricted arc,contact

area,

break,erosion,force,fretting



[contact]gap voltage withstand, 25 ,322make, 343mass, 196materials for vacuum, 190,196melting, 200opening, 116softening, 201spirals, 183structure, 195surface, 198temperature, 346theory, 195travel, 275,342welding, 2 16

continuous current rating, 245continuous overload, 295 copper-bismuth, 191copper-chrome, 191 corrective actions, 79 critical gas flow, 175

blast nozzle, 144

current, asymmetrical, 29

current, chopping, 120time constant, orX/R, 300

in air,oil, ihtermpters, 122level, 122numbers, 123vacuum, 124

current injection technique, 80

current injection,268,276current intermption,1current limitingeactors,262current overload

continuous, 295eight hours, 299four hours, 299

current source, 268

current transients, 30 current zero pause, 10

Ddc component, 301de-energizing a capacitor bank,

330

definite-purpose circuitbreakers,253

delta to wye, 46 density of points of contact, 197derating, 281derating factorK, 14derating factorR, 252design testing, 259 deteriorating bearings, 342dielectric recovery region, 85

diffuse arc, 9, 182diffusion welding, 190diffusive cooling, 136 direct test, 263 direction of forces, 58dry withstand, 232double frequency recovery

voltage, 90,99

Dwight, 68

currents, 294eight-hour factor, 299 electric arcs, see rcsElectric Power Club, 226electromagnetic forces,57,203

Electro-Magnetic TransientsProgram or @"P) , 79

electromechanical forces, 266emergency load current, 298 energizing a cable, 306energizing a capacitor bank, 325 equivalent circuit, 79 ex-cos, 102

exhaust valves, 146



firstphase to clear, 6,10 1factor S, 302fast reclosing, 112, 281ferroresonance, 3 13film resistance, 197

Fleming’s left hand rule,8,60Fortesque, 54four-hour factor, 299frequency 16 and 2/3 Hz, 304frequency 25 Hz, 304fretting, 199Frick, 67Frind, 25,136

fuzzy logic, 348

Ggap voltage withstand, 22,325gas blast circuit breakers, 14 1gas bubble, 150gas compressor, 341gas density, 345

gas exhaust, 266gas mixture,181gas moisture, 345general-purpose circuit breaker,

Global Warming Potential (GWP),

global warming radiative forcing,

253

165

166glow

abnonnal, 5discharge, 3luminous, 4normal, 5

Greenwood, 79 ,203

grounded capacitor bank, 329grounding, 275

HHammarlund, 79hammer blow, 216Harner, 277heater elements, 343heating of copper contact, 203

heating o f arc plates, 137Hermann, 25high altitude, 293high current tests, 260high pressure arcs,high TRV, 293high X/R,293,299hydraulic mechanisms, 220

hydraulic pump, 341hydrogen fluoride, 164hydrogen thermal conductivity,

152hydrogen, 15 1

I

EC, 25,255

impulse levels, 233independent pole operation, 322indirect tests, 264inductive load currents, 18influence of wear, aging, cold

temperature, control voltage,322

inherent TRV test circuit, 276

initial exponential, 101initial opening velocity, 212initial rate of rise, 103InitialTransient Recovery Voltage

0 , 1 0 7inrush current, 334inrush frequency nd current, 311insulating oil, 150insulating plates, 136



International ElectrotechnicalCommission @C),

interrupter wear,interrupters in series,interrupter’s nozzle,intermpting time,interruption of

alternating currents,capacitive circuits,direct current,inductive circuits,resistive circuits,

interruption theories,Browne’s,Cassie’s,Mayr’s,modem,hince’s,Slepian’s,

IREQ,

isolated capacitor bank,

KKelman,KEMA Amhem,KEMA-Powertest,Kesselring,Kogelschat,

LLAPEM,lattice diagram, 81, lighting impulse,

standard lighting impulse,

test methodtest method

lighting stroke,

lineopen ended,shorted end,

line charging,line-to-ground fault,line to line,line-to-line fault,

Lingal, 161linkages,low oil content,low pressure (vacuum) arcs,Lowke,L W ensors,Ludwig,luminous glow,

Mmagnetic blow-outassist,magnetic blow-out coil,

magnetic fluxremnants,magnetic pressure,

major failures,making switch,materials contacts,maximum arc energy input,maximum arcing time,maximum force,Mayr’s Theory,measuringasymmetricalcurrents,

mechanism spring charge,melting point,Mendelhall,metal fluorides,methane,minimum arcing time,modern theories,



Moissan,molten metal bridge,monitored parameters,MOSKVA,MVA Method, -Nnaphthenic base petroleumils,

negative-phase-sequencereactance,

negative-phase-sequence,

network modeling,neural networks,neutral shift,Niemayer,,nitrogen accumulator,non-contact transducers,non-self-sustaining discharge,normal glow,nozzle

axial,

conducting,cross blast,D’Laval,insulating,radial,single, double flow,throat diameter,

number of contacts,

0

oil circuit breaker,compensating chambers,

baffle chamber,explosion chamber,interrupting pot,oil poor circuit breaker,

plain break,suicide breakers,

opening controL,opening speed or SF,,opening speed or vacuum,opening speed,operating characteristics,operating duty cycle,operating mechanism,

operating pressure,operating springs,operating times,operating voltage rating, operations counter,optical motion sensors,overload currents,overvoltage factor,

overvoltages,oxide film,oxyfluorides,

parallel conductors,parallel current injection,

partial discharges,peak makingper unit,per unit method,perpendicular conductors, pi (n)ype circuit, pinch-off,plastically deformed,

pneumatic mechanisms,point of contact,positive column voltage,positive-phase-sequence reactance,

positive-phase-sequence, power frequency overvoltage, power frequency rating,

power loss from arc,Power-Tech,



Sfactors, 302

safe overload, 297saturation current, 2Schulman, 183,187self-sustained discharge,2 ,4series current injection, 271

series, parallel compensation, 11 service capability, 252

sF6

arc temperature profile, 167by-products neutralization, 165circuit breaker, 169corrosive effects of, 164'dielectric strength, 162electronegativity, 163environmental, 165 green house effects, 165heat transfer, 163liquefaction, 170 molecular weight, 162ozone depletion, 165peak thermal conductivity, 167sonic velocity, 162thermal conductivity, 163thermal conductivity, 168time constant, 168

Shade, 25 shock ionization, 3short circuit current rating, 246

short circuit current testet-up,

short circuit current, 29short circuit generator, 260short circuit testing, 257short line fault recovery voltage,

short line faults, 85 ,245

short time current rating, 25short time overloads, 297

260

104

preferred number series, 246preferred ratings, 225Prince's Theory, 21proactive signals, 346proximity sensors, 342PSM, 258 puffer interrupter, 173

RR10 series, 246race theory. See Slepian's TheoryRagaller, 25 rate of change of voltage, 328rate of change of withstand

capability, 327rated dielectric 232 rated transient recovery voltage,

rated voltage range factor, 23ratings of a circuit breaker, 228 reactance in parallel, 45 reactance in series, 45reactive signal, 348 reactor currents, 31 reclosing duty, 252

derating factors, 3 16reclosing of a line, 116reflection coefficient, 83refraction coefficient, 83reignition circuit, 273

reignition definition, 116reignitionvoltage, 10, 15, 16 related required capabilities, 229Renard, 246 repulsion forces, 205resistance of contacts, 196restrike definition, 116reversed polarity,118

rod gaps, 235running time, 342

239



Siemens,single flow nozzle, single frequency recovery voltage,

single loopest,single phase tests,Slepian'sSLF parameters, SLF TRVnetwork,solar radiation, solidly grounded,Sorrensen,spring mechanism,standard altitude,steady state losses,

Strom,subtransient,successivereignitions,'ll9sulfide coat,sulfur dioxide,

fluorides,sulfurhexafluoride,suppression of bounce,

surge arrester,surge impedance,surge suppressers,Swanson,switching diode, switching impulseswitching overvoltages,symmetrical components,

symmetrical current ratings,symmetrical shortsymmetrical, ungroundederminal,

synchronous switching,capacitance,closing,interruption,

synthetic tests, advantageddisadvantages of,

parallel current injection,current injection,

T

temperature rise, terminal, orbolted faults,test a one minute, Hz,

test cycle,test duties,test lighting mpdse,test methodtest methodtesting requirements,testing stations,test lighting impulse, theories of ac interruption,thermal and dielectric regions,

equilibrium,thionyl fluoride,three phase system,total force,transient alternating current com-

transient direct current component,

transient overvoltage component,

Transient RecoveryVoltages

ponents,

calculations of,four parameter method,IEC and ANSI,

shaping capacitors, for terminal fault,two parameter method,

High Voltage Circuit Breakers

Documents

Transcript of High Voltage Circuit Breakers