Electrical Protective Relay Theory and Applications

Preface

Continuous change in protective relaying has been caused by two different influences. One is the fact that therequirements imposed by power systems are in a constant state of change, and our understanding of the basicconcepts has sharpened considerably over the years. The other is that the means of implementing the fundamentalconcepts of fault location and removal and system restoration are constantly growing more sophisticated.

It is primarily because of these changing constraints that this text has been revised and expanded. It began withcontributions from two giants of the industry, J. Lewis Blackburn and George D. Rockefeller. From the nucleus oftheir extensive analyses and writings, and the desire to cover each new contingency with new relaying concepts, thisvolume has evolved. New solutions to age-old problems have become apparent as greater experience has beengained. No problem is without benefit in the solution found.

This new edition weeds out those relaying concepts that have run their course and have been replaced by moreperceptive methods of implementation using new solid-state or microprocessor-based devices.

No single technological breakthrough has been more influential in generating change than the microprocessor.Initially, the methods of translating a collection of instantaneous samples of sine waves into useful current,direction, and impedance measurements were not obvious. Diligent analysis and extensive testing allowed theseuseful functions to be obtained and to be applied to the desired protective functions. This text attempts to describe,in the simplest possible terms, the manner in which these digital measurements are accomplished in present-daydevices.

In addition to those already mentioned, huge contributions were made in the development and refinement of theconcepts described in this book by Hung Jen Li, Walter Hinman, Roger Ray, James Crockett, Herb Lensner, AlRegotti, Fernando Calero, Eric Udren, James Greene, Liancheng Wang, Elmo Price, Solveig Ward, JohnMcGowan, and Cliff Downs. Some of these names may not be immediately recognizable, but all have made animpact with their thoughtful, accurate, well-reasoned writings, and they all deserve the gratitude of the industry forthe wealth of knowledge they have contributed to this book. I am keenly aware of the high quality of the technicalofferings of these people, and I am particularly grateful for the warmth and depth of their friendship.

Walter A. Elmore

iii

Contents

Preface iii

1 Introduction and General Philosophies 1

Revised by W. A. Elmore

1 Introduction 12 Classification of Relays 1

2.1 Analog/Digital/Numerical 23 Protective Relaying Systems and Their Design 2

3.1 Design Criteria 33.2 Factors Influencing Relay Performance 43.3 Zones of Protection 4

4 Applying Protective Relays 44.1 System Configuration 54.2 Existing System Protection and Procedures 54.3 Degree of Protection Required 54.4 Fault Study 54.5 Maximum Loads, Transformer Data, and Impedances 6

5 Relays and Application Data 65.1 Switchboard Relays 65.2 Rack-Mounted Relays 7

6 Circuit-Breaker Control 87 Comparison of Symbols 9

2 Technical Tools of the Relay Engineer: Phasors, Polarity, and Symmetrical Components 11


1 Introduction 112 Phasors 11

2.1 Circuit Diagram Notation for Current and Flux 112.2 Circuit Diagram Notation for Voltage 12

v

2.3 Phasor Notation 122.4 Phasor Diagram Notation 132.5 Phase Rotation vs. Phasor Rotation 15

3 Polarity in Relay Circuits 153.1 Polarity of Transformers 153.2 Polarity of Protective Relays 153.3 Characteristics of Directional Relays 163.4 Connections of Directional Units to Three-Phase Power Systems 17

4 Faults on Power Systems 184.1 Fault Types and Causes 184.2 Characteristics of Faults 20

5 Symmetrical Components 215.1 Basic Concepts 215.2 System Neutral 235.3 Sequences in a Three-Phase Power System 235.4 Sequence Impedances 245.5 Sequence Networks 265.6 Sequence Network Connections and Voltages 275.7 Network Connections for Fault and General Unbalances 285.8 Sequence Network Reduction 295.9 Example of Fault Calculation on a Loop-Type Power System 325.10 Phase Shifts Through Transformer Banks 375.11 Fault Evaluations 39

6 Symmetrical Components and Relaying 42

3 Basic Relay Units 43


1 Introduction 432 Electromechanical Units 43

2.1 Magnetic Attraction Units 432.2 Magnetic Induction Units 452.3 D’Arsonval Units 472.4 Thermal Units 47

3 Sequence Networks 473.1 Zero Sequence Networks 473.2 Composite Sequence Current Networks 483.3 Sequence Voltage Networks 49

4 Solid-State Units 504.1 Semiconductor Components 504.2 Solid-State Logic Units 524.3 Principal Logic Units 52

5 Basic Logic Circuits 545.1 Fault-Sensing Data Processing Units 545.2 Amplification Units 595.3 Auxiliary Units 59

6 Integrated Circuits 636.1 Operational Amplifier 636.2 Basic Operational Amplifier Units 656.3 Relay Applications of Operational Amplifier 68

7 Microprocessor Architecture 70

vi Contents

4 Protection Against Transients and Surges 71

W. A. Elmore

1 Introduction 711.1 Electrostatic Induction 711.2 Electromagnetic Induction 721.3 Differential- and Common-Mode Classifications 72

2 Transients Originating in the High-Voltage System 732.1 Capacitor Switching 732.2 Bus Deenergization 732.3 Transmission Line Switching 742.4 Coupling Capacitor Voltage Transformer (CCVT) Switching 742.5 Other Transient Sources 74

3 Transients Originating in the Low-Voltage System 743.1 Direct Current Coil Interruption 743.2 Direct Current Circuit Energization 753.3 Current Transformer Saturation 753.4 Grounding of Battery Circuit 75

4 Protective Measures 754.1 Separation 754.2 Suppression at the Source 774.3 Suppression by Shielding 774.4 Suppression by Twisting 774.5 Radial Routing of Control Cables 784.6 Buffers 784.7 Optical Isolators 784.8 Increased Energy Requirement 79

5 Instrument Transformers for Relaying 81

W. A. Elmore

1 Introduction 812 Current Transformers 81

2.1 Saturation 812.2 Effect of dc Component 82

3 Equivalent Circuit 824 Estimation of Current Transformer Performance 82

4.1 Formula Method 834.2 Excitation Curve Method 834.3 ANSI Standard: Current Transformer Accuracy Classes 85

5 European Practice 875.1 TPX 885.2 TPY 885.3 TPZ 88

6 Direct Current Saturation 887 Residual Flux 898 MOCT 919 Voltage Transformers and Coupling Capacitance Voltage Transformers 91

9.1 Equivalent Circuit of a Voltage Transformer 919.2 Coupling Capacitor Voltage Transformers 929.3 MOVT/EOVT 93

10 Neutral Inversion 93

Contents vii

6 Microprocessor Relaying Fundamentals 95

W. A. Elmore

1 Introduction 952 Sampling Problems 973 Aliasing 974 How to Overcome Aliasing 98

4.1 Antialiasing Filters 984.2 Nonsynchronous Sampling 98

5 Choice of Measurement Principle 995.1 rms Calculation 1005.2 Digital Filters 1005.3 Fourier-Notch Filter 1005.4 Another Digital Filter 1015.5 dc Offset Compensation 1015.6 Symmetrical Component Filter 1025.7 Leading-Phase Identification 1025.8 Fault Detectors 102

6 Self-Testing 1036.1 Dead-Man Timer 1036.2 Analog Test 1036.3 Check-Sum 1036.4 RAM Test 1036.5 Nonvolatile Memory Test 103

7 Conclusions 104

7 System Grounding and Protective Relaying 105


1 Introduction 1052 Ungrounded Systems 105

2.1 Ground Faults on Ungrounded Systems 1052.2 Ground Fault Detection on Ungrounded Systems 107

3 Reactance Grounding 1083.1 High-Reactance Grounding 1083.2 Resonant Grounding (Ground Fault Neutralizer) 1093.3 Low-Reactance Grounding 109

4 Resistance Grounding 1104.1 Low-Resistance Grounding 1104.2 High-Resistance Grounding 111

5 Sensitive Ground Relaying 1125.1 Ground Overcurrent Relay with Conventional Current Transformers 1125.2 Ground Product Relay with Conventional Current Transformers 1135.3 Ground Overcurrent Relay with Zero Sequence Current Transformers 114

6 Ground Fault Protection for Three-Phase, Four-Wire Systems 1146.1 Unigrounded Four-Wire Systems 1146.2 Multigrounded Four-Wire Systems 115

8 Generator Protection 117

Revised by C. L. Downs

1 Introduction 1172 Choice of Technology 117

viii Contents

3 Phase Fault Detection 1173.1 Percentage Differential Relays (Device 87) 1183.2 High Impedance Differential Relays (Device 87) 1193.3 Machine Connections 1193.4 Split-Phase 119

4 Stator Ground Fault Protection 1204.1 Unit-Connected Schemes 1204.2 95% Ground Relays 1204.3 Neutral-to-Ground Fault Detection (Device 87N3) 1214.4 100% Winding Protection 122

5 Backup Protection 1235.1 Unbalanced Faults 1235.2 Balanced Faults 124

6 Overload Protection 1266.1 RTD Schemes (Device 49) 1266.2 Thermal Replicas (Device 49) 126

7 Volts per Hertz Protection 1268 Overspeed Protection 1269 Loss-of-Excitation Protection 127

9.1 Causes of Machine Loss of Field 1279.2 Hazard 1279.3 Loss-of-Field Relays 1289.4 KLF and KLF-1 Curves 1299.5 Two-Zone KLF Scheme 129

10 Protection Against Generator Motoring 13010.1 Steam Turbines 13110.2 Diesel Engines 13110.3 Gas Turbines 13110.4 Hydraulic Turbines 131

11 Inadvertent Energization 13212 Field Ground Detection 134

12.1 Brush-Type Machine 13512.2 Brushless Machines 13612.3 Injection Scheme for Field Ground Detection 136

13 Alternating-Current Overvoltage Protection for Hydroelectric Generators 13614 Generator Protection at Reduced Frequencies 13615 Off-Frequency Operation 13816 Recommended Protection 13917 Out-of-Step Protection 13918 Bus Transfer Systems for Station Auxiliaries 139

18.1 Fast Transfer 13918.2 Choice of Fast Transfer Scheme 14018.3 Slow Transfer 142

19 Microprocessor-Based Generator Protection 143

9 Motor Protection 145

Revised by C. L. Downs

1 Introduction 1451.1 General Requirements 1451.2 Induction Motor Equivalent Circuit 1461.3 Motor Thermal Capability Curves 146

Contents ix

2 Phase-Fault Protection 1473 Ground-Fault Protection 1474 Locked-Rotor Protection 1495 Overload Protection 1536 Thermal Relays 153

6.1 RTD-Input-Type Relays 1546.2 Thermal Replica Relays 154

7 Low-Voltage Protection 1558 Phase-Rotation Protection 1559 Negative Sequence Voltage Protection 155

10 Phase-Unbalance Protection 15611 Negative Sequence Current Relays 15712 Jam Protection 15713 Load Loss Protection 15714 Out-of-Step Protection 15815 Loss of Excitation 15816 Typical Application Combinations 159

10 Transformer and Reactor Protection 163

Revised by J. J. McGowan

1 Introduction 1632 Magnetizing Inrush 163

2.1 Initial Inrush 1632.2 Recovery Inrush 1652.3 Sympathetic Inrush 165

3 Differential Relaying for Transformer Protection 1663.1 Differential Relays for Transformer Protection 1663.2 General Guidelines for Transformer Differential Relaying Application 171

4 Sample Checks for Applying Transformer Differential Relays 1734.1 Checks for Two-Winding Banks 1734.2 Checks for Multiwinding Banks 1784.3 Modern Microprocessor Relay 180

5 Typical Application of Transformer Protection 1805.1 Differential Scheme with Harmonic Restraint Relay Supervision 1805.2 Ground Source on Delta Side 1825.3 Three-Phase Banks of Single-Phase Units 1835.4 Differential Protection of a Generator-Transformer Unit 1835.5 Overexcitation Protection of a Generator-Transformer Unit 1845.6 Sudden-Pressure Relay (SPR) 1855.7 Overcurrent and Backup Protection 1855.8 Distance Relaying for Backup Protection 1925.9 Overcurrent Relay with HRU Supplement 192

6 Typical Protective Schemes for Industrial and Commercial Power Transformers 1937 Remote Tripping of Transformer Bank 1978 Protection of Phase-Angle Regulators and Voltage Regulators 1979 Zig-Zag Transformer Protection 202

10 Protection of Shunt Reactors 20310.1 Shunt Reactor Applications 20310.2 Rate-of-Rise-of-Pressure Protection 20510.3 Overcurrent Protection 20510.4 Differential Protection 206

x Contents

10.5 Reactors on Delta System 20710.6 Turn-to-Turn Faults 209

11 Station-Bus Protection 213

Revised by Solveig Ward

1 Introduction 2131.1 Current Transformer Saturation Problem and Its Solutions on Bus Protection 2131.2 Information Required for the Preparation of a Bus Protective Scheme 2151.3 Normal Practices on Bus Protection 215

2 Bus Differential Relaying with Overcurrent Relays 2162.1 Overcurrent Differential Protection 2162.2 Improved Overcurrent Differential Protection 216

3 Multirestraint Differential System 2174 High Impedance Differential System 219

4.1 Factors that Relate to the Relay Setting 2214.2 Factors that Relate to the High-Voltage Problem 2214.3 Setting Example for the KAB Bus Protection 222

5 Differential Comparator Relays 2226 Protecting a Bus that Includes a Transformer Bank 2237 Protecting a Double-Bus Single-Breaker with Bus Tie Arrangement 2248 Other Bus Protective Schemes 226

8.1 Partial Differential Relaying 2268.2 Directional Comparison Relaying 2278.3 Fault Bus (Ground-Fault Protection Only) 227

12 Line and Circuit Protection 229

Revised by Elmo Price

1 Introduction 2291.1 Classification of Electric Power Lines 2291.2 Techniques for Line Protection 2291.3 Seleting a Protective System 2291.4 Relays for Phase- and Ground-Fault Protection 2301.5 Multiterminal and Tapped Lines and Weak Feed 230

2 Overcurrent Phase- and Ground-Fault Protection 2312.1 Fault Detection 2312.2 Time Overcurrent Protection 2322.3 Instantaneous Overcurrent Protection 2372.4 Overcurrent Ground-Fault Protection 238

3 Directional Overcurrent Phase- and Ground-Fault Protection 2393.1 Criteria for Phase Directional Overcurrent Relay Applications 2393.2 Criteria for Ground Directional Overcurrent Relay Applications 2393.3 Directional Ground-Relay Polarization 2393.4 Mutual Induction and Ground-Relay Directional Sensing 2433.5 Applications of Negative Sequence Directional Units for Ground Relays 2443.6 Selection of Directional Overcurrent Phase and Ground Relays 244

4 Distance Phase and Ground Protection 2474.1 Fundamentals of Distance Relaying 2474.2 Phase-Distance Relays 2504.3 Ground-Distance Relays 2544.4 Effect of Line Length 2574.5 The Infeed Effect on Distance-Relay Application 260

Contents xi

4.6 The Outfeed Effect on Distance-Relay Applications 2614.7 Effect of Tapped Transformer Bank on Relay Application 2614.8 Distance Relays with Transformer Banks at the Terminal 2624.9 Fault Resistance and Ground-Distance Relays 2654.10 Zero Sequence Mutual Impedance and Ground-Distance Relays 265

5 Loop-System Protection 2675.1 Single-Source Loop-Circuit Protection 2675.2 Multiple-Source Loop Protection 269

6 Short-Line Protection 2706.1 Definition of Short Line 2706.2 Problem Associated with Short-Line Protection 2706.3 Current-Only Scheme for Short-Line Protection 2706.4 Distance Relay for Short-Line Protection 270

7 Series-Capacitor Compensated-Line Protection 2737.1 A Series-Capacitor Compensated Line 2737.2 Relaying Quantities Under Fault Conditions 2737.3 Distance Protection Behavior 2757.4 Practical Considerations 276

8 Distribution Feeder Protection 2768.1 Relay Coordination with Reclosers and Sectionalizers on a Feeder 2778.2 Coordinating with Low-Voltage Breaker and Fuse 277

Appendix A: Equation (12-2) 281Appendix B: Impedance Unit Characteristics 281

B.1 Introduction 281B.2 Basic Application Example of a Phase Comparator 284B.3 Basic Application Example of a Magnitude Comparator 285B.4 Practical Comparator Applications in Distance Relaying 285B.5 Reverse Characteristics of an Impedance Unit 294B.6 Response of Distance Units to Different Types of Faults 298B.7 The Influence of Current Distribution Factors and Load Flow 302B.8 Derived Characteristics 305B.9 Apparent Impedance 305B.10 Summary 306

Appendix C: Infeed Effect on Ground-Distance Relays 306C.1 Infeed Effect on Type KDXG, LDAR, and MDAR Ground-Distance Relays 306C.2 Infeed Effect on Type SDG and LDG Ground-Distance Relays 307

Appendix D: Coordination in Multiple-Loop Systems 308D.1 System Information 308D.2 Relay Type Selection 308D.3 Relay Setting and Coordination 309

13 Backup Protection 323

Revised by E. D. Price

1 Introduction 3232 Remote vs. Local Backup 323

2.1 Remote Backup 3232.2 Local Backup and Breaker Failure 3242.3 Applications Requiring Remote Backup with Breaker-Failure Protection 326

3 Breaker-Failure Relaying Applications 3273.1 Single-Line/Single-Breaker Buses 3273.2 Breaker-and-a-Half and Ring Buses 328

xii Contents

4 Traditional Breaker-Failure Scheme 3294.1 Timing Characteristics of the Traditional Breaker-Failure Scheme 3294.2 Traditional Breaker-Failure Relay Characteristics 3304.3 Microprocessor Relays 331

5 An Improved Breaker-Failure Scheme 3325.1 Problems in the Traditional Breaker-Failure Scheme 3325.2 The Improved Breaker-Failure Scheme 3335.3 Type SBF-1 Relay 334

6 Open Conductor and Breaker Pole Disagreement Protection 3367 Special Breaker-Failure Scheme for Single-Pole Trip-System Application 337

14 System Stability and Out-of-Step Relaying 339

W. A. Elmore

1 Introduction 3392 Steady-State Stability 3393 Transient Stability 3404 Relay Quantities During Swings 3415 Effect of Out-of-Step Conditions 343

5.1 Distance Relays 3435.2 Directional Comparison Systems 3445.3 Phase-Comparison or Pilot-Wire Systems 3445.4 Underreaching Transfer-Trip Schemes 3445.5 Circuit Breakers 3445.6 Overcurrent Relays 3445.7 Reclosing 344

6 Out-of-Step Relaying 3456.1 Generator Out-of-Step Relaying 3456.2 Transmission-Line Out-of-Step Relaying 346

7 Philosophies of Out-of-Step Relaying 3467.1 Utility Practice 347

8 Types of Out-of-Step Schemes 3478.1 Concentric Circle Scheme 3478.2 Blinder Scheme 348

9 Relays for Out-of-Step Systems 3489.1 Electromechanical Types 3489.2 Solid-State Types 349

10 Selection of an Out-of-Step Relay System 351

15 Voltage Stability 353

L. Wang

1 Introduction 3531.1 Small-Disturbance Instability 3531.2 Large-Disturbance Instability 3551.3 Voltage Instability Incidents 356

2 Voltage Instability Indices 3572.1 Indices Based on Current Operating Condition 3572.2 Indices Based on Stressed System Conditions 3602.3 Summary 362

3 Voltage Instability Protection 3623.1 Reactive Power Control 3623.2 Load Tap Changer Blocking Schemes 3623.3 Load Shedding 362

Contents xiii

16 Reclosing and Synchronizing 365

Revised by S. Ward

1 Introduction 3652 Reclosing Precautions 3653 Reclosing System Considerations 366

3.1 One-Shot vs. Multiple-Shot Reclosing Relays 3663.2 Selective Reclosing 3663.3 Deionizing Times for Three-Pole Reclosing 3663.4 Synchronism Check 3663.5 Live-Line/Dead-Bus, Live-Bus/Dead-Line Control 3673.6 Instantaneous-Trip Lockout 3673.7 Intermediate Lockout 3673.8 Compatibility with Supervisory Control 3673.9 Inhibit Control 3683.10 Breaker Supervision Functions 3683.11 Factors Governing Application of Reclosing 368

4 Considerations for Applications of Instantaneous Reclosing 3684.1 Feeders with No-Fault-Power Back-Feed and Minimum Motor Load 3694.2 Single Ties to Industrial Plants with Local Generation 3694.3 Lines with Sources at Both Ends 369

5 Reclosing Relays and Their Operation 3695.1 Review of Breaker Operation 3695.2 Single-Shot Reclosing Relays 3695.3 Multishot Reclosing Relays 371

6 Synchronism Check 3776.1 Phasing Voltage Synchronism Check Characteristic 3776.2 Angular Synchronism Check Characteristic 378

7 Dead-Line or Dead-Bus Reclosing 3798 Automatic Synchronizing 379

17 Load-Shedding and Frequency Relaying 381


1 Introduction 3812 Rate of Frequency Decline 3813 Load-Shedding 3834 Frequency Relays 384

4.1 KF Induction-Cylinder Underfrequency Relay 3844.2 Digital Frequency Relays 3854.3 Microprocessor-Based Frequency Relay 385

5 Formulating a Load-Shedding Scheme 3855.1 Maximum Anticipated Overload 3855.2 Number of Load-Shedding Steps 3865.3 Size of the Load Shed at Each Step 3865.4 Frequency Settings 3875.5 Time Delay 3885.6 Location of the Frequency Relays 388

6 Special Considerations for Industrial Systems 389

xiv Contents

7 Restoring Service 3908 Other Frequency Relay Applications 391

Bibliography 395

Index 399

Contents xv

1

Introduction and General Philosophies

Revised by: W. A. ELMORE

1 INTRODUCTION

Relays are compact analog, digital, and numericaldevices that are connected throughout the powersystem to detect intolerable or unwanted conditionswithin an assigned area. They are, in effect, a form ofactive insurance designed to maintain a high degree ofservice continuity and limit equipment damage. Theyare ‘‘silent sentinels.’’ Although protective relays willbe the main emphasis of this book, other types ofrelays applied on a more limited basis or used as partof a total protective relay system will also be covered.

2 CLASSIFICATION OF RELAYS

Relays can be divided into six functional categories:

Protective relays. Detect defective lines, defectiveapparatus, or other dangerous or intolerableconditions. These relays generally trip one ormore circuit breaker, but may also be used tosound an alarm.

Monitoring relays. Verify conditions on the powersystem or in the protection system. These relaysinclude fault detectors, alarm units, channel-monitoring relays, synchronism verification, andnetwork phasing. Power system conditions thatdo not involve opening circuit breakers duringfaults can be monitored by verification relays.

Reclosing relays. Establish a closing sequence for acircuit breaker following tripping by protectiverelays.

Regulating relays. Are activated when an operat-ing parameter deviates from predeterminedlimits. Regulating relays function through sup-plementary equipment to restore the quantity tothe prescribed limits.

Auxiliary relays. Operate in response to the open-ing or closing of the operating circuit tosupplement another relay or device. Theseinclude timers, contact-multiplier relays, sealingunits, isolating relays, lockout relays, closingrelays, and trip relays.

Synchronizing (or synchronism check) relays. As-sure that proper conditions exist for intercon-necting two sections of a power system.

Many modern relays contain several varieties of thesefunctions. In addition to these functional categories,relays may be classified by input, operating principle orstructure, and performance characteristic. The follow-ing are some of the classifications and definitionsdescribed in ANSI/IEEE Standard C37.90 (see alsoANSI/IEEE C37.100 ‘‘Definitions for Power Switch-gear’’):

InputsCurrentVoltagePowerPressureFrequencyTemperatureFlowVibration

1

Operating Principle or StructuresCurrent balancePercentageMultirestraintProductSolid stateStaticMicroprocessorElectromechanicalThermal

Performance CharacteristicsDifferentialDistanceDirectional overcurrentInverse timeDefinite timeUndervoltageOvervoltageGround or phaseHigh or low speedPilot

Phase comparisonDirectional comparisonCurrent differential

A separate volume, Pilot Protective Relaying, coverspilot systems (those relaying functions that involve acommunications channel between stations.

2.1 Analog/Digital/Numerical

Solid-state (and static) relays are further categorizedunder one of the following designations.

2.1.1 Analog

Analog relays are those in which the measuredquantities are converted into lower voltage but similarsignals, which are then combined or compared directlyto reference values in level detectors to produce thedesired output (e.g., SA-1 SOQ, SI-T, LCB, circuitshield relays).

2.1.2 Digital

Digital relays are those in which the measured acquantities are manipulated in analog form andsubsequently converted into square-wave (binary)voltages. Logic circuits or microprocessors comparethe phase relationships of the square waves to make atrip decision (e.g., SKD-T, REZ-1).

2.1.3 Numerical

Numerical relays are those in which the measured acquantities are sequentially sampled and converted intonumeric data form. A microprocessor performsmathematical and/or logical operations on the datato make trip decisions (e.g., MDAR, MSOC, DPU,TPU, REL-356, REL-350, REL-512).

3 PROTECTIVE RELAYING SYSTEMS ANDTHEIR DESIGN

Technically, most relays are small systems withinthemselves. Throughout this book, however, the termsystem will be used to indicate a combination of relaysof the same or different types. Properly speaking, theprotective relaying system includes circuit breakers andcurrent transformers (ct’s) as well as relays. Relays,ct’s, and circuit breakers must function together. Thereis little or no value in applying one without the other.

Protective relays or systems are not required tofunction during normal power system operation, butmust be immediately available to handle intolerablesystem conditions and avoid serious outages anddamage. Thus, the true operating life of these relayscan be on the order of a few seconds, even though theyare connected in a system for many years. In practice,the relays operate far more during testing and main-tenance than in response to adverse service conditions.

In theory, a relay system should be able to respondto an infinite number of abnormalities that canpossibly occur within the power system. In practice,the relay engineer must arrive at a compromise basedon the four factors that influence any relay application:

Economics. Initial, operating, and maintenanceAvailable measures of fault or troubles. Fault

magnitudes and location of current transformersand voltage transformers

Operating practices. Conformity to standards andaccepted practices, ensuring efficient systemoperation

Previous experience. History and anticipation ofthe types of trouble likely to be encounteredwithin the system

The third and fourth considerations are perhaps betterexpressed as the ‘‘personality of the system and therelay engineer.’’

Since it is simply not feasible to design a protectiverelaying system capable of handling any potentialproblem, compromises must be made. In general, only

2 Chapter 1

those problems that, according to past experience, arelikely to occur receive primary consideration. Natu-rally, this makes relaying somewhat of an art. Differentrelay engineers will, using sound logic, design sig-nificantly different protective systems for essentiallythe same power system. As a result, there is littlestandardization in protective relaying. Not only maythe type of relaying system vary, but so will the extentof the protective coverage. Too much protection isalmost as bad as too little.

Nonetheless, protective relaying is a highly specia-lized technology requiring an in-depth understandingof the power system as a whole. The relay engineermust know not only the technology of the abnormal,but have a basic understanding of all the systemcomponents and their operation in the system. Relay-ing, then, is a ‘‘vertical’’ speciality requiring a‘‘horizontal’’ viewpoint. This horizontal, or totalsystem, concept of relaying includes fault protectionand the performance of the protection system duringabnormal system operation such as severe overloads,generation deficiency, out-of-step conditions, and soforth. Although these areas are vitally important to therelay engineer, his or her concern has not always beenfully appreciated or shared by colleagues. For thisreason, close and continued communication betweenthe planning, relay design, and operation departmentsis essential. Frequent reviews of protective systemsshould be mandatory, since power systems grow andoperating conditions change.

A complex relaying system may result from poorsystem design or the economic need to use fewer circuitbreakers. Considerable savings may be realized byusing fewer circuit breakers and a more complex relaysystem. Such systems usually involve design compro-mises requiring careful evaluation if acceptable protec-tion is to be maintained. It should be recognized thatthe exercise of the very best relaying applicationprinciples can never compensate for the absence of aneeded circuit breaker.

3.1 Design Criteria

The application logic of protective relays divides thepower system into several zones, each requiring its owngroup of relays. In all cases, the four design criterialisted below are common to any well-designed andefficient protective system or system segment. Since itis impractical to satisfy fully all these design criteriasimultaneously, the necessary compromises must beevaluated on the basis of comparative risks.

3.1.1 Reliability

System reliability consists of two elements: depend-ability and security. Dependability is the degree ofcertainty of correct operation in response to systemtrouble, whereas security is the degree of certainty thata relay will not operate incorrectly. Unfortunately,these two aspects of reliability tend to counter oneanother; increasing security tends to decrease depend-ability and vice versa. In general, however, modernrelaying systems are highly reliable and provide apractical compromise between security and depend-ability. The continuous supervision made possible bynumerical techniques affords improvement in bothdependability and security. Protective relay systemsmust perform correctly under adverse system andenvironmental conditions.

Dependability can be checked relatively easily in thelaboratory or during installation by simulated tests ora staged fault. Security, on the other hand, is muchmore difficult to check. A true test of system securitywould have to measure response to an almost infinitevariety of potential transients and counterfeit troubleindications in the power system and its environment. Asecure system is usually the result of a good back-ground in design, combined with extensive modelpower system or EMTP (electromagnetic transientprogram) testing, and can only be confirmed in thepower system itself and its environment.

3.1.2 Speed

Relays that could anticipate a fault are utopian. But,even if available, they would doubtlessly raise thequestion of whether or not the fault or trouble reallyrequired a trip-out. The development of faster relaysmust always be measured against the increasedprobability of more unwanted or unexplained opera-tions. Time is an excellent criterion for distinguishingbetween real and counterfeit trouble.

Applied to a relay, high speed indicates that theoperating time usually does not exceed 50ms (threecycles on a 60-Hz base). The term instantaneousindicates that no delay is purposely introduced in theoperation. In practice, the terms high speed andinstantaneous are frequently used interchangeably.

3.1.3 Performance vs. Economics

Relays having a clearly defined zone of protectionprovide better selectivity but generally cost more.High-speed relays offer greater service continuity byreducing fault damage and hazards to personnel, but

Introduction and General Philosophies 3

also have a higher initial cost. The higher performanceand cost cannot always be justified. Consequently,both low- and high-speed relays are used to protectpower systems. Both types have high reliabilityrecords. Records on protective relay operations con-sistently show 99.5% and better relay performance.

3.1.4 Simplicity

As in any other engineering discipline, simplicity in aprotective relay system is always the hallmark ofgood design. The simplest relay system, however, isnot always the most economical. As previouslyindicated, major economies may be possible with acomplex relay system that uses a minimum numberof circuit breakers. Other factors being equal,simplicity of design improves system reliability—ifonly because there are fewer elements that canmalfunction.

3.2 Factors Influencing Relay Performance

Relay performance is generally classed as (1) correct,(2) no conclusion, or (3) incorrect. Incorrect operationmay be either failure to trip or false tripping. The causeof incorrect operation may be (1) poor application, (2)incorrect settings, (3) personnel error, or (4) equipmentmalfunction. Equipment that can cause an incorrectoperation includes current transformers, voltage trans-formers, breakers, cable and wiring, relays, channels,or station batteries.

Incorrect tripping of circuit breakers not associatedwith the trouble area is often as disastrous as a failureto trip. Hence, special care must be taken in bothapplication and installation to ensure against this.

‘‘No conclusion’’ is the last resort when no evidenceis available for a correct or incorrect operation. Quiteoften this is a personnel involvement.

3.3 Zones of Protection

The general philosophy of relay applications is todivide the power system into zones that can beprotected adequately with fault recognition andremoval producing disconnection of a minimumamount of the system.

The power system is divided into protective zonesfor

1. Generators2. Transformers3. Buses

4. Transmission and distribution circuits5. Motors

A typical power system and its zones of protection areshown in Figure 1-1. The location of the currenttransformers supplying the relay or relay systemdefines the edge of the protective zone. The purposeof the protective system is to provide the first line ofprotection within the guidelines outlined above. Sincefailures do occur, however, some form of backupprotection is provided to trip out the adjacent breakersor zones surrounding the trouble area.

Protection in each zone is overlapped to avoid thepossibility of unprotected areas. This overlap isaccomplished by connecting the relays to currenttransformers, as shown in Figure 1-2a. It shows theconnection for ‘‘dead tank’’ breakers, and Figure 1-2bthe ‘‘live tank’’ breakers commonly used with EHVcircuits. Any trouble in the small area between thecurrent transformers will operate both zone A and Brelays and trip all breakers in the two zones. InFigure 1-2a, this small area represents the breaker, andin Figure 1-2b the current transformer, which isgenerally not part of the breaker.

4 APPLYING PROTECTIVE RELAYS

The first step in applying protective relays is to statethe protection problem accurately. Although develop-ing a clear, accurate statement of the problem canoften be the most difficult part, the time spent will paydividends—particularly when assistance from others is

Figure 1-1 A typical system and its zones of protection.

4 Chapter 1

desired. Information on the following associated orsupporting areas is necessary:

System configurationExisting system protection and any known deficien-

ciesExisting operating procedures and practices and

possible future expansionsDegree of protection requiredFault studyMaximum load and current transformer locations

and ratiosVoltage transformer locations, connections, and

ratiosImpedance of lines, transformers, and generators

4.1 System Configuration

System configuration is represented by a single-linediagram showing the area of the system involved in theprotection application. This diagram should show indetail the location of the breakers; bus arrangements;taps on lines and their capacity; location and size of thegeneration; location, size, and connections of thepower transformers and capacitors; location and ratioof ct’s and vt’s; and system frequency.

Transformer connections are particularly impor-tant. For ground relaying, the location of all ground‘‘sources’’ must also be known.

4.2 Existing System Protection and Procedures

The existing protective equipment and reasons for thedesired change(s) should be outlined. Deficiencies inthe present relaying system are a valuable guide toimprovements. New installations should be so speci-fied. As new relay systems will often be required tooperate with or utilize parts of the existing relaying,details on these existing systems are important.

Whenever possible, changes in system protectionshould conform with existing operating proceduresand practices. Exceptions to standard procedures tendto increase the risk of personnel error and may disruptthe efficient operation of the system. Anticipatedsystem expansions can also greatly influence the choiceof protection.

4.3 Degree of Protection Required

To determine the degree of protection required, thegeneral type of protection being considered should beoutlined, together with the system conditions oroperating procedures and practices that will influencethe final choice. These data will provide answers to thefollowing types of questions. Is pilot, high-, medium-,or slow-speed relaying required? Is simultaneoustripping of all breakers of a transmission line required?Is instantaneous reclosing needed? Are generatorneutral-to-ground faults to be detected?

4.4 Fault Study

An adequate fault study is necessary in almost all relayapplications. Three-phase faults, line-to-ground faults,and line-end faults should all be included in the study.Line-end fault (fault on the line side of an openbreaker) data are important in cases where one breakermay operate before another. For ground-relaying, thefault study should include zero sequence currents andvoltages and negative sequence currents and voltages.These quantities are easily obtained during the courseof a fault study and are often extremely useful insolving a difficult relaying problem.

Figure 1-2 The principle of overlapping protection around

a circuit breaker.


4.5 Maximum Loads, Transformer Data, andImpedances

Maximum loads, current and voltage transformerconnections, ratios and locations, and dc voltage arerequired for proper relay application. Maximum loadsshould be consistent with the fault data and based onthe same system conditions. Line and transformerimpedances, transformer connections, and groundingmethods should also be known. Phase sequence shouldbe specified if three-line connection drawings areinvolved.

Obviously, not all the above data are necessary inevery application. It is desirable, however, to reviewthe system with respect to the above points and,wherever applicable, compile the necessary data.

In any event, no amount of data can ensure asuccessful relay application unless the protectionproblems are first defined. In fact, the applicationproblem is essentially solved when the availablemeasures for distinguishing between tolerable andintolerable conditions can be identified and specified.

5 RELAYS AND APPLICATION DATA

Connected to the power system through the currentand voltage transformers, protective relays are wiredinto the control circuit to trip the proper circuitbreakers. In the following discussion, typical connec-tions for relays mounted on conventional switchboardsand for rack-mounted solid-state relays will be used toillustrate the standard application practices andtechniques.

5.1 Switchboard Relays

Many relays are supplied in a rectangular case that ispermanently mounted on a switchboard located in thesubstation control house. The relay chassis, in someimplementations, slides into the case and can beconveniently removed for testing and maintenance.The case is usually mounted flush and permanentlywired to the input and control circuits. In the Flexitestcase, the electrical connections are made throughsmall, front-accessible, knife-blade switches. A typicalswitchboard relay is shown in Figure 1-3; its corre-sponding internal schematic is shown in Figure 1-4.While the example shown is an electromechanicalrelay, many solid-state relays are in the Flexitest casefor switchboard mounting.

The important designations in the ac schematic forthe relay, such as that illustrated in Figure 1-5, are

Phase rotationTripping directionCurrent and voltage transformer polarities

Figure 1-3 A typical switchboard type relay. (The CR

directional time overcurrent relay in the Flexitest case.)

Figure 1-4 Typical internal schematic for a switchboard-

mounted relay. (The circuit shown is for the CR directional

time overcurrent relay of Figure 1-3.)

6 Chapter 1

Relay polarity and terminal numbersPhasor diagram

All these designations are required for a directionalrelay. In other applications, some may not apply. Inaccordance with convention, all relay contacts areshown in the position they assume when the relay isdeenergized.

A typical control circuit is shown in Figure 1-6.Three phase relays and one ground relay are shownprotecting this circuit. Any one could trip theassociated circuit breaker to isolate the trouble orfault area. A station battery, either 125Vdc or 250Vdc,is commonly used for tripping. Lower-voltage batteriesare not recommended for tripping service when longtrip leads are involved.

In small stations where a battery cannot be justified,tripping energy is obtained from a capacitor tripdevice. This device is simply a capacitor charged,through a rectifier, by the ac line voltage. An exampleof this arrangement is presented in Figure 1-7. Whenthe relay contacts close, the discharge of the energy inthe capacitor through the trip coil is sufficient to trip

the breaker. Line voltage cannot be used directly since,of course, it may be quite low during fault conditions.

5.2 Rack-Mounted Relays

Solid-state and microprocessor relays are usually rack-mounted (Fig. 1-8). Since these relays involve morecomplex and sophisticated circuitry, different levels ofinformation are required to understand their opera-tion. A block diagram provides understanding of thebasic process. Figure 1-9 is a block diagram for theMDAR microprocessor relay. Detailed logic diagramsplus ac and dc schematics are also required for acomplete view of the action to be expected from theserelays.

Figure 1-5 Typical ac schematic for a switchboard-

mounted relay. (The connections are for the CR phase and

CRC ground directional time overcurrent relay of Figure 1-3.)

Figure 1-6 Typical dc schematic for a switchboard-

mounted relay. (The connections are for three phase type

CR and one CRC ground directional time overcurrent relays

of Figure 1-3 applied to trip a circuit breaker.)

Figure 1-7 Typical capacitor trip device schematic.


6 CIRCUIT-BREAKER CONTROL

Complete tripping and closing circuits for circuitbreakers are complex. A typical circuit diagram isshown in Figure 1-10. In this diagram, the protectiverelay circuits, such as that shown in Figure 1-6, areabbreviated to a single contact marked ‘‘prot relays.’’While the trip circuits must be energized from a sourceavailable during a fault (usually the station battery),the closing circuits may be operated on ac. Suchbreakers have control circuits similar to those shown inFigure 1-10, except that the 52X, 52Y, and 52CCcircuits are arranged for ac operation.

The scheme shown includes red light supervision ofthe trip coil, 52X/52Y antipump control, and low-pressure and latch checks that most breakers contain insome form.

Figure 1-8 A typical rack type relay. (The SBFU static

circuit breaker failure relay.)

Figure 1-9 Block diagram of MDAR relay.

8 Chapter 1

7 COMPARISON OF SYMBOLS

Various symbols are used throughout the world torepresent elements of the power system. Table 1-1compiles a few of the differences.

Figure 1-10 A typical control circuit schematic for a circuit

breaker showing the tripping and closing circuits.

Table 1-1 Comparison of Symbols

Element

U.S.

practice

European

practice

Normally open contact

Normally closed contact

Form C

Breaker

Fault

Current transformer

Transformer

Phase designations (typical) A,B,C

(preferred)

1, 2, 3

RST

Component designations

(positive, negative, zero)

1, 2, 0 1, 2, 0

Current I I

Voltage V U


2

Technical Tools of the Relay Engineer: Phasors, Polarity, andSymmetrical Components


1 INTRODUCTION

In addition to a general knowledge of electrical powersystems, the relay engineer must have a good workingunderstanding of phasors, polarity, and symmetricalcomponents, including voltage and current phasorsduring fault conditions. These technical tools are usedfor application, analysis, checking, and testing ofprotective relays and relay systems.

2 PHASORS

A phasor is a complex number used to representelectrical quantities. Originally called vectors, thequantities were renamed to avoid confusion with spacevectors. A phasor rotates with the passage of time andrepresents a sinusoidal quantity. A vector is stationaryin space.

In relaying, phasors and phasor diagrams are usedboth to aid in applying and connecting relays and forthe analysis of relay operation after faults.

Phasor diagrams must be accompanied by a circuitdiagram. If not, then such a circuit diagram must beobvious or assumed in order to interpret the phasordiagram. The phasor diagram shows only the magni-tude and relative phase angle of the currents andvoltages, whereas the circuit diagram illustrates onlythe location, direction, and polarity of the currents andvoltages. These distinctions are important. Confusiongenerally results when the circuit diagram is omitted orthe two diagrams are combined.

There are several systems and many variations ofphasor notation in use. The system outlined below isstandard with most relay manufacturers.

2.1 Circuit Diagram Notation for Current andFlux

The reference direction for the current or flux can beindicated by (1) an identified directional arrow in thecircuit diagram, as shown in Figure 2-1, or (2) thedouble subscript method, such as Iab, defined as thecurrent flowing from terminal a to terminal b, as inFigure 2-2.

In all cases, the directional arrow or doublesubscript indicates the actual or assumed direction ofcurrent (or flux) flow through the circuit during thepositive half-cycle of the ac wave.

Figure 2-1 Reference circuit diagram illustrating single

subscript notation.

11

2.2 Circuit Diagram Notation for Voltage

The relative polarity of an ac voltage may be shown inthe circuit diagram by (1) a þ mark at one end of thelocating arrow (Fig. 2-1) or (2) the double subscriptnotation (Fig. 2-2). In either case, the meaning of thenotation must be clearly understood. Failure toproperly define notation is the basis for muchconfusion among students and engineers.

The notation used in this text is defined as follows:

The letter ‘‘V’’ is used to designate voltages. Forsimplicity, only voltage drops are used. In thissense, a generator rise is considered a negativedrop. Some users assign the letter ‘‘E’’ togenerated voltage. In much of the world, ‘‘U’’is used for voltage.

If locating arrows are used for voltage in the circuitdiagram with a single subscript notation, theþmark at one end indicates the terminal of actualor assumed positive potential relative to the otherin the half-cycle.

If double subscript notation is used, the order of thesubscripts indicates the actual or assumed direc-tion of the voltage drop when the voltage is in thepositive half-cycle.

Thus, the voltage between terminals a and b may bewritten as either Vab or Eab. Voltage Vab or Eab ispositive if terminal a is at a higher potential thanterminal b when the ac wave is in the positive half-cycle. During the negative half-cycle of the ac wave,Vab or Eab is negative, and the actual drop for thathalf-cycle is from terminal b to terminal a.

2.3 Phasor Notation

Figure 2.3a demonstrates the relationship between aphasor and the sinusoid it represents. At a chosen time(in this instance at the time at which the phasor hasadvanced to 308), the instantaneous value of thesinusoid is the projection on the vertical of the pointof the phasor.

Phasors must be referred to some reference frame.The most common reference frame consists of the axisof real quantities x and the axis of imaginary quantitiesy, as shown in Figure 2-3b. The axes are fixed in theplane, and the phasors rotate, since they are sinusoidalquantities. (The convention for positive rotation iscounterclockwise.) The phasor diagram thereforerepresents the various phasors at any given commoninstant of time.

Theoretically, the length of a phasor is proportionalto its maximum value, with its projections on the realand imaginary axes representing its real and imaginarycomponents at that instant. By arbitrary convention,however, the phasor diagram is constructed on the

Figure 2-2 Reference circuit diagram illustrating double

subscript notation. (Current arrows not required but are

usually shown in practice.)

Figure 2-3a Phasor generation of sinusoid.

Figure 2-3b Reference axis and nomenclature for phasors.

12 Chapter 2

basis of rms values, which are used much morefrequently than maximum values. The phasor diagramindicates angular relationships under the chosenconditions, normal or abnormal.

For reference and review, the various forms, forrepresentation of point P in Figure 2-3b are as follows:

Rectan- Expo-

gular Complex nential Polar Phasor

form form form form form

aþ jb ¼ jcjðcos yþ j sin yÞ ¼ jcjejy ¼ jcjffy� ¼ c ð2-1Þa� jb ¼ jcjðcos y� j sin yÞ ¼ jcje�jy ¼ jcjff�y� ¼ cc ð2-2Þ

where

a ¼ real valueb ¼ imaginary valuejcj ¼ modulus or absolute value ðmagnitudeÞy ¼ argument or amplitude ðrelative positionÞ

If c is a phasor, then cc is its conjugate. Thus, if

c ¼ aþ jb

then

cc ¼ a� jb

Some references use c* to represent conjugate.The absolute value of the phasor is jcj:

jcj ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffia2 þ b2

pð2-3Þ

By adding Eqs. (2-1) and (2-2), we obtain

a ¼ 1

2ðcþ ccÞ ð2-4Þ

Substracting Eqs. (2-1) and (2-2) yields

jb ¼ 1

2ðc� ccÞ ð2-5Þ

In addition to the use of a single term such as c for aphasor, _cc; �cc, and c* have also been used.

2.3.1 Multiplication Law

The absolute value of a phasor product is the productof the absolute values of its components, and theargument is the sum of the component arguments:

EI ¼ jEj6jIjffðy1 þ y2Þ ð2-6Þor

EII ¼ jEjejy16jIje�jy2

¼ jEj6jIjffðy1 � y2Þ ð2-7Þ

2.3.2 Division Law

The division law is the inverse of multiplication:

E

I¼ jEjejy1

jIjejy2 ¼ jEjjIj ffðy1 � y2Þ ð2-8Þ

2.3.3 Powers of Complex Numbers

The product of a phasor times its conjugate is

ðjIjejyÞZ ¼ jIjZejZy ð2-9ÞThus, I2 equals jIj2ej2y:ffiffiffiffiffiffiffiffiffiffi

jIjejyZq

¼ffiffiffiffiffijIjZ

pejyZ

� �ð2-10Þ

The product of a phasor times its conjugate is

III ¼ jIjejy6jIje�jy

¼ jIj2ejðy�yÞ

¼ jIj2 ð2-11ÞOther reference axes used frequently are shown inFigure 2-4. Their application will be covered in laterchapters.

2.4 Phasor Diagram Notation

In Figure 2-5, the phasors all originate from a commonorigin. This method is preferred. In an alternativemethod, shown in Figure 2-6, the voltage phasors aremoved away from a common origin to illustrate thephasor addition of voltages in series (closed system).Although this diagram notation can be useful, it is not

Figure 2-4 Other reference axes for phasors used in relaying

and power systems.

Phasors, Polarity, and Symmetrical Components 13

generally recommended since it often promotes confu-sion by combining the circuit and phasor diagrams.

Notation for three-phase systems varies consider-ably in the United States; the phases are labeled a, b, cor A, B, C or 1, 2, 3. In other countries, thecorresponding phase designation of r, s, t is frequentlyused.

The letter designations are preferred and used here toavoid possible confusion with symmetrical componentsnotation. A typical three-phase system, with its separatecircuit and phasor diagrams, is shown in Figure 2-7.The alternative closed-system phasor diagram is shownin Figure 2-8. With this type of diagram, one tends tolabel the three corners of the triangle a, b, and c—thereby combining the circuit and phasor diagrams.The resulting confusion is apparent when one notesthat, with a at the top corner and b at the lower rightcorner, the voltage drop from a to b would indicate theopposite arrow from that shown on Vab.

However, when it is considered that, always,Vab ¼ Van þ Vnb, it is evident that Vab ¼ Van � Vbn.

Similarly, Vbc ¼ Vbn � Vcn, and Vca ¼ Vcn � Van. Theassociated phasors are shown in Figure 2-7.

Neutral (n) and ground (g) are often incorrectlyinterchanged. They are not the same. The voltage fromn to g is zero when no zero sequence voltage exists.With zero sequence current flowing, there will be avoltage between neutral and ground, Vng ¼ Vo.

Figure 2-5 Open-type phasor diagram for the basic

elements (resistor, reactor, and capacitor) connected in series.

Figure 2-6 Alternative closed-type phasor diagram for the

basic circuit of Figure 2-5.

Figure 2-7 Designation of the voltages and currents in a

three-phase power system.

Figure 2-8 Alternative closed system phasor diagram for

the three-phase power system of Figure 2-7.

14 Chapter 2

Ground impedance (Rg or RL) resulting in a rise instation ground potential can be an important factor inrelaying. This will be considered in later chapters.

According toANSI/IEEEStandard 100, ‘‘the neutralpoint of a system is that point which has the samepotential as the point of junction of a group of equalnonreactive resistances if connected at their free ends tothe appropriate main terminals or lines of the system.’’

2.5 Phase Rotation vs. Phasor Rotation

Phase rotation, or preferably phase sequence, is theorder in which successive phase phasors reach theirpositive maximum values. Phasor rotation is, byinternational convention, counterclockwise in direc-tion. Phase sequence is the order in which the phasorspass a fixed point.

All standard relay diagrams are for phase rotationa, b, c. It is not uncommon for power systems to haveone or more voltage levels with a, c, b rotation; thenspecific diagrams must be made accordingly. Theconnection can be changed from one rotation to theother by completely interchanging all b and c connec-tions.

3 POLARITY IN RELAY CIRCUITS

3.1 Polarity of Transformers

The polarity indications shown in Figures 2-9 and 2-10apply for both current and voltage transformers, orany type of transformer with either subtractive oradditive polarity.

The polarity marks X or &—— indicate

The current flowing out at the polarity-markedterminal on the secondary side is essentially inphase with the current flowing in at the polarity-marked terminal on the primary side.

The voltage drop from the polarity-marked to thenon-polarity-marked terminal on the primaryside is essentially in phase with the voltage dropfrom the polarity-marked to the non-polarity-marked terminals on the secondary side.

The expression ‘‘essentially in phase’’ allows for thesmall phase-angle error.

3.2 Polarity of Protective Relays

Polarity is always associated with directional-typerelay units, such as those indicating the direction ofpower flow. Other protective relays, such as distancetypes, may also have polarity markings associated withtheir operation. Relay polarity is indicated on theschematic or wiring diagrams by a small þ mark aboveor near the terminal symbol or relay winding. Twosuch marks are necessary; a mark on one windingalone has no meaning.

Typical polarity markings for a directional unit areshown in Figure 2-11. In this example, the markingsindicate that the relay will operate when the voltagedrop from polarity to nonpolarity in the voltage coil isin phase with the current flow from polarity tononpolarity in the current coil. This applies irrespec-tive of the maximum sensitivity angle of the relay. Ofcourse, the levels must be above the relay pickupquantities for the relay to operate.Figure 2-9 Polarity and circuit diagram for transformers.

Figure 2-10 Polarity and circuit diagram for conventional

representation of current and linear coupler transformer.


3.3 Characteristics of Directional Relays

Directional units are often used to supervise the actionof fault responsive devices such as overcurrent units.The primary function of the directional units is to limitrelay operation to a specified direction. These highlysensitive units operate on load in the tripping direction.

Directional units can conveniently serve to illustratethe practical application of phasors and polarity. Inaddition to polarity, these units have a phase-anglecharacteristic that must be understood if they are to beproperly connected to the power system. The char-acteristics discussed below are among the mostcommon.

3.3.1 Cylinder-Type Directional Unit

As shown in Figure 2-12, the cylinder-type unit hasmaximum torque when I, flowing in the relay windingfrom polarity to nonpolarity, leads V drop frompolarity to nonpolarity by 308. The relay minimumpickup values are normally specified at this maximumtorque angle. As current Ipq lags or leads thismaximum torque position, more current is required(at a constant voltage) to produce the same torque.Theoretically, at 1208 lead or 608 lag, no torque resultsfrom any current magnitude. In practice, however, this

zero torque line is a zone of no operation and not athin line through the origin, as commonly drawn.

3.3.2 Ground Directional Unit

As shown in Figure 2-13, the ground directional unitusually has a characteristic of maximum torque when Iflowing from polarity to nonpolarity lags V drop frompolarity to nonpolarity by 608. Although this char-acteristic may be inherent in the unit’s design, anauxiliary phase shifter is generally required in analogrelays.

3.3.3 Watt-Type Directional Unit

The characteristic of the watt-type unit is as shown inFigure 2-14. It has maximum sensitivity when relaycurrent and voltage are in phase.

Figure 2-11 Polarity markings for protective relays.

Figure 2-12 Phase-angle characteristics of the cylinder-type

directional relay unit.

Figure 2-13 Phase-angle characteristics of a ground direc-

tional relay unit.

Figure 2-14 Phase-angle characteristics of a watt-type

directional relay unit.

16 Chapter 2

3.4 Connections of Directional Units to Three-Phase Power Systems

The relay unit’s individual characteristic, as discussedso far, is the characteristic that would be measured ona single-phase test. Faults on three-phase powersystems can, however, produce various relationsbetween the voltages and currents. To ensure correctrelay operation, it is necessary to select the proper

quantities to apply to the directional units. For allfaults in the operating zone of the relay, the faultcurrent and voltage should produce an operatingcondition as close to maximum sensitivity as possible.Fault current generally lags its unity power factorposition by 20 to 858, depending on the system voltageand characteristics.

Four types of directional element connections (Fig.2-15) have been used for many years. The proper

Figure 2-15 Directional element connections.


system quantities are selected to yield the bestoperation, considering the phase-angle characteristicof the directional unit. A study of these connectionsreveals that none is perfect. All will provide incorrectoperation under some fault conditions. These condi-tions are, moreover, different for each connection.Fortunately, the probability of such fault conditionsoccurring in most power systems is usually very low.

For phase directional measurements, the standard908 connection is the one best suited to most powersystems. Here, the system quantities applied to the relayare 908 apart at unity power factor, balanced current.With this connection, maximum sensitivity can occur atvarious angles, depending on relay design, as inconnection 4. The 908 connection is one standard forphase relays. The 908 angle is that between the unitypower factor current and the voltage applied to the relay.

Some experts use a dual numbered system to describethe relationship of the system quantities and to identifythe nature of the relaying unit itself. For example, the90–608 connection is one in which the unity power-factor current applied to the relay and flowing in therelay trip direction leads the voltage applied to the relayby 908. The nature of the relay referred to is such thatthe maximum sensitivity occurs when the systemcurrent lags its unity power phase position by 608.The relay has its maximum sensitivity in this case whenthe current applied to it (into the polarity marker andout nonpolarity) leads the voltage applied to it (voltagedrop polarity to nonpolarity) by 308. Since this issomewhat confusing, it is recommended that the systemquantities that are applied to the relay be definedindependent of the characteristics of the relay, and thatthe characteristics of the relay be described independentof the system quantities with which it is used.

Figure 2-16 is a composite circuit diagram illustrat-ing the ‘‘phase-a’’ connections for these four connec-tions that have been used over the years, together withthe connection for a ground directional relay. Thephasor diagrams are shown in Figure 2-15a for thephase relays and Figure 2-17 for a commonly usedground relay.

4 FAULTS ON POWER SYSTEMS

A fault-proof power system is neither practical noreconomical. Modern power systems, constructed withas high an insulation level as practical, have sufficientflexibility so that one or more components may be outof service with minimum interruption of service. Inaddition to insulation failure, faults can result from

electrical, mechanical, and thermal failure or anycombination of these.

4.1 Fault Types and Causes

To ensure adequate protection, the conditions existingon a system during faults must be clearly understood.These abnormal conditions provide the discriminatingmeans for relay operation. The major types and causesof failure are listed in Table 2-1.

Relays must operate for several types of faults:

Three-phase (a-b-c, a-b-c-g)Phase-to-phase (a-b, b-c, c-a)Two-phase-to-ground (a-b-g, b-c-g, c-a-g)Phase-to-ground (a-g, b-g, c-g)

Unless preceded by or caused by a fault, opencircuits on power systems occur infrequently. Conse-quently, very few relay systems are designed specifi-cally to provide open-circuit protection. One exceptionis in the lower-voltage areas, where a fuse can be open.Another is in EHV, where breakers are equipped withindependent pole mechanisms.

Simultaneous faults in two parts of the system aregenerally impossible to relay properly under allconditions. If both simultaneous faults are in therelays’ operating zone, at least one set of relays is likelyto operate, with the subsequent sequential operationof other relays seeing the faults. When faults appearboth internal and external simultaneously, some relayshave difficulty determining whether to trip or not.Fortunately, simultaneous faults do not happen very

Table 2-1 Major Types and Causes of Failures

Type Cause

Insulation Design defects or errors

Improper manufacturing

Improper installation

Aging insulation

Contamination

Electrical Lightning surges

Switching surges

Dynamic overvoltages

Thermal Coolant failure

Overcurrent

Overvoltage

Ambient temperatures

Mechanical Overcurrent forces

Earthquake

Foreign object impact

Snow or ice

18 Chapter 2

Figure 2-16 Directional unit connections (phase ‘‘a’’ only) for four types of connections plus the ground directional relay

connections.


often and are not a significant cause of incorrectoperations.

4.2 Characteristics of Faults

4.2.1 Fault Angles

The power factor, or angle of the fault current, isdetermined for phase faults by the nature of the sourceand connected circuits up to the fault location and, forground faults, by the type of system grounding as well.The current will have an angle of 80 to 858 lag for aphase fault at or near generator units. The angle will beless out in the system, where lines are involved.

Typical open-wire transmission line angles are asfollows:

7.2 to 23 kV: 20 to 458 lag23 to 69 kV: 45 to 758 lag69 to 230 kV: 60 to 808 lag230 kV and up: 75 to 858 lag

At these voltage levels, the currents for phase faultswill have the angles shown where the line impedancepredominates. If the transformer and generator impe-dances predominate, the fault angles will be higher.Systems with cables will have lower angles if the cableimpedance is a large part of the total impedance to thefault.

4.2.2 System Grounding

System grounding significantly affects both the magni-tude and angle of ground faults. There are three classesof grounding: ungrounded (isolated neutral), impe-dance-grounded (resistance or reactance), and effec-tively grounded (neutral solidly grounded). Anungrounded system is connected to ground throughthe natural shunt capacitance, as illustrated in Figure2-18 (see also Chap. 7). In addition to load, small(usually negligible) charging currents flow normally.

In a symmetrical system, where the three capaci-tances to ground are equal, g equals n. If phase a isgrounded, the triangle shifts as shown in Figure 2-18.Consequently, Vbg and Vcg become approximately

ffiffiffi3

ptimes their normal value. In contrast, a ground on onephase of a solidly grounded radial system will result ina large phase and ground fault current, but little or noincrease in voltage on the unfaulted phases (Fig. 2-19).

4.2.3 Fault Resistance

Unless the fault is solid, an arc whose resistance varieswith the arc length and magnitude of the fault currentis usually drawn through air. Several studies indicatethat for currents in excess of 100A the voltage acrossthe arc is nearly constant at an average of approxi-mately 440V/ft.

Arc resistance is seldom an important factor inphase faults except at low system voltages. The arcdoes not elongate sufficiently for the phase spacingsinvolved to decrease the current flow materially. Inaddition, the arc resistance is at right angles to thereactance and, hence, may not greatly increase the totalimpedance that limits the fault current.

Figure 2-17 Phasor diagram for the ground directional

relay connection shown in Figure 2-16. (Phase ‘‘a’’-to-ground

fault is assumed on a solidly grounded system.)

Figure 2-18 Voltage plot for a solid phase ‘‘a’’-to-ground

fault on an ungrounded system.

Figure 2-19 Voltage plot for a solid phase ‘‘a’’-to-ground

fault on a solidly grounded system.

20 Chapter 2

For ground faults, arc resistance may be animportant factor because of the longer arcs that canoccur. Also, the relatively high tower footing resistancemay appreciably limit the fault current.

Arc resistance is discussed in more detail in Chapter12.

4.2.4 Distortion of Phases During Faults

The phasor diagrams in Figure 2-20 illustrate the effectof faults on the system voltages and currents. Thediagrams shown are for effectively grounded systems.In all cases, the dotted or uncollapsed voltage triangleexists in the source (the generator) and the maximumcollapse occurs at the fault location. The voltage at

other locations will be between these extremes,depending on the point of measurement.

5 SYMMETRICAL COMPONENTS

Relay application requires a knowledge of systemconditions during faults, including the magnitude,direction, and distribution of fault currents, and oftenthe voltages at the relay locations for various operatingconditions. Among the operating conditions to beconsidered are maximum and minimum generation,selected lines out, line-end faults with the adjacentbreaker open, and so forth. With this information, therelay engineer can select the proper relays and settingsto protect all parts of the power system in a minimumamount of time. Three-phase fault data are used forthe application and setting of phase relays and single-phase-to-ground fault data for ground relays.

The method of symmetrical components is thefoundation for obtaining and understanding faultdata on three-phase power systems. Formulated byDr. C. L. Fortescue in a classic AIEE paper in 1918,the symmetrical components method was given its firstpractical application to system fault analysis by C. F.Wagner and R. D. Evans in the late 1920s and early1930s. W. A. Lewis and E. L. Harder addedmeasurably to its development in the 1930s.

Today, fault studies are commonly made with thedigital computer and can be updated rapidly inresponse to system changes. Manual calculations arepractical only for simple cases.

A knowledge of symmetrical components is impor-tant in both making a study and understanding thedata obtained. It is also extremely valuable inanalyzing faults and relay operations. A number ofprotective relays are based on symmetrical compo-nents, so the method must be understood in order toapply these relays successfully.

In short, the method of symmetrical components isone of the relay engineer’s most powerful technicaltools. Although the method and mathematics are quitesimple, the practical value lies in the ability to thinkand visualize in symmetrical components. This skillrequires practice and experience.

5.1 Basic Concepts

The method of symmetrical components consistsof reducing any unbalanced three-phase system ofphasors into three balanced or symmetrical systems:

Figure 2-20 Phasor diagrams for the various types of faults

occurring on a typical power system.


the positive, negative, and zero sequence components.This reduction can be performed in terms of current,voltage, impedance, and so on.

The positive sequence components consist of threephasors equal in magnitude and 1208 out of phase (Fig.2-21a). The negative sequence components are threephasors equal in magnitude, displaced 1208 with aphase sequence opposite to that of the positivesequence (Fig. 2-21b). The zero sequence componentsconsist of three phasors equal in magnitude and inphase (Fig. 2-21c). Note all phasors rotate in acounterclockwise direction.

In the following discussion, the subscript 1 willidentify the positive sequence component, the subscript2 the negative sequence component, and the subscript 0the zero sequence component. For example, Va1 is thepositive sequence component of phase-a voltage, Vb2

the negative sequence component of phase-b voltage,and Vc0 the zero sequence component of phase-cvoltage. All components are phasor quantities, rotatingcounterclockwise.

Since the three phasors in any set are always equalin magnitude, the three sets can be expressed in termsof one phasor. For convenience, the phase-a phasor isused as a reference. Thus,

Positive

sequence

Va1 ¼ Va1

Vb1 ¼ a2Va1

Vc1 ¼ aVa1

Negative

sequence

Va2 ¼ Va2

Vb2 ¼ aVa2

Vc2 ¼ a2Va2

Zero

sequence

Va0 ¼ Va0

Vb0 ¼ Va0

Vc0 ¼ Va0

ð2-12Þ

The coefficients a and a2 are operators that, whenmultiplied with a phasor, result in a counterclockwiseangular shift of 120 and 2408, respectively, with no

change in magnitude:

a ¼ 1ff120�¼ �0:5þ j0:866 ð2-13Þ

a2 ¼ 1ff240�¼ �0:5� j0:866 ð2-14Þ

a3 ¼ 1ff360�¼ 1:0þ j0 ð2-15Þ

From these equations, useful combinations can bederived

1þ aþ a2 ¼ 0

1� a2 ¼ffiffiffi3

pff30� ð2-16Þ

or

a2 � 1 ¼ffiffiffi3

pff210�

a� 1 ¼ffiffiffi3

pff150� ð2-17Þ

or

1� a ¼ffiffiffi3

pff � 30�

a2 � a ¼ffiffiffi3

pff270� ð2-18Þ

or

a� a2 ¼ffiffiffi3

pff90� ð2-19Þ

Any three-phase system of phasors will always bethe sum of the three components:

Va ¼ Va1 þ Va2 þ Va0 ð2-20ÞVb ¼ Vb1 þ Vb2 þ Vb0

¼ a2Va1 þ aVa2 þ Va0 ð2-21ÞVc ¼ Vc1 þ Vc2 þ Vc0

¼ aVa1 þ a2Va2 þ Va0 ð2-22Þ

Since phase a has been chosen as a reference, thesubscripts are often dropped for convenience. Thus,

Va ¼ V1 þ V2 þ V0

and

Ia ¼ I1 þ I2 þ I0

Vb ¼ a2V1 þ aV2 þ V0

ð2-23Þ

and

Ib ¼ a2I1 þ aI2 þ I0

Vc ¼ aV1 þ a2V2 þ V0

ð2-24ÞFigure 2-21 Sequence components of voltages.

22 Chapter 2

and

Ic ¼ aI1 þ a2I2 þ I0 ð2-25ÞQuantities V1;V2;V0; I1; I2, and I0, can always beassumed to be the phase-a components. Note that theb and c components always exist, as indicated by Eq.(2-12). Note that dropping the phase subscripts shouldbe done with great care. Where any possibility ofmisunderstanding can occur, the additional effort ofusing the double subscripts will be rewarded.

Equations (2-20) to (2-22) can be solved to yield thesequence components for a general set of three-phasephasors:

Va1 ¼ 1

3ðVag þ aVbg þ a2VcgÞ

and

Ia1 ¼ 1

3ðIa þ aIb þ a2IcÞ

Va2 ¼ 1

3ðVag þ a2Vbg þ aVcgÞ ð2-26Þ

and

Ia2 ¼ 1

3ðIa þ a2Ib þ aIcÞ

Va0 ¼ 1

3ðVag þ Vbg þ VcgÞ ð2-27Þ

and

I0 ¼ 1

3ðIa þ Ib þ IcÞ ð2-28Þ

A sequence component cannot exist in only onephase. If any sequence component exists by measure-ment or calculation in one phase, it exists in all threephases, as shown in Eq. (2-12) and Figure 2-21.

5.2 System Neutral

Figure 2-22 describes the definition of power-systemneutral and contrasts it with ground. Neutral isestablished by connecting together the terminals ofthree equal resistances as shown with each of the otherresistor terminals connected to one of the phases. Wecan thus write

Vag ¼ Van þ Vng

Vbg ¼ Vbn þ Vng

Vcg ¼ Vcn þ Vng ð2-29Þ

From Eq. (2-28),

V0 ¼ 1

3ðVag þ Vbg þ VcgÞ

Substituting Eq. (2-29), we obtain

V0 ¼ 1

3ðVan þ Vng þ Vbn þ Vng þ Vcn þ VngÞ

Since Van þ Vbn þ Vcn ¼ 0,

V0 ¼ 1

3ð3VngÞ

V0 ¼ Vng

Neutral and ground are distinctly independent anddiffer in voltage by V0.

Grounding and its influence on relaying arediscussed in Chapters 7 and 12.

5.3 Sequences in a Three-Phase Power System

Several important assumptions are made to greatlysimplify the use of symmetrical components inpractical circumstances. Interconnections of the threesequence networks allow any series or shunt disconti-nuity to be investigated. For the rest of the power-system network, it is assumed that the impedances inthe individual phases are equal and the generatorphase voltages are equal in magnitude and displaced1208 from one another.

Based on this premise, in the symmetrical part of thesystem, positive sequence current flow produces onlypositive sequence voltage drops, negative sequencecurrent flow produces only negative sequence voltagedrops, and zero sequence current flow produces onlyzero sequence voltage drops. For an unsymmetricalsystem, interaction occurs between components. For aparticular series or shunt discontinuity being repre-

Figure 2-22 Power system neutral.


sented, the interconnection of the networks producesthe required interaction.

Any circuit that is not continuously transposed willhave impedances in the individual phases that differ.This fact is generally ignored in making calculationsbecause of the immense simplification that results.From a practical viewpoint, ignoring this effect, ingeneral, has no appreciable influence.

5.4 Sequence Impedances

Quantities Z1, Z2, and Z0 are the system impedances tothe flow of positive, negative, and zero sequencecurrents, respectively. Except in the area of a fault orgeneral unbalance, each sequence impedance is con-sidered to be the same in all three phases of thesymmetrical system. A brief review of these quantitiesis given below for synchronous machinery, transfor-mers, and transmission lines.

5.4.1 Synchronous Machinery

Three different positive sequence reactance values arespecified. X00

d indicates the subtransient reactance, X0d

the transient reactance, and Xd the synchronousreactance. These direct-axis values are necessary forcalculating the short-circuit current value at differenttimes after the short circuit occurs. Since the sub-transient reactance values give the highest initialcurrent value, they are generally used in systemshort-circuit calculations for high-speed relay applica-tion. The transient reactance value is used for stabilityconsideration and slow-speed relay application.

The unsaturated synchronous reactance is used forsustained fault-current calculation since the voltage isreduced below saturation during faults near the unit.Since this generator reactance is invariably greaterthan 100%, the sustained fault current will be less thanthe machine rated load current unless the voltageregulator boosts the field substantially.

The negative sequence reactance of a turbinegenerator is generally equal to the subtransient X00

d

reactance. X2 for a salient-pole generator is muchhigher. The flow of negative sequence current ofopposite phase rotation through the machine statorwinding produces a double frequency component inthe rotor. As a result, the average of the subtransientdirect-axis reactance and the subtransient quadrature-axis reactance gives a good approximation of negativesequence reactance.

The zero sequence reactance is much less than theothers, producing a phase-to-ground fault current

magnitude ½3=ðx1 þ x2 þ x0Þ� greater than the three-phase fault current magnitude ð1=x1Þ. Since themachine is braced for only three-phase fault currentmagnitude, it is seldom possible or desirable to groundthe neutral solidly.

The armature winding resistance is small enough tobe neglected in calculating short-circuit currents. Thisresistance is, however, important in determining the dctime constant of an asymmetrical short-circuit current.

Typical reactance values for synchronous machin-ery are available from the manufacturer or handbooks.However, actual design values should be used whenavailable.

5.4.2 Transformers

The positive and negative sequence reactances of alltransformers are identical. Values are available fromthe nameplate. The zero sequence reactance is eitherequal to the other two sequence reactances or infiniteexcept for the three-phase, core-type transformers. Ineffect, the magnetic circuit design of the latter unitsgives them the effect of an additional closed deltawinding. The resistance of the windings is very smalland neglected in short-circuit calculations.

The sequence circuits for a number of transformerbanks are shown in Figure 2-23. The impedancesindicated are the equivalent leakage impedancesbetween the windings involved. For two-windingtransformers, the total leakage impedance ZLH ismeasured from theL winding, with the H windingshort-circuited. ZHL is measured from the H windingwith the L winding shorted. Except for a 1:1 transfor-mer ratio, the impedances have different values inohms. On a per unit basis, however, ZLH equals ZHL.

For three-winding and autotransformer banks,there are three leakage impedances:

Impedance

Winding

measured

from

Shorted

winding

Open

winding

ZHM(ZHL) H M(L) L(T)

ZHL(ZHT) H L(T) M(L)

ZML(ZLT) M(L) L(T) H

Both winding conventions shown above are incommon use. In the first convention, the windingsare labeled H (high), L (low), M (medium); in thesecond H (high), L (low), and T (tertiary). Unfortu-nately, the L winding in the second convention is

24 Chapter 2

equivalent toM in the first. The tertiary windingvoltage is generally the lowest.

On a common kVA base, the equivalent wyeleakage impedances are obtained from the followingequations:

ZH ¼ 1

2ðZHM þ ZHL � ZMLÞ

or

ZH ¼ 1

2ðZHL þ ZHT � ZLTÞ

ZM ¼ 1

2ðZHM þ ZML � ZHLÞ

or

ZL ¼ 1

2ðZHL þ ZLT � ZHTÞ

ZL ¼ 1

2ðZHL þ ZML � ZHMÞ

or

ZT ¼ 1

2ðZHT þ ZLT � ZHLÞ

As a check, ZH plus ZM equals ZHM, and so on. Thewye is a mathematical equivalent valid for current andvoltage calculations external to the transformer bank.The junction point of the wye has no physicalsignificance. One equivalent branch, usually ZMðZLÞ,

Figure 2-23 Equivalent positive, negative, and zero sequence circuits for some common and theoretical connections for two-

and three-winding transformers.


may be negative. On some autotransformers, ZH isnegative.

The equivalent diagrams shown in Figure 2-23 aresatisfactory when calculations are to be made relativeto one segment of a power system. However, a morecomplex representation is required when phase cur-rents and voltages are to be determined at points in thesystem having an intervening transformer betweenthem and the point of discontinuity being examined.For delta-wye transformers, a 308 phase shift must beaccommodated. For ANSI standard transformers, thehigh-voltage phase-to-ground voltage leads the low-voltage phase-to-ground voltage by 308, irrespective ofwhich side the delta or wye is on. This phase shift maybe included in the equivalent per unit diagram byshowing a 1 ff30�:1 ratio for it.

The phase shift in the negative sequence network forthe delta-wye transformer is the same amount, but inthe opposite direction, to that in the positive sequencenetwork. The phase shift then, for an ANSI standardtransformer, would be 1 ff � 30�:1 in the negativesequence per unit diagram.

The phase shift must be used in all the combinationsof Figure 2-23 where a wye and delta winding coexist.This effect is extremely important when considerationis being given to the behavior of devices on both sidesof such a transformer.

5.4.3 Transmission Lines

For transmission lines, the positive and negativesequence reactances are the same. As a rule of thumb,the 60-Hz reactance is roughly 0:8O=mi for singleconductor overhead lines and 0:6O=mi for bundledoverhead lines.

The zero sequence impedance is always differentfrom the positive and negative sequence impedances. Itis a loop impedance (conductor plus earth and/orground wire return), in contrast to the one-way

impedance for a positive and negative sequence. Zerosequence impedance can vary from 2 to 6 times X1; arough average for overhead lines is 3 to 3.5 times X1.

The resistance terms for the three sequences areusually neglected for overhead lines, except for lower-voltage lines and cables. In the latter cases, line anglesof 30 to 608may exist, and resistance can be significant.A good compromise is to use the impedance valuerather than reactance and neglect the angular differ-ence in fault calculations. This yields a lower current toassure that the relay will be set sensitively enough.

Zero sequence mutual impedance resulting fromparalleled lines can be as high as 50 to 70% of the zerosequence self-impedance. This mutual impedancebecomes an increasingly important factor as morelines are crowded into common rights of way.

5.5 Sequence Networks

With the system assumed to be balanced or symme-trical to the point of unbalance or fault, the threesequence components are independent and do notreact with each other. Thus, three network diagramsare required to separate the three sequence compo-nents for individual consideration: one for positive,one for negative, and one for zero sequence. Thesesequence network diagrams consist of one phase toneutral of the power system, showing all the compo-nent parts relevant to the problem under considera-tion. Typical diagrams are illustrated in Figures 2-24through 2-26.

The positive sequence network (Fig. 2-24) mustshow both the generator voltages and impedances ofthe generators, transformers, and lines. Balanced loadsmay be shown from any bus to the neutral bus.Generally, however, balanced loads are neglected.Compared to the system low-impedance high-anglequantities, they have a much higher impedance at a

Figure 2-24 Example system and positive sequence network.

26 Chapter 2

very low angle. In short, balanced loads complicate thecalculations and generally do not affect the faultcurrents significantly.

With two exceptions, the negative sequence network(Fig. 2-25) will be a duplicate of the positive sequencenetwork: (1) There will be no generator voltages, sincesynchronous machines generate a positive sequenceonly, and (2) the negative sequence reactance ofsynchronous machinery may be different from thepositive, as previously described. For all practicalcalculations involving faults or discontinuities remotefrom the generating plant, however, X1 is assumed tobe equal to X2.

The zero sequence network (Fig. 2-26) is quitedifferent from the other two. First of all, it has novoltage: Rotating machinery does not produce zerosequence voltage. Also, the transformer connectionsrequire special consideration and grounding impe-dances must be included. Figure 2-23 shows the zerosequence circuits for many transformers.

A three-line system diagram is usually not requiredto determine the zero sequence network, but if aquestion arises as to the flow of zero sequence currents,the three-line diagram can be useful. From this three-phase system diagram, the zero sequence networkrequirements can be resolved by determining whetheror not equal and in-phase currents can exist in each ofthe three phases. If the zero sequence currentcomponent can flow, the zero sequence network mustreflect its path.

For simplicity, Figure 2-27 shows the generatorssolidly grounded. In practice, however, solid ground-ing is used only in very special cases.

5.6 Sequence Network Connections andVoltages

The current flow direction and voltage connectionsillustrated in Figure 2-28 must be followed for Eqs.(2-29), (2-30), and (2-31) to apply. Current referencedirection in any circuit element must be the same in allthree networks to avoid confusion. Current flow in oneor more of the networks may reverse for some types ofunbalances, particularly if the networks are complex.Reverse flow should be treated as a negative current toensure that it will be properly subtracted whendetermining the phase currents.

Each sequence network is, of course, a per unitdiagram representing one of the three phases of thesymmetrical power system. Therefore, a resistor(reactor, impedance) connected between the systemneutral and ground, as shown in Figure 2-28, must bemultiplied by 3 as indicated. In the system, 3I0 flowsthrough R; in the zero sequence network, however, I0flows through 3R, producing an equivalent voltagedrop.

Figure 2-25 Negative sequence network for example

system.

Figure 2-26 Zero sequence network for example system.

Figure 2-27 Example system (generators shown solidly grounded for simplification).


5.7 Network Connections for Fault and GeneralUnbalances

The sequence networks can be interconnected at apoint of discontinuity, such as a fault. In such areas,negative and zero sequence voltages are generated, aspreviously described. Sequence network connectionsfor various types of common faults are shown inFigures 2-29 through 2-32. From the three-phasediagrams of the fault area, the sequence networkconnections representing the fault can be derived.These diagrams do not show fault impedance, andfault studies do not include this effect except in veryrare cases. The single-sequence impedance Z1;Z2;Z0

(practically equivalent to X1;X2;X0) shown in the

figures is the net impedance between the neutral busand selected fault location. Based on zero load, allgenerated voltages (VAN) are equal and in phase.

Since the three-phase fault is balanced, symmetricalcomponents are not required for this calculation.However, since the positive sequence network repre-sents the system, the network can be connected asshown in Figure 2-29 to represent the fault.

For a phase-a-to-ground fault, the three networksare connected in series (Fig. 2-30). Figure 2-31illustrates a phase-b-c-to-ground fault and its sequencenetwork interconnection. The phase-b-to-phase-c faultand its sequence connections are shown in Figure 2-32.

Fault studies normally include only three-phasefaults and single-phase-to-ground faults. Three-phasefaults are the most severe phase faults, whereas single-phase-to-ground faults are the most common. Studiesof the latter faults provide useful information forground relaying.

A fundamental study of both series and shuntunbalances was made by E. L. Harder in 1937. Theshunt unbalances summarized in Figure 2-33 are takenfrom Harder’s study. Note that all the faults shown inFigures 2-29 through 2-32 are also represented inFigure 2-33.

Figure 2-28 Sequence network connections and voltages.

Figure 2-29 Three-phase fault and its network connection.

Figure 2-30 Phase-to-ground fault and its sequence net-

work connections.

28 Chapter 2

In Figure 2-33, the entire symmetrical power systemup to a point x of the shunt connection is represented bya rectangular box. Inside the topmost box for each shuntcondition is a four-line representation of the shunt to beconnected to the systemat point x.The three lower boxesfor each shunt condition are the positive, negative, andzero sequence representations of the shunt.

The sequence connections for the series unbalances,such as open phases and unbalanced series impe-dances, are shown in Figure 2-34. As before, thesediagrams are taken from E. L. Harder’s study. Hereagain, the diagrams inside the topmost box for eachseries condition represent the area under study, frompoint x on the diagrams left to point y on the right. Thepower system represented by the box is open between xand y to insert the circuits shown inside the box. Pointsx and y can be any distance apart, as long as there is notap or other system connection between them. The

positive, negative, and zero sequence interconnectionsfor the discontinuity shown in the top box areillustrated in the three boxes below it.

Simultaneous faults require two sets of interconnec-tions from either Figures 2-33 or 2-34 or both. Asshown in Figure 2-35, ideal or perfect transformers canbe used to isolate the two restrictions. Perfecttransformers are 100% efficient and have ratios of1:1, 1:a, 1:a2.

It is sometimes necessary to use two transformers asshown in Figure 2-35f. In this case, the first transfor-mer (ratios 1 :e�j30� , 1 :ej30

�, and 1:1) represents the

wye-delta transformer, and the second transformer(ratios 1:a2, 1:a, 1:1) represents the b-to-neutral fault.These can be replaced by an equivalent transformerwith ratios 1 : e�j150� , 1 :ej150

�, and 1:1.

Figure 2-35a, for example, represents an openphase-a conductor with a simultaneous fault to groundon the x side. The sequence networks are connected forthe open conductor according to Figure 2-34j, withthree 1:1 perfect transformers providing the restric-tions required by Figure 2-33f. The manual calcula-tions required, which involve the solution ofsimultaneous equations, may be quite tedious.

5.8 Sequence Network Reduction

When manual calculations are performed, the com-plete system networks are reduced to the singleimpedance values of Figures 2-29 through 2-32.

To simplify this reduction, with negligible effect onthe results, the following basic assumptions are some-times made:

All generated voltages are equal and in phase.All resistance is neglected, or the reactance of

machines and transformers is added directlywith line impedances.

All shunt reactances are neglected, including loads,charging, and magnetizing reactances.

All mutual reactances are neglected, except onparallel lines.

By using these assumptions, the positive sequencenetwork can be drawn with a single-source voltage Van

connected to the generator impedances by a bus.If voltages are different, either the voltages must be

retained in the network or Thevenin theorem andsuperposition must be used to reduce the network andcalculate fault currents and voltages. Note that for theseries unbalances of Figure 2-34, a difference in

Figure 2-31 Double phase-to-ground fault and its sequence

network connections.

Figure 2-32 Phase-to-phase fault and its sequence network

connections.


voltage—either magnitude, phase angle, or both—isrequired for current to flow.

The single-sequence impedances Z1, Z2 and Z0 ofFigures 2-29 through 2-32 will be different for eachfault location because of the different network reduc-tions. During the network reduction, the distribution ofcurrents in the various branches should be calculated,both as a check and to determine the current flowthrough the relays involved in a fault. These distribu-tion factors are calculated with the assumption that 1

per unit current flows in these single-sequence impe-dances at the fault or point of discontinuity.

Network reduction calculations for the system ofFigure 2-24 are illustrated in Figures 2-36, 2-37, and 2-38. In these figures, X1, X2, and X0 are the impedancesbetween the neutral bus and the fault at bus G. I1R, I1L,I2R, I2L, I0R, I0L are the per unit distribution factors. I1,I2, and I0 are all assumed to be equal to 1 per unit.

Analog or digital studies should be tailored toproduce outputs that allow each branch current in

Figure 2-33 Sequence network interconnections for shunt balanced and unbalanced conditions.

30 Chapter 2

each network to be identified. For single-phase-to-ground faults, 3I0 is required for relays.

When using the computer for sequence networkreduction, the impedance data are input for thepositive and zero sequence networks, along with busand fault node points. The network is then solved forthree-phase and single-phase-to-ground faults. Tabu-lated printed data are provided for phase-a faultcurrent and three-phase fault voltages, along with thecorresponding 3I0, 3V0 values for the phase-to-groundfault. I2 and V2 values should also be obtained fornegative sequence relays.

These voltage and current values are needed for notonly faults near the relay, but also those several buses orlines away. Among the operating conditions normallyconsidered are maximum and minimum generation,selected lines out of service, and line-end faults where theadjacent breaker is open. This information allows the

correct relay types and settings to be selected in aminimal amount of time for the entire power system.

The following steps must be performed for calculat-ing fault currents and voltages:

Obtain a complete single-line diagram for the entiresystem, including generators, transformers, andtransmission lines, along with the positive,negative, and zero sequence impedances foreach component.

Prepare a single-line impedance diagram from thesystem diagram or establish the nodes in a digitalstudy for the positive, negative, and zerosequence networks.

Reduce the impedance values of all networkbranches to a common base. Values may beexpressed as per unit on a common kVA base oras ohms impedance on a common voltage base.

Figure 2-34 Sequence network interconnections for series balanced and unbalanced conditions.


Obtain, or have the computer obtain, the equivalentsingle impedance of each sequence network,current distribution factors, and equivalent sourcevoltage for the positive-phase sequence network.All quantities must be referred to the proper base.

Interconnect the networks or utilize the computerprogram to represent the fault type involved, andcalculate the total fault current at the fault.

Determine the current distribution and voltages asrequired in the system. Total fault current is

seldom of use as relays generally see a fraction ofthat current except for radial circuits.

5.9 Example of Fault Calculation on a Loop-Type Power System

For the typical loop system shown in Figure 2-39, thegenerator units at stations D, S, or E could each becombinations of several machines. Alternatively, they

Figure 2-35 Representations for simultaneous unbalances.

32 Chapter 2

could represent the equivalent of a complex system upto the bus. All the impedances have been reduced to acommon base, as indicated in the diagram. Thepositive sequence network for this system is shown inFigure 2-40, the zero sequence network in Figure 2-41.The negative sequence network is equal to Figure 2-40,except that Van is not present.

To perform this sample calculation of a phase-to-ground fault on the bus at station D, the networksmust be reduced to a single reactance value betweenthe neutral bus and fault point. Of the several delta

loops, at least one must be converted to wye-equivalentin order to reduce the networks. After one loop ischosen (arbitrarily), the equivalent X, Y, Z branchesfor an equivalent wye are dotted in as shown in Figures2-40 and 2-41.

The X, Y, Z conversion from delta to wye-equivalent is a simple process: The X branch of thewye-equivalent is the product of the two deltareactances on either side divided by the sum of thethree delta impedances. The same relation applies to

Figure 2-36 Network reduction for example system

(Figure 2-24) fault at bus G.

Figure 2-37 Network reduction and current distribution.

Figure 2-38 Final network reduction for fault at bus G in

Figure 2-24.

Figure 2-39 Single line diagram for a typical loop-type

power system.


the Y and Z branches. Thus, in Figures 2-40 and 2-41,the networks are reduced as follows:

Positive and negative

sequence networks

Zero sequence

network

X1 ¼ 2462862

¼ j10:84 X0 ¼ 96675

¼ j0:72

Y1 ¼ 2861062

¼ j4:52 Y0 ¼ 666075

¼ j4:8

Z1 ¼ 2461062

¼ j3:87 Z0 ¼ 966075

¼ j7:2

The networks now reduce to the simpler forms shownin Figure 2-40c. Since the two upper branches ofeach network are in parallel, they can be reduced asfollows:


sequence networks

Zero sequence

network

0:4716 0:5284 0:2594 0:7406

44:52649:8794:39

56:86162:2219:0

¼ 23:52 ¼ 42:07

Figure 2-40 Positive sequence network reduction for the

system of Figure 2-39.

Figure 2-41 Zero sequence network reduction for the


34 Chapter 2

These reductions are shown in Figures 2-40d and2-41c. The remaining branches are in parallel and canalso be reduced:


sequence networks Zero sequence network

0:4621 0:5379 0:8106 0:1894

X1 ¼ X2 ¼ 34:3664074:36

X0 ¼ 42:7961052:79

¼ j18:48% ¼ j8:11%

The numbers written above the equations are thedistribution factors for the parallel circuits. Thesefactors are expressed as the ratio of each term in thenumerator and denominator. Determining these fac-tors provides a convenient check on the calculations,since the sum of the two fractions must be 1.

Distribution factors can be determined by workingback through the reduction. The factors should bewritten on the diagrams as shown in Figure 2-42.

The distribution factors for the upper parallelbranches of Figure 2-40c are determined as follows:

Positive sequence network

44:52% branch : 0:528460:5379 ¼ 0:2842

49:87% branch : 0:471660:5379 ¼ 0:2537

0:5379 ðcheckÞ

The distribution factors in the zero sequencenetwork are

Zero sequence network

162:2% branch : 0:259460:1894 ¼ 0:0491

56:8% branch : 0:740660:1895 ¼ 0:1403

0:1894 ðcheckÞ

In turn, these distribution factors are added to thediagram, as shown in Figure 2-42b.

The delta current distribution factors are obtainedfrom the X, Y, Z equivalents. The conversiontechnique is straightforward: The voltage drop acrosstwo of the wye branches is equivalent to the dropacross the delta branch. Calculating from Figure 2-40c,we obtain

Positive sequence network

0:53796j10:84þ 0:28426j4:52

j28¼ 0:2541

0:53796j10:84þ 0:25376j3:87

j24¼ 0:2838

�0:2537þ j3:87þ 0:28426j4:52

j10¼ 0:0301

Zero sequence network

0:18946j0:72þ 0:14036j4:52

j6¼ 0:1350

0:18946j0:72þ 0:04916j7:20

j24¼ 0:0544

�0:0491þ j7:20þ 0:14036j4:80

j60¼ 0:0053

Figure 2-42 shows the complete per unit distributionfor the original network of Figure 2-39.

The three networks are connected in series for thephase-to-ground fault (Fig. 2-28). For convenience, thesequence currents are calculated in per unit values:

Figure 2-42 Per unit current distribution for AG fault at D.


I1 ¼ I2 ¼ I0

¼ j1:0

j0:1848þ j0:1848þ j0:0811

¼ 1:0

0:4507

¼ 2:22 p:u:

The 100% (1 p.u. base) current is

IB ¼ kVA baseffiffiffi3

pkV

¼ 100;000ffiffiffi3

p6110

¼ 524:86A at 110 kV

I1 ¼ I2 ¼ I0

¼ 2:226524:86

¼ 1164:55A at 110 kV

The current flowing in each branch of the networks cannow be determined by multiplying the actual fault cur-rent by the distribution factor. These currents may beexpressed in either per unit or ampere values. Currentsin the fault are calculated for each phase as follows:

Ia ¼ 3I1 ¼ 3I2 ¼ 3I0 ¼ 6:66 p:u:

or

Ia ¼ 3493:66A110 kV

Ib ¼ ða2I1 þ aI2 þ I0Þ¼ ð�I1 þ I0Þ ¼ 0

Ic ¼ ðaI1 þ a2I2 þ I0Þ¼ ð�I1 þ I0Þ ¼ 0

For each branch, the per unit positive, negative, andzero sequence currents can then be used to determinethe individual phase currents by using Eqs. (2-23),(2-24), and (2-25). These are recorded in Figure 2-43.

Next, the sequence and phase voltages at each bus aredetermined as inFigure 2-28. It is convenient to calculatethe voltages in per unit values. Note that the impedanceslisted in Figure 2-39 appear in percent, rather thanohms, and may be converted easily to per unit.

In the following calculations, the values in parenth-eses are volts, converted from the per unit values forthe 110-kV system of Figure 2-39:

Vline-to-neutral ¼ 1:0 p:u:

¼ 110;000Vffiffiffi3

p

¼ 63;508:53V

From Figure 2-42, first the sequence and phasevoltages are calculated at bus S:

V1 ¼ j1:0� 0:62976j0:24

¼ j1:0� j0:1511

¼ j0:8489 p:u: ð53;912:39VÞV2 ¼ 0� 0:62976j0:24

¼ �j0:1511 p:u: ð9596:14VÞV0 ¼ 0� 0:12076j0:09

¼ �j0:0109 p:u: ð692:24VÞVag ¼ V1 þ V2 þ V0

¼ j0:6869 p:u: ð43;624:01VÞVbg ¼ a2V1 þ aV2 þ V0

¼ 0:8489ff � 30� þ 0:1511ff þ 30� � j0:0109

¼ 0:7352� j0:4245þ 0:1309þ j0:0756� j0:0109

¼ 0:8661� j0:3598

¼ 0:9379ff � 22:56� p:u: ð59;594:65VÞVcg ¼ aV1 þ a2V2 þ V0

¼ 0:8489ff210� þ 0:1511ff150� � j0:0109

¼ �0:7352� j0:4245� 0:1309þ j0:0756� j0:0109

¼ �0:8661� 0:3598

¼ 0:9379ff202:56� p:u: ð59;564:65VÞ

Next, the sequence and phase voltages are calculated atbus D, the fault location:

V1 ¼ j1:0� 1:02536j0:40

¼ j1:0� j0:4101

¼ j0:5899 p:u: ð37;463:68VÞV2 ¼ 0� 1:02536j0:40

¼ �j0:4101 p:u: ð26;044:85VÞV0 ¼ 0� 1:79866j0:1

¼ �j0:1798 p:u: ð11;418:83VÞVag ¼ 0

Vbg ¼ 0:5899ff � 30� þ 0:4101ff30� � j0:1798

¼ 0:5109� j0:2950þ 0:3552þ j0:2051� j0:1798

¼ 0:8661� j0:2697

¼ 0:9071ff � 17:30� p:u: ð57;608:59VÞVcg ¼ 0:5899ff210� þ 0:4101ff150� � j0:1798

¼ 0:5109� j0:2950� 0:3552þ j0:2051� j0:1798

¼ �0:8661� j0:2697

¼ 0:9071ff197:30� p:u: ð57;608:59VÞ

Similarly, the sequence and phase voltages can be

36 Chapter 2

calculated at bus E:

Vag ¼ j0:6352 p:u: ð40;340:62VÞVbg ¼ 0:9502ff � 24:30� p:u: ð60;345:80VÞVcg ¼ 0:9502ff204:30� p:u: ð60;345:80VÞ

Finally, the voltages are calculated at bus R:

Vag ¼ j0:3646 p:u: ð23;155:21VÞVbg ¼ 0:8909ff13:59� p:u: ð56;579:75VÞVcg ¼ 0:8909ff193:59� p:u: ð56;579:75VÞ

The sequence voltages calculated above, as shown inFigure 2-43, complete the analysis of the single-phase-to-ground fault at bus D in the system of Figure 2-39.All the distributed current and voltage values for thesystem are displayed in Figure 2-43.

5.10 Phase Shifts Through Transformer Banks

In these fault calculations, the phase shifts through thewye-delta transformer banks were not considered. Inthis example, only a 110-kV system fault, with itscurrents and voltages, was involved. The effect of thephase shift through the transformer banks could not,however, have been neglected if currents and voltageswere required for the opposite side of the powertransformers.

If the transformer bank is wye-connected on thehigh-voltage side, as shown in Figure 2-44, the generalequations for one phase are

IA ¼ nðIa � IcÞ ð2-30ÞVan ¼ nðVAn � VBnÞ ¼ nVAB ð2-31Þ

The lowercase subscripts represent high-side quantitiesand the capital letter subscripts low-side quantities. In

Figure 2-43 Current and voltage distribution for a single phase-to-ground fault at bus ‘‘D’’ of the system of Figure 2-37.


the balanced or symmetrical transformer bank, thesequences are independent.

Consequently, positive sequence only is first appliedto Eqs. (2-30) and (2-31):

IA1 ¼ nðIa1 � Ic1Þ¼ nðIa1 � aIa1Þ¼ nð1� aÞIa1¼ n

ffiffiffi3

pIa1ff � 30�

IA1 ¼ NIA1ff � 30� ð2-32ÞIa1 ¼ IA1

Nff30�

Va1 ¼ nðVA1 � VB1Þ¼ nðVA1 � a2VA1Þ ð2-33Þ¼ nð1� a2ÞVA1

Va1 ¼ nffiffiffi3

pVA1ff30�

Va1 ¼ NVA1ff30� ð2-34ÞVA1 ¼ Va1

Nff � 30� ð2-35Þ

Next, only negative sequence quantities are applied toEqs. (2-30) and (2-31):

IA2 ¼ nðIa2 � Ic2Þ¼ nðIa2 � a2Ia2Þ¼ nð1� a2ÞIa2¼ n

ffiffiffi3

pIa2ff30�

IA2 ¼ NIa2ff30� ð2-36ÞIa2 ¼ IA2

Nff � 30� ð2-37Þ

Va2 ¼ nðVA2 � VB2Þ¼ nðVA2 � aVA2Þ¼ nð1� aÞVA2

¼ nffiffiffi3

pVA2ff � 30�

Va2 ¼ NVA2ff � 30�

VA2 ¼ Va2

Nff30�

If a power transformer bank is connected delta on thehigh-voltage side, as shown in Figure 2-45, the generalequations for one phase are

Ia ¼ 1

nðIA � IBÞ ð2-38Þ

VA ¼ 1

nðVa � VcÞ ð2-39Þ

Applying only positive sequence quantities to Eqs.(2-38) and (2-39),

Ia1 ¼ 1

nðIA1 � IB1Þ

¼ 1

nðIA1 � a2IA1Þ

¼ 1

nð1� a2ÞIA1

¼ffiffiffi3

pIA1

nff30�

Ia1 ¼ IA1

Nff30� ð2-40Þ

IA1 ¼ NIa1ff � 30� ð2-41Þ

Figure 2-44 Connections and phasors for an ANSI

standard power transformer bank with the wye connection

on the high side (Van leads VAN by 308).

Figure 2-45 Connections and phasors for an ANSI

standard power transformer bank with the delta connection

on the high side (Van leads VAN by 308).

38 Chapter 2

VA1 ¼ 1

nðVa1 � Vc1Þ

¼ 1

nðVa1 � aVa1Þ

¼ 1

nð1� aÞVa1

¼ffiffiffi3

pVa1

nff � 30�

VA1 ¼ Va1

Nff � 30� ð2-42Þ

Va1 ¼ NVA1ff30� ð2-43Þ

Then, applying only negative sequence quantities toEqs. (2-38) and (2-39), we obtain

Ia2 ¼ 1

nðIA2 � IB2Þ

¼ 1

nðIA2 � aIA2Þ

¼ 1

nð1� aÞIA2

¼ffiffiffi3

pIA2

nff � 30�

Ia2 ¼ IA2

Nff � 30� ð2-44Þ

IA2 ¼ NIa2ff30� ð2-45ÞVA2 ¼ 1

nðVa2 � Vc2Þ

¼ 1

nðVa2 � a2Va2Þ

¼ 1

nð1� a2ÞVa2

¼ffiffiffi3

pVa2

nff30�

VA2 ¼ Va2

Nff30� ð2-46Þ

Va2 ¼ NVA2ff � 30� ð2-47Þ

If the bank is connected according to ANSIstandards, the formulas are the same and not depen-dent on whether the wye or the delta is on the high side.In either case, the positive sequence quantities areshifted 308 in one direction, while the negative sequencequantities are shifted 308 in the opposite direction.These relations for ANSI standard connections aresummarized in Table 2-2. Zero sequence quantities arenot affected by phase shift. These either pass directlythrough the bank or, more commonly, are blocked bythe connections. Thus, in a wye-delta bank, zerosequence current and voltage on one side cannot passthrough the bank to the other side.

5.11 Fault Evaluations

The sample calculation of a phase-to-ground fault on aloop system (see Sec. 5.9) was made at no load; that is,before the fault all currents throughout the systemwere zero.

With a ground fault, current flows in not only thefaulted phase ‘‘a,’’ but also the unfaulted ‘‘b’’ and ‘‘c’’phases. The positive and zero sequence distributionfactors on any loop system will be different. Conse-quently, the positive, negative, and zero sequencecurrents will not add up to zero in the unfaultedphases. On a radial system (one with a source at oneend only for both the positive and zero sequences), thethree network distribution factors will all be equal to 1.For a phase-a-to-ground fault on these circuits, Ibequals Ic, which equals 0.

In practice, only 3I0 and related 3V0;V2, and I2values would be recorded for a phase-to-ground fault.The phase currents and voltages shown in Figure 2-43were provided for academic purposes.

The reason for showing 3I0, rather than the faultedphase current, can be seen from Figure 2-43. In mostcircuits, there is a significant difference between the Iaand 3I0 currents in any loop network. In a radialsystem, however, Ia is equal to 3I0 and ground relaysoperate on 3I0.

On phase-to-ground faults, the phase relays willreceive current and may start to operate. Coordinationbetween ground and phase relays is usually notnecessary. The principal reason there are so fewcoordination problems is that phase relays must beset above load (5A secondary), whereas ground relaysare conventionally set at 0.5 to 1.0A secondary. Sincethe ground relays are more sensitive, they will generally

Table 2-2 Phase Shift Relations for Power Transformer

Banks

High side in terms

of low sideaLow side in terms

of high sidea

Ia1 ¼ IA1

Nff30� IA1 ¼ NIa1ff � 30�

Va1 ¼ NVA1ff30� VA1 ¼ Va1

Nff � 30�

Ia2 ¼ IA2

Nff � 30� IA2 ¼ NIa2ff30�

Va2 ¼ NVA2ff � 30� VA2 ¼ VA2

Nff30�

aThe lowercase subscripts represent high-side quantities, and the

capital letter subscripts low-side quantities.


not miscoordinate with the phase relays. If higherground settings are used, the likelihood of miscoordi-nation is increased.

Under any fault condition, the total current flowinginto the ground must equal the total current flowing upthe neutrals. With an autotransformer, however,current can flow down the neutral. In this case, thefault current plus the autotransformer neutral currentequals the current up the other transformer neutrals.

The convention that current flows up the neutralwhen current is flowing down into the earth at the faulthas given rise to the idea that the grounded wye-deltatransformer bank is a ground source, a source of zerosequence current. This long-established idea is not, infact, correct. The fault is the true source. It is a

converter of positive sequence into negative sequenceand, for ground faults, into zero sequence current.

This is illustrated by a voltage plot for various faultson a simple system (Fig. 2-46). For simplicity, assumeZ1 equals Z2 equals Z0. During faults, the voltageinside the generators does not change unless the faultpersists long enough for the internal flux to change. Noappreciable voltage change should occur in high- ormedium-speed relaying.

For a solid three-phase fault, the voltage at the faultis zero. Therefore, high positive sequence-phasecurrents flow to produce the gradient shown in theplot of Figure 2-46. For a phase-to-phase fault,negative sequence voltage is produced by the faultitself. Negative sequence current, then, flows through-

Figure 2-46 Voltage gradient for various types of faults.

40 Chapter 2

out the system. The same general conditions also applyto phase-to-ground faults, except that since Va is zero,V2 and V0 are negative.

In summary, the positive sequence voltage is alwayshighest at the generators or sources and lowest at afault. In contrast, negative and zero sequence voltagesare always highest at the fault and lowest at the‘‘sources.’’

The phasor diagrams of Figure 2-20 illustrate thesame phenomena, from a different viewpoint. In athree-phase fault, the voltages collapse symmetrically,except inside the generator. The three currents havea large symmetrical increase and lagging shift ofangle.

Other phase faults shown in Figure 2-20 arecharacterized by the relative collapse of two of thephase-to-neutral voltages, compared to the relatively

normal third phase-to-neutral voltage. Two of thephase currents have a large lagging increase.

For a single-phase-to-ground fault, on the otherhand, one phase-to-neutral voltage is collapsed relativeto the other two phases. Similarly, one phase currenthas a large value and lags the line-to-ground voltage.

With wye-delta transformers between the fault andmeasurement point, the positive sequence quantitiesshift 308 in one direction, and the negative sequencequantities shift 308 in the opposite direction. As aresult, a phase-to-ground fault on the wye side of abank has the appearance of a phase-to-phase fault onthe delta side.

Figures 2-47 and 2-48 offer a final look at sequencecurrents and voltages for faults. Note that the positivesequence currents and voltages, shown in the left-handcolumns, have approximately the same phase relations

Figure 2-47 Sequence currents for various faults. Assumes

Z1 ¼ Z2 ¼ Z0.

Figure 2-48 Sequence voltages for various faults. Assumes

Z1 ¼ Z2 ¼ Z0.


for all types of faults. At the fault are variousnonsymmetrical currents and voltages, as shown inthe far right-hand column. The negative and, some-times, the zero sequence quantities provide the transi-tion between the symmetrical left-hand column andnonsymmetrical right-hand column. These quantitiesrotate and change to produce the nonsymmetrical, orunbalanced, quantity when added to the positivesequence.

These phasors can be constructed easily byremembering which fault quantity should be mini-mum or maximum. In a phase c-a fault, for example,phase-b current will be small. Thus, Ib2 will tend tobe opposite Ib1. Since phase-b voltage will berelatively uncollapsed, Vb1 and Vb2 will tend to bein phase. After one sequence phasor is established,the others can be derived from Eq. (2-12) andFigure 2-21.

6 SYMMETRICAL COMPONENTS ANDRELAYING

Since ground relays operate from zero sequencequantities, all ground relay types use symmetricalcomponents. A number of other protective relays usecombinations of the sequence quantities, as summar-ized in Table 2-3.

A zero sequence ð3I0Þ current filter is obtained byconnecting three current transformers in parallel. Azero sequence ð3V0Þ voltage filter is provided by thewye-grounded-broken-delta connection for a voltagetransformer or an auxiliary. Positive and negativesequence current and voltage filters are described inChapter 3.

Table 2-3 Protective Relays Using Symmetrical

Component Quantities for Their Operation

Device no. Application

Sequence quantities

used

50N, 51N Ground overcurrent I059N Ground voltage V0

67N Ground directional

overcurrent

I0 with I0 or V0I0or V2I2

32N Ground product

overcurrent

I20 or I0, V0

21N Ground distance I0, V0

I0, V0, V1þV2

87 Phase and ground pilot K1I1þK2I2þK0I046 Phase unbalance voltage V2

46 Phase unbalance current I2Blown fuse detection V0 and not I0

42 Chapter 2

3

Basic Relay Units


1 INTRODUCTION

Protective relays for power systems are made up of oneor more fault-detecting or decision units, along withany necessary logic networks and auxiliary units.Because a number of these fault-detecting or decisionunits are used in a variety of relays, they are calledbasic units. Basic units fall into several categories:electromechanical units, sequence networks, solid-stateunits, integrated circuits, and microprocessor architec-ture. Combinations of units are then used to formbasic logic circuits applicable to protective relays.

2 ELECTROMECHANICAL UNITS

Four types of electromechanical units are widely used:magnetic attraction, magnetic induction, D’Arsonval,and thermal units.

2.1 Magnetic Attraction Units

Three types of magnetic attraction units are incommon use: plunger (solenoid), clapper, and polar.The plunger unit, shown in Figure 3-1, is typically usedin SC, SV, and ITH relays; the clapper-type unit (Fig.3-2) in SG, AR, ICS, IIT, and MG relays; and thepolar-type unit (Fig. 3-3) in HCB, HU, and PM-typerelays.

2.1.1 Plunger Units

Plunger units have cylindrical coils with an externalmagnetic structure and a center plunger. When the

current or voltage applied to the coil exceeds thepickup value, the plunger moves upward to operate aset of contacts. The force F which moves the plunger isproportional to the square of the current in the coil.

The plunger unit’s operating characteristics arelargely determined by the plunger shape, internalcore, magnetic structure, coil design, and magneticshunts. Plunger units are instantaneous in that nodelay is purposely introduced. Typical operating timesare 5 to 50 msec, with the longer times occurring nearthe threshold values of pickup.

The unit shown in Figure 3-1a is used as a high-dropout instantaneous overcurrent unit. The steelplunger floats in an air gap provided by a nonmag-netic ring in the center of the magnetic core. When thecoil is energized, the plunger assembly moves upward,carrying a silver disk that bridges three stationarycontacts (only two are shown). A helical springabsorbs the ac plunger vibrations, producing goodcontact action. The air gap provides a ratio ofdropout to pickup of 90% or greater over a two-to-one pickup range. The pickup range can be variedfrom a two-to-one to a four-to-one range by theadjusting core screw. When the pickup range isincreased to four to one, the dropout ratio willdecrease to approximately 45%.

The more complex plunger unit shown in Figure3-1b is used as an instantaneous overcurrent or voltageunit. An adjustable flux shunt permits more precisesettings over the nominal four-to-one pickup range.This unit is relatively independent of frequency,operating on dc, 25-Hz, or nominal 60-Hz frequency.It is available in high- and low-dropout versions.

43

2.1.2 Clapper Units

Clapper units have a U-shaped magnetic frame with amovable armature across the open end. The armatureis hinged at one side and spring-restrained at the other.When the associated electrical coil is energized, thearmature moves toward the magnetic core, opening orclosing a set of contacts with a torque proportional tothe square of the coil current. The pickup and dropoutvalues of clapper units are less accurate than those ofplunger units. Clapper units are primarily applied asauxiliary or go/no-go units.

Four clapper units are shown in Figure 3-2. Thoseillustrated in Figures 3-2a and 3-2b have the samegeneral design, but the first is for dc service and thesecond for ac operation. In both units, upwardmovement of the armature releases a target, whichdrops to provide a visual indication of operation (thetarget must be reset manually). The dc ICS unit (Fig.3-2a) is commonly used to provide a seal-in around themain protective relay contacts. The ac IIT unit (Fig.3-2b) operates as an instantaneous overcurrent orinstantaneous trip unit. It is equipped with a lag-loopto smooth the force variations due to the alternatingcurrent input. Its adjustable core provides pickupadjustment over a nominal four-to-one range.

The SG (Fig. 3-2c) and MG clapper units provide awide range of contact multiplier auxiliaries: The SGhas provisions for four contacts (two make and twobreak), and the MG will accept six. The AR clapperunit (Fig. 3-2d) operates in 2 to 4 msec, with fourcontacts suitable for breaker tripping.

Figure 3-1 Plunger-type units.

Figure 3-2 Four clapper units.

44 Chapter 3

2.1.3 Polar Units

Polar units (Fig. 3-3) operate from direct currentapplied to a coil wound around the hinged armature inthe center of the magnetic structure. A permanentmagnet across the structure polarizes the armature-gappoles, as shown. The nonmagnetic spacers, located atthe rear of the magnetic frame, are bridged by twoadjustable magnetic shunts. This arrangement enablesthe magnetic flux paths to be adjusted for pickup andcontact action. With balanced air gaps (Fig. 3-3a), theflux paths are as shown and the armature will float inthe center with the coil deenergized. With the gapsunbalanced (Fig. 3-3b), some of the flux is shuntedthrough the armature. The resulting polarization holdsthe armature against one pole with the coil deener-gized. The coil is arranged so that its magnetic axis is inline with the armature and at a right angle to thepermanent magnet axis. Current in the coil magnetizesthe armature either north or south, increasing ordecreasing any prior polarization of the armature. If,as shown in Figure 3-3b, the magnetic shunt adjust-ment normally makes the armature a north pole, it willmove to the right. Direct current in the operating coil,which tends to make the contact end a south pole, will

overcome this tendency and the contact will move tothe left. Depending on design and adjustments, thispolarizing action can be gradual or quick. The left-gapadjustment (Fig. 3-3b) controls the pickup value, theright-gap adjustment the reset value. Some units useboth an operating and a restraining coil on thearmature. The polarity of the restraint coil tends tomaintain the contacts in their initial position. Currentof sufficient magnitude applied to the operating coilwill provide a force to overcome the restraint, causingthe contacts to change position. A combination ofnormally open or normally closed contacts is available.These polar units operate on alternating currentthrough a full-wave rectifier and provide very sensitive,high-speed operation on very low energy levels.

The operating equation of the polar unit is

K1Iop �K2Ir ¼ K3

fð3-1Þ

where K1 and K2 are adjusted by the magnetic shunts;K3 is a design constant; f is the permanent magneticflux; Iop is the operating current; and Ir is the restraintcurrent in milliamperes.

2.2 Magnetic Induction Units

There are two general types of magnetic inductionunits: induction disc and cylinder units. The inductiondisc unit (Fig. 3-4) is typically used in CO, CV, CR,IRV, IRD, CW, CA, and CM relays. The cylinder unit(see Fig. 3-6) is most commonly used in KD-line, KC,KDXG, KF, KRD, KRC, and KRP relays.

2.2.1 Induction Disc Units

Originally, induction disc units were based on thewatthour meter design. Modern units, however,although using the same operating principles are quitedifferent. All operate by torque derived from theinteraction of fluxes produced by an electromagnetwith those from induced currents in the plane of arotatable aluminum disc. The E unit in Figure 3-4a hasthree poles on one side of the disc and a commonmagnetic member or ‘‘keeper’’ on the opposite side.The main coil is on the center leg. Current I in the maincoil produces flux, which passes through the air gapand disc to the keeper. (A small portion of the flux isshunted off through the side air gap.) The flux fT

returns as fL through the left-hand leg and fR throughthe right-hand leg, where fT ¼ fL þ fR. A short-circuited lagging coil on the left leg causes fL to lagboth fR and fT, producing a split-phase motor action.(The phasors are shown in Figure 3-5.)

Figure 3-3 Polar-type unit

Basic Relay Units 45

Flux fL induces voltage Vs, and current Is flows,essentially in phase, in the shorted lag coil. Flux fT is

the total flux produced by main coil current I. Thethree fluxes cross the disc air gap and induce eddycurrents in the disc. These eddy currents react with thepole fluxes and produce the torque that rotates thedisc. With the same reference direction for the threefluxes as shown in Figure 3-5b, the flux shifts from leftto right and rotates the disc clockwise, as viewed fromthe top.

There are many alternative versions of the inductiondisc unit. The unit shown in Figure 3-4, for example,may have a single current or voltage input. The discalways moves in the same direction, regardless of thedirection of the input. If the lag coil is open, no torquewill exist. Other units can thus control torque in theinduction disc unit. Most commonly, a directional unitis connected in the lag coil circuit. When thedirectional unit’s contact is closed, the induction discunit has torque; when the contact is open, the unit hasno torque.

Induction disc units are used in power or directionalapplications by substituting an additional input coil forthe lag coil in the E unit. The phase relation betweenthe two inputs determines the direction of theoperating torque.

A spiral spring on the disc shaft conducts current tothe moving contact. This spring, together with theshape of the disc (an Archimedes spiral) and design ofthe electromagnet, provides a constant minimumoperating current over the contact travel range. Apermanent magnet with adjustable keeper (shunt)

Figure 3-4 Induction disc unit.

Figure 3-5 Phasors and operations of the ‘‘E’’ unit

induction disc.

46 Chapter 3

dampens the disc, and magnetic plugs in the electro-magnet control the degree of saturation. The springtension, damping magnet, and magnetic plugs allowseparate and relatively independent adjustment of theunit’s inverse-time current characteristics.

2.2.2 Cylinder Units

The operation of a cylinder unit is similar to that of aninduction motor with salient poles for the statorwindings. Shown in Figure 3-6 , the basic unit usedfor relays has an inner steel core at the center of thesquare electromagnet, with a thin-walled aluminumcylinder rotating in the air gap. Cylinder travel islimited to a few degrees by the moving contact

attached to the top of the cylinder and the stationarycontacts. A spiral spring provides reset torque.

Operating torque is a function of the product of thetwo operating quantities applied to the coils wound onthe four poles of the electromagnet and the sine of theangle between them. The torque equation is

T ¼ KI1I2 sinf12 �Ks ð3-2Þwhere K is a design constant; I1 and I2 are the currentsthrough the two coils; f12 is the angle between I1 andI2; and Ks is the restraining spring torque. Differentcombinations of input quantities can be used fordifferent applications, system voltages or currents, ornetwork voltages.

2.3. D’Arsonval Units

In the D’Arsonval unit, shown in Figure 3-7, amagnetic structure and an inner permanent magnetform a two-pole cylindrical core. A moving coil loop inthe air gap is energized by direct current, which reactswith the air gap flux to create rotational torque. TheD’Arsonval unit operates on very low energy input,such as that available from dc shunts, bridge networks,or rectified ac. The unit can also be used as a dccontact-making milliammeter or millivoltmeter.

2.4 Thermal Units

Thermal units consist of bimetallic strips or coils thathave one end fixed and the other end free. As thetemperature changes, the different coefficients ofthermal expansion of the two metals cause the freeend of the coil or strip to move. A contact attached tothe free end will then operate based on temperaturechange.

3 SEQUENCE NETWORKS

Static networks with three-phase current or voltageinputs can provide a single-phase output proportionalto positive, negative, or zero sequence quantities.These networks, also known as sequence filters, arewidely used.

3.1 Zero Sequence Networks

In zero sequence networks, three current transformersecondaries, connected in parallel, provide 3I0 from Ia,Figure 3-6 Cylinder unit.


Ib, and Ic inputs. Similarly, the secondaries of three-phase voltage transformers, connected in series withthe primary in grounded wye, provide 3V0.

3.2 Composite Sequence Current Networks

The network shown in Figure 3-8 can be adapted for avariety of single-phase outputs. Output filter voltage

VF is obtained from input currents Ia, Ib, and Ic, withneutral ð3I0Þ return. By using Thevenin’s theorem,these three-phase networks can be reduced to a simpleequivalent circuit, as shown in Figure 3-8b. VF is theopen circuit voltage at the output, and Z theimpedance looking back into the three-phase network.Zs is the self-impedance of the three-winding reactor’ssecondary with mutual impedance Xm.

The open circuit voltage (Fig. 3-8a) with switch ropen and switch s closed is the drop from VFðþÞ toVFð�Þ.

VF ¼ jðIc � IbÞXm þ IaR1 þ 3I0R0 ð3-3ÞFrom the basic symmetrical component equations [Eq.(2-24) and (2-25) in Chapter 2], we have

Ic � Ib ¼ jffiffiffi3

pI1 � j

ffiffiffi3

pI2 ð3-4Þ

Substituting this and the sequence equation (2-23) forIa in Eq. (3-3),

VF ¼�ffiffiffi3

pXmI1 þ

ffiffiffi3

pXmI2 þR1I1

þR1I2 þR1I0 þ 3I0R0

¼ðR1 �ffiffiffi3

pXmÞI1 þ ðR1 þ

ffiffiffi3

pXmÞI2

þ ðR1 þ 3R0ÞI0

ð3-5Þ

Varying Xm, R1, R0, and the connections producesdifferent output characteristics. In some applications,the currents Ib and Ic are interchanged, changing

Figure 3-7 D’Arsonval-type unit.

Figure 3-8 Composite sequence current network.

48 Chapter 3

Eq. (3-5) to

VF ¼ðR1 þffiffiffi3

pXmÞI1

þ ðR1 �ffiffiffi3

pXmÞI2 þ ðR1 þ 3R0ÞI0

ð3-6Þ

Note that the choice of design constant Xm ¼ R1=ffiffiffi3

pcauses the I2 term in Eq. (3-6) to become 0. Withswitch r closed and switch s open, the zero sequenceresponse of Eq. (3-5) and (3-6) is eliminated. The zerosequence drop across R1 is 2=3R1Iao � 1=3R1

ðIb0 þ Ic0Þ ¼ 0. The switches r and s are used inFigure 3-8 as a convenience for description only.Several typical sequence network combinations aregiven in Table 3-1.

3.3 Sequence Voltage Networks

Sequence voltage networks may be constructed toprovide a single-phase output proportional to eitherpositive or negative sequence voltage. A network incommon use is shown in Figure 3-9. Since this networkis connected phase to phase, there is no zero sequencevoltage effect.

The network is best explained through the phasordiagram (Fig. 3-10). By design, the phase angle of ZþR is 608 lagging. For convenience, consider switches sto be closed and switches r open. Impedance ZþR isthus connected across voltage Vab, and the autotrans-former across voltage Vbc. With only positive sequencevoltages (Fig. 3-10a), the current Iab1 through ZþRlags Vab1 by 608. The drop Vby1 across the auto-transformer to the tap is in phase with voltage Vbc

across the entire transformer. The tap is chosen so thatjVxb1j ¼ jVby1j. The filter output Vxy ¼ VF is thephasor sum of these two voltages.

With only negative voltages applied (Fig. 3-10b),Vxb2 is equal and opposite to Vby2, that is,

Table 3-1 Typical Sequence Network Combinations

VF reduces from:

Network type Switch r Switch s Xm ¼ Figure 3-8 notes Equation To equal

Positive sequence closed open R1=ffiffiffi3

pInterchange Ib and Ic (3-6) 2R1I1

Negative sequence closed open R1=ffiffiffi3

pAs shown (3-5) 2R1I2

HCB composite open closed R1=ffiffiffi3

pInterchange Ib and Ic (3-6) 2R1I1 þ ðR1 þ 3R0ÞI0

HCB-1 and SKB

composites

open closed 1:46R1 or

0.191 ohms

As shown (3-5) �0:2I1 þ 0:462I2þðR1 þ 3R0ÞI0

Data for Tap C of three taps available.

Figure 3-9 Sequence voltage network.

Figure 3-10 Phasor diagrams for the sequence voltage

network of Figure 3-9 with ‘‘s’’ closed and ‘‘r’’ open.


Vxy ¼ VF ¼ 0. Thus, this is a positive sequence net-work.

A negative sequence network can be made byreversing the b and c leads or, in Figure 3-9, byopening s and closing r. Then Figure 3-10a conditionsapply to a negative sequence, giving an output VF;Figure 3-10b conditions apply to a positive sequencewith VF ¼ 0.

This interchange of b and c leads to either thecurrent or voltage networks offers a very convenienttechnique for checking the networks. For example, thenegative sequence current network should have nooutput on a balanced power-system load but byinterchanging the b and c leads it should produce fulloutput on test.

4 SOLID-STATE UNITS

4.1 Semiconductor Components

Solid-state relays use various low-power components:diodes, transistors, thyristors, and associated resistorsand capacitors. These components have been designedinto logic units used in many relays. Before these logicunits are described in detail, the semiconductorcomponents and their characteristics will be reviewed(Fig. 3-11). Relays use silicon-type components almostexclusively because of their stability over a widetemperature range.

4.1.1 Diode

The diode (Fig. 3-11a) is a two-terminal device thatconducts in one direction but does not conduct in theother. The device manifests a voltage drop forconduction in the forward direction of approximately0.7V. The limit of voltage to be applied in the reversedirection is defined by the rating of the diode. Failureof the diode is expected if a voltage in excess of therating is applied in the reverse direction.

These devices are used in dc circuits to blockinteraction between circuits, for ac test circuits togenerate a half-wave rectified current wave shape, or asa protective device around a coil to minimize thevoltage associated with coil current interruption.

4.1.2 Zener Diode

The zener diode (Fig. 3-11b) differs from the diodedescribed above in having a sharp and reproduciblereverse breakdown voltage, called the zener voltage. Ifthe current is limited to within rated values, the diode

recovers its nonconducting characteristics when thereverse voltage falls below the zener value. They areused for surge protection, voltage-regulating functions,and other applications in which a distinct conductionlevel is desired.

Where conduction is desired in both directions witha threshold at a level at which conduction occurs, theback-to-back zener (Fig. 3-11c), commonly known as avolt trap or zener clipper, is used. The characteristics ofthese devices are essentially the same in both theforward and reverse direction.

4.1.3 Varistor, Thermistor

The characteristics of the varistor are shown in Fig.3-11d. It has a voltage-dependent nonlinear charac-teristic. The thermistor depicted in Figure 3-11e is anonlinear device whose resistance varies with tempera-ture.

4.1.4 Transistor

In relaying, the transistor is used primarily as a switch.For this function, it is shifted from a nonconducting toconducting state by the base current Ib. The transistor

Figure 3-11 Semiconductor components and their charac-

teristics.

50 Chapter 3

is nonconducting until Ib is increased to a value atwhich the transistor conducts, and a collector currentIc and emitter current Ie flow (Fig. 3-12). The emittercurrent Ie is the sum of Ib and Ic. Very small values ofIb are able to control much larger values of Ic and Ie(Fig. 3-13).

4.1.5 Thyristor

The thyristor (Fig. 3-14) is a diode with a thirdelectrode (the gate). The thyristor is also known as asilicon-controlled rectifier (SCR). With forward vol-tage applied, the thyristor will not conduct until gatecurrent Ig is applied to trigger conduction. The higherthe gate current, the lower the anode-to-cathodevoltage (VF) required to start anode conduction. Afterconduction is established and the gate current isremoved, the anode current IF continues to flow. Theminimum anode current required to sustain conduc-tion is called the holding current IH.

4.1.6 Unijunction Transistor

The unijunction transistor (Fig. 3-15) has two bases, b1and b2, and one emitter e. When Ve reaches the peakvalue Vp, the device conducts and passes current Ie.Current will continue to flow as long as Ve does not fallbelow the minimum value Vv. The unijunctiontransistor is used for oscillator and timing circuits.

Figure 3-12 The transistor and equivalent electrical sym-

bols.

Figure 3-13 Typical characteristic curves of transistor.

Figure 3-14 The thyristor and its characteristics.


4.2 Solid-State Logic Units

4.2.1 Basic Principles

Solid-state logic units are combinations of solid-statecomponents designed to use dc voltage signals toperform logic functions. A logic unit has only twostates: no output, represented by 0 (zero), and output,represented by 1 (one). Two logic conventions are usedto indicate the voltages associated with the 0 and 1states. In normal logic, 0 is equivalent to zero voltageand 1 to normal voltage. In reverse logic, thecorresponding voltage equivalents are reversed; 0 isequivalent to normal voltage and 1 to zero voltage.

In positive logic, inputs and outputs are positive; innegative logic, both inputs and outputs are negative.Relay systems normally use positive logic, althoughsome elements may use negative signal inputs andoutputs.

Logic units are shown diagrammatically in theirquiescent state, that is, the normal or ‘‘at-rest’’ state. Thequiescent state corresponds to the normally deenergizedrepresentation in electromechanical relay circuitry.

4.2.2 Logic Unit Representation

Logic units are represented by characteristic functionsymbols (Fig. 3-16). Two sets of symbols are in

common use in the United States. In the commercial/military system, the type of function is indicated by thedistinctive geometrical shape of the symbol. In solid-state relaying, the name of the logic function is simplywritten in a rectangle or block, or a distinctive symbolsuch as ‘‘&’’ is used inside a block. The Europeanpractice is similar to this. Convention dictates thatinputs are shown on the left-hand side and outputs onthe right-hand side. The symbols and terminology usedcomply with IEEE Standard 91-1973 (ANSI StandardU32 14-1973), ‘‘Graphic Symbols for Logic Dia-grams.’’

When a logic function has only two inputs, itsoutput is usually simple to determine. For three ormore inputs, particularly with combination logicfunctions, a logic or truth table offers a convenientmethod of determining the output. A logic table for afunction with three inputs and one output is shown inFigure 3-17. The table lists all possible combinations ofzeros and ones for the inputs. Each output could be 0or 1, depending on the function.

4.3 Principal Logic Units

In this section, the major units used in relaying will bedescribed. Detailed circuit descriptions will be kept to aminimum. For simplicity, the diagrams will show onlytwo inputs per function and include electromechanicalcontact equivalents.

4.3.1 AND Unit

The AND logic element is shown in Figure 3-18. Thesimplest type consists of forward-biased diodes andresistors (Fig. 3-18a). The symbolic representation andelectromechanical equivalents for this unit are given inFigure 3-18b, the logic table in Figure 3-18c. Theforward-biased diodes shunt the output terminal, and

Figure 3-16 Examples of logic symbols. Figure 3-17 Example of logic table.

Figure 3-15 The unijunction transistor and its character-

istics.

52 Chapter 3

no output voltage can appear unless all input diodeshave a reverse bias that equals or exceeds the forwardbias. Since inputs are either 0 or 1, there is no in-between state that would allow partial output voltage.Thus, the output is either 0 or 1, as shown in the logictable. Three variations of the AND element areprovided in Figure 3-19.

4.3.2 OR Unit

The OR unit is shown in Figure 3-20. Again, thesimplest type of unit consists of resistors and diodes.The symbolic representation and electromechanicalequivalents for the unit are illustrated in Figure 3-20b,the logic table in Figure 3-20c. Since the diodes are notbiased, an input voltage applied to any input willproduce an output voltage at X.

Three variations of the OR unit, comparable to thoseof the AND element, are shown together with theirelectromechanical equivalents in Figure 3-21a, b, and c.By comparing Figure 3-21a with Figure 3-19b, it is clearthat the inverse OR unit is equivalent to the negationAND. Similarly, the negation OR of Figure 3-21bis equivalent to the inverse AND of Figure 3-19a.

4.3.3 NOT Unit

The negation, or NOT, unit (Fig. 3-22) is a frequentlyused logic element. This unit changes the state of the

input from 0 to 1 or vice versa. For convenience, thesymbol depicted in Figure 3-22a is replaced by thatshown in Figure 3-22b for negated inputs and thatshown in Figure 3-22c for negated outputs.

The NOT unit, together with its electromechanicalequivalent and logic table, is illustrated in Figure 3-23.Although the NOT unit can be included in logicdiagrams as a separate unit, it is usually combined withother units using the symbols shown in Figure 3-22.

4.3.4 Time-Delay Units

Time-delay units are used in the normal manner toprovide ON and/or OFF delays. The U.S. andEuropean symbolic representations are presented inFigure 3-24. The X value in Figure 3-24 is the pickuptime, that is, the time that elapses between an inputsignal being received and an output signal appearing.The Y value is the dropout time, that is, the time afterthe input signal is removed until the output signal goesto 0. In Figure 3-24, W-X is the range of the pickuptime and Y-Z that of the dropout time. Any of thevalues can be 0. Time values are always in millisecondsunless otherwise indicated.

Figure 3-18 AND logic.

Figure 3-19 Variations of AND logic.

Figure 3-20 OR logic.

Figure 3-21 Variations of OR logic.


5 BASIC LOGIC CIRCUITS

In describing basic logic circuits, two types of diagramsare used: the logic block diagram and logic circuitschematic diagram. In the logic block diagram, theunits are represented by their logic symbols, and thelogic symbol blocks are interconnected to provide acomplete functional representation of the system. Inthe logic circuit schematic diagram, the elements areshown schematically. Unit interconnections aredepicted in the same way as in normal schematicdiagrams. The logic block diagram is useful in showingthe complete system in functional form; the logicschematic circuit diagram indicates how the logic unitsoperate. In the following discussion, logic schematicdiagrams will be used.

5.1 Fault-Sensing Data Processing Units

In solid-state relays, fault-sensing and data-processinglogic circuits use power-system inputs (voltage, cur-

rent, phase angle, frequency, and so on) to determine ifany intolerable system conditions exist within therelay’s zone of protection. The conventional functionsobtained by logic circuits are listed in Table 3-2.

5.1.1 Magnitude Comparison

There are two basic types of magnitude comparisonlogic units: fixed-reference and variable-reference.

Fixed-Reference

The logic circuit used for an instantaneous overcurrentunit (Fig. 3-25) is basically a dc-level detector. Inputcurrent from the current transformer secondary istransformed to a current-derived voltage on thesecondary of the input transformer. This voltage islimited by zener clipper Z1 and resistor R2. For lowinput currents, the voltage is proportional to thecurrent, as determined by R1 and R3. The minimumpickup is adjusted via the setting of R1. A low R1setting diverts more current through R1 and R3, andless to the phase splitter.

The phase splitter consists of a resistor-capacitornetwork, transformer, and bridge rectifier. The outputvoltage of the phase splitter is shown in the upper partof Figure 3-25. When this voltage equals the zenervoltage of Z2, Z2 will conduct, providing a basecurrent that turns on Q1. Q1 then turns on Q2,providing an output current through D2 and R9. Q2provides positive feedback through R7 and D1,compounding the effect on the level detector.

Figure 3-22 Negation symbol convention.

Figure 3-23 NOT logic.

Figure 3-24 Examples of time delay units.

Table 3-2 Conventional Functions Obtained by Fault

Sensing and Data Processing Logic Circuits

Conventional

function Logic circuits

Typical relay

types

Instantaneous

overcurrent

Magnitude

comparison with

fixed reference

SI-T 50B

Time overcurrent Magnitude

comparison with

fixed reference

and time

50D 51

Ground distance Magnitude

comparison with

variable reference

SDG-T

Phase distance Block-block

comparison

SKD-T

Directional Coincident-time

(ring modulator)

SRGU

54 Chapter 3

The dropout current can be adjusted by resistor R7,normally set for a dropout/pickup ratio of about 0.97.Positive feedback provides the equivalent of snapaction, and the 3% bandwidth prevents the equivalentof chattering for current values close to minimumpickup. This type of circuit could also be used forovervoltage.

Variable-Reference

This logic unit discriminates between the value of anoperate voltage and the smallest of three restraintvoltages. Shown in Figure 3-26, this type of circuitforms the decision logic element for the SDG grounddistance relay described in Chapter 12. The restraintvoltages (Vx, Vy, and Vz) and operating voltage areconnected in opposition through tunnel diode TD1and diodes D25, D26, and D27. When the operatingvoltage is larger than any of the three restraintvoltages, current will flow through TD1. A smallcurrent through TD1 drives it to a high voltage,turning on Q1 and producing a voltage across theoutput terminal. The tunnel diode characteristicprovides a sharp turn-on point, which serves as aneffective triggering action.

Since double phase-to-ground faults may causeover-reach of the ground distance relay, a desensitizercircuit is included. This circuit consists of threeminimum voltage networks. A portion k of each of

the restraint voltages is input to the desensitizer circuit.When any combination of two restraint voltages issmaller than the third restraint voltage, an outputproduces a blocking action through D86, preventingQ1 from turning on. When the operating voltagebecomes larger than the largest restraint voltage,reverse bias is applied to D86 through D38, turningon Q1.

5.1.2 Phase-Angle Comparison

Phase-angle comparator logic circuitry produces anoutput when the phase angle between two quantities iswithin certain critical limits. Either of these twoquantities, the polarizing (or reference quantity) andoperating quantity, may be current or voltage.

Two types of phase-angle comparator logic circuitryare in common use: block-block and coincident-timecomparison.

Block-Block Comparison

The block-block type of phase-angle comparator usesthe zero-crossing detector principle to generate squarewaves. Additional logic circuitry provides an output ifthe operating quantity leads the polarizing quantity.Phase relations for the operating condition are shownin Figure 3-27. An output is obtained if the operating

Figure 3-25 Magnitude comparison dc level detector as an instantaneous overcurrent unit.


input leads the polarizing input by 0 to 1808.Conversely, no output (restraint) occurs if the operat-ing input lags the polarizing input by 0 to 1808. Thephase relation for this restraint condition is given inFigure 3-28.

Half-cycle square waves are generated at each zerocrossing of the respective input quantities. The polarityof the square waves is the same as that of thegenerating quantity during corresponding half-cycles.

One half of the circuit of a block-block type ofcomparator (Fig. 3-29) makes the comparison duringthe positive half-cycles. The input diodes, arrays DAand DB, limit the input voltage to 1.5V and the outputof transformers T1 and T2 to about 12V.

For the operating condition shown in Figure 3-27,the leading operating input makes the base of Q1positive before the polarizing input can make the baseof Q3 positive. Thus, Q1 turns on first, which then

Figure 3-26 Magnitude comparator circuit.

Figure 3-27 Phase relationship of block-block circuit for

operate condition.

Figure 3-28 Phase relationship of block-block circuit for

restraint condition.

56 Chapter 3

turns on Q2. Since Q5 has not been gated, it is in theblock state, permitting an output through Q2, R8, andthe output diodes. When Q2 turns on, D3 is reverse-biased through D4 from the 20-V supply. This preventsthe flow of base current from turning on Q4 asotherwise would occur as the lagging polarizing inputbecomes positive and turns on Q3. Since Q4 cannotturn on, a half-cycle of output occurs. Similarly, duringthe negative half-cycle the leading operating inputprovides an output in the other half of the circuit,which connects through its negative half-cycle diode tothe output.

If, however, the polarizing input leads the operatinginput (Fig. 3-28), the base of Q3 becomes positivebefore the base of Q1. Q3 turns on first and then turnson Q4. The current flowing through Z1, Q4, R6, andR7 produces a voltage drop across R7. This voltagedrop gates thyristor Q5, causing it to conduct andshort the output to negative. As the lagging operatinginput becomes positive, it turns on Q1 and Q2 and(since Q5 is conducting) the current through Q2 andR8 is shunted to negative. The operating input thatremains when the polarizing half-cycle is completedcannot produce an output, because Q5 continues toconduct. Recovery is determined by the anode-to-cathode current, and R8 is set to allow sufficientholding current from the 20-V supply to maintain Q5in a conducting state until the operating quantity ispractically 0.

Coincident-Time Comparison (Ring Modulator)

Functioning like a biased bridge rectifier, the ringmodulator type of phase-angle comparator producesan output when the operating quantity leads or lags thepolarizing quantity by 908 or less. This characteristicmakes the ring modulator applicable as a directionalsensing unit.

Figure 3-30 shows the operating principles of thebridge under several input conditions. Current inputsare depicted, but combinations of current and voltagecan also be used. Solid arrows indicate the inputoperating quantities, open arrows the input polarizingquantities. Actual current is the phasor sum of thecurrents shown. The in-phase conditions are illustratedin Figure 3-30.

In the bridge rectifier, two diodes are forward-biased by the larger current, and the magnitude of theoutput is determined by the smaller current. When theoperating current is larger (top half of Fig. 3-30a), D1and D3 are forward-biased, with the return throughR1 and R2 blocked by D4. The polarizing current isshown in two parts, each one half of IPOL. The halfgoing down through the transformer from point Aflows backward through D3 and R1 to the polarizingterminal. However, since the operating current islarger, net current in D3 is forward. The outputvoltage IPOLR1 is proportional to the smaller current.

If the operating current reverses and is still largerthan the polarizing current, D4 and D2 are forward-biased, with the return through R2 and R1 blocked byD1. Polarizing current going up from point A flowsbackward throughD2and up throughR2 (net current inD2 is forward). The output voltage�IPOLR2 is reversed.

With reversed but still smaller polarizing current,part of the polarizing current would flow up throughR1, around through D1 (which is forward-biased bythe operating current) and back. The other part wouldflow up through R1, down through D3, and back.Again, the output voltage �IPOLR1 is reversed.

If the polarizing current is larger, as in the bottomhalf of Figure 3-30a, D1 is forward-biased through R1,and D4 is forward-biased through R2. If R1 equals R2,the net output from the polarizing current is 0. Thesmaller operating current flows through D1, R1, R2,and back through D4. Net current in D4 is forwardbecause of the larger value of polarizing current. Thenet output then is IOP (R1þR2), or 2IOPR1. Reversingeither the polarizing or operating current will reversethe output voltage.

The output will not be 0 as long as the smallercurrent is above a threshold or pickup value. When the

Figure 3-29 Block-block type phase-angle comparator

circuit.


sum of the currents through a diode is 0, as through D3in Figure 3-30a, the output is still IOPR1. Any tendencyof either current to flow in another path, because D3 isnot conducting, will result in one of the two

components becoming larger, that is, the zero condi-tion no longer exists.

Figure 3-30a shows the ring modulator operationwhen the operating current is larger than the polarizing

Figure 3-30 Principle of operation of ring-modulator type phase-angle comparator.

58 Chapter 3

current and leads it by 908. At time 0, half of IPOL flowsup through R1, down through D3, and returns up thelower half of the transformer to A. The other half flowsdown through R2, up through D2, and down the upperhalf of the transformer to A. The net output is 0 sinceIOP is 0. As IOP increases from 0 to equal IPOL (pointP), IOP flows through D2, R2, R1, and D3, producingan output of �2IOPR1, where R1 and R2 are equal.When IOP equals IPOL, the current in D2 goes to 0. AsIOP becomes larger than IPOL, D1 conducts. Thenegative IPOL all flows through R1; half passes throughD1, and the other half continues through D3. IOP flowsthrough D1 and D3, producing an output of�2IPOLR1.

At time Q, when IOP again equals IPOL, thepolarizing current is about to become the largercurrent and forward-bias D1 and D4. The outputchanges to þ2IOPR1 and decreases. When IOP crossesthe zero axis, the output is 0. As the operating currentbecomes negative, so does the output, which reaches amaximum of �2IOPR.

Further analysis shows that there is a maximumpositive or negative output each time IOP¼ IPOL andalternate one-half-cycle periods (4.17msec) of positiveand negative outputs. These outputs are crosshatchedin Figure 3-30b. Similar results are obtained if thepolarizing quantity is greater than the operatingquantity, but with IOP leading IPOL by 908.

5.2 Amplification Units

5.2.1 Breaker Trip Coil Initiator

The breaker trip coil initiator circuit both providespower amplification for a trip coil and isolates thecontrol circuitry from the tripping energy source (thestationbattery).A typical circuit is shown inFigure 3-31.

Q1 turns on when the input voltage from the fault-sensing and data-processing circuit exceeds 2V. Q1then turns on Q2, allowing C2 to charge through R6.When the voltage across C2 reaches the ‘‘firingvoltage’’ of the unijunction transistor Q3, the capacitorenergy discharges through T1. This discharge reducesthe voltage across the capacitor, turning off Q3 untilthe charge on C2 builds up again.

In this way, a repetitive train of pulses is generatedas long as the input signal exists. These pulses aretransformed through LA, Q4, T2 primary, LB, and Z4to trip the circuit breaker. The time delay of this circuitis approximately 1msec. T1 has two secondaries, thesecond of which is connected to a similar Q4 circuitryfor double trip.

Except for the transformer T2, the devices asso-ciated with Q4 provide security. Zener Z1 clips high-voltage transients on the battery leads to a level of one-third of the Q4 rating. This voltage clipping preventsthe false operation of Q4 from surges and overvoltage.The two-winding reactor LA-LB suppresses anytransients that could be transmitted through theinterwinding capacitance of T1 or between the tripcircuit and other logic circuit wiring. Zener Z4 preventsshock excitation from setting up high-frequencyoscillation, which might reverse the current throughQ4 and return it to a blocking state.

Capacitor C3 is initially charged through R9 and Z3when the breaker or switch is closed, bypassing T2 toavoid a false indication. When Q4 fires, C3 dischargesthrough Q4, Z2, and R8. This discharge provides aholding current for Q4 of about 1msec, long enoughfor the current through the inductive trip coil to reachthe required holding current for Q4.

5.3 Auxiliary Units

5.3.1 Annunciator Circuits

Two types of circuits are used to provide light andalarm indications: One is for circuit-breaker-tripoperations and the other for general use.

Typical breaker-trip indicator and alarm logic areshown in Figure 3-32. Transformer T2 is in the trip

Figure 3-31 Breaker trip coil initiation circuit.


circuit, as in Figure 3-31. The transformer core usessquare-hystersis loop material to produce a very smallexciting current and negligible inductive reactancewhen saturated. When trip current flows (after Q4fires), the circuit of R1, C1, R2, and R3 stretches a 2-msec pulse at the secondary of T2 into 6msec, at 20V,at the output of Q2. The input signal turns on both Q1and Q2 to charge capacitor C2. When the voltagebuilds up to the ‘‘intrinsic standoff ratio’’ of theunijunction transistor (VP of Fig. 3-15), Q3 firesand gates Q4, energizing the indicating light. Theconduction of Q4 also gates Q5 through R10 from thedrop across R11. Q5 energizes the alarm relay. Even ifthe indicating light circuit is open, Q5 will still begated.

A general indicator circuit is shown in Figure 3-33.The normal condition is a 1 input, which makes Q1conducting. For indication, the 1 is removed, turningoff Q1. Then C1 charges through R3 and R7. Whenthe voltage across C1 reaches the firing point of Q2, Q2is turned on, gating Q4 and Q5 to energize theindicating light and alarm relay.

The indicating lights are the solid-state equivalent ofmechanical indicating targets. Red lights are used toindicate tripping or which sensing unit signaled a trip,amber lights general alarms, blue lights testing. Sixty-volt lamps operated at 48V or 120-V lamps at 97Vprovide a filament life of more than 30,000 hr.

5.3.2 Coordinating and Loop Logic Timers

Fixed time-delay timers are used extensively in logiccircuitry. A typical circuit of this type is shown inFigure 3-34. With an input, Q1 is normally conductingand shorts C1 through R4. Removing the input turnsoff Q1 and permits C1 to charge through R3 and R4.When the voltage across C1 reaches the zener voltageof Z1 plus the potential hill of D1 and Q2, base currentwill flow, turning on Q2. Turning on Q2 removesvoltage from the output. The fixed time interval isbetween removal of input to removal of output.Although normally used for short delays, the judicious

Figure 3-32 Breaker trip indicator and alarm circuit.

Figure 3-33 Indicator and alarm circuit.

60 Chapter 3

selection of values for R3, R4, C1, and Z1 provides awide range of available time delays. Similar circuitrycan provide a delay between an ON input and ONoutput, or other variations. Also, timers can be madeadjustable by making elements such as R3 adjustable.

5.3.3 Toggle or Latching Circuits

Toggle or latching circuits, known as flip-flops, arebistable units similar to a latched-in or toggle-typerelay. An operating signal will make the unit changestate; removal of the signal will leave the unit in thenew state. A momentary reset signal will restore theunit to its original state. Normally, a momentaryoperating signal will change the output from 0 to 1 anda momentary reset signal will change the output from 1back to 0. The typical circuit shown in Figure 3-35 issimplified to aid in the explanation of its operation.

The circuit depicted is in the reset state with a‘‘clear’’ output and no ‘‘set’’ output. The voltagedividers R3, R5, and R8 provide base voltage to Q2.Since Q2 is conducting, the set output is shunted to

negative. Q1 is not conducting. Its base, suppliedthrough R6, R4, and R2, is a negative. There is,however, a clear output from the R3-R5-R8 voltagedivider.

Closing the set input switch S1 momentarilyreverses this condition. Voltage divider R1-R2 pro-vides base drive to turn on Q1. The base drive for Q2 isthen shunted through Q1 to negative, and Q2 is turnedoff. When S1 is opened, Q1 will remain on, throughvoltage divider R2-R4-R6. Q2 will remain off, sinceR5-R8 ties the base of Q2 to negative. Thus, with Q1on and Q2 off, there is a set output but no clear output.

When a momentary signal is applied to the resetinput, voltage divider R7-R8 provides base drive toturn on Q2 again. This ties the base of Q1 to negative,turning it off. Q2 then remains on, even after the resetsignal is removed. The unit is now back to its reset ornormal state.

Figure 3-36a is a symbolic representation of anormal flip-flop. A modification to the normal flip-flopis to desensitize it by holding Q2 in a saturatedcondition. When saturated, Q2 keeps conducting evenwhen Q1 turns on. This prevents a spurious set signalfrom producing a set output. The modified flip-flopmust first be ‘‘armed’’ by introducing an input armsignal (Fig. 3-36b). This signal removes the desensitiz-ing bias from Q2 and allows it to turn off when thenormal set input signal is applied.

The flip-flop can also be combined with AND logic,so that two or more separate set input signals must bereceived simultaneously to produce an output. Thismodification may also be provided with desensitizing,again requiring an arming signal. These modified flip-flops are commonly used for the final trip logic unit insolid-state relaying systems.

Figure 3-34 Typical logic timer circuit.

Figure 3-35 Flip-flop circuit. Figure 3-36 Flip-flop logic symbols.


5.3.4 Isolator and Buffer Circuits

Output and input isolators separate and electricallyisolate dc circuits between logic units. Used on theinput and output of each separately packaged relay,buffers protect the logic circuit from transients andsurge on interconnecting leads and circuitry. Bothisolator and buffer circuits protect solid-state relaysagainst undesirable operation on spurious signals.

Input Isolator

A typical input isolator circuit is shown in Figure 3-37.A 20-V input to the pulsing circuit of R1-R3-C1-D2charges the capacitor C1. When the capacitor voltagereaches the breakdown voltage of the four-layer diodeD2, a pulse is transmitted through T1. The dischargeof C1 turns off D2 until the voltage across C1 buildsup again. Thus, a series of pulses continues as long asthe input signal exists. Zener Z1 provides surgeprotection clipping at 20V. The pulses are rectifiedand accumulated on C2. C2-R4-R5 provide a steady dcinput to Q1 until the input is removed. Q1 conducts,turning on Q2, providing a 20-V output.

Output Isolator

The input section of the output isolator circuit(Figure 3-38) is similar to the breaker-trip coilinitiation circuit shown in Figure 3-31. The outputisolator circuit differs in that a four-layer diode D1,rather than a unijunction transistor, provides a pulsechain through T1. An input voltage turns on Q1 and

Q2, charging C2. The voltage across C2 triggers D1, asdescribed above. The pulses, rectified and filtered, areapplied to the base of Q3, turning it on and producingan output. Zener Z1 provides surge protection clippingat 20V.

Input Buffer

The input buffer circuit is shown in Figure 3-39. Anormal 20-V signal will result in approximately 90% ofthe voltage appearing across R3 and capacitor C1.When the voltage on C1 builds up to around 5 to 7V,current flows through Z2, D1, and R4. Q1 then turnson, which produces an output. R5 is required when theoutput drives a PNP stage, but is omitted for an NPNstage. For internal logic circuitry, Q1 can be turned onby an unbuffered input.

There are three types of buffering: (1) A high-frequency, high-voltage surge on the input, such as the1.0- to 1.5-MHz, 2500-V standard test surge, isdropped across R1 and clipped to 20V by Z1; (2) allsignals of 150 to 200 msec are delayed by means of R1-R2-C1; (3) a minimum threshold voltage of 6V isrequired to turn on Q1. Thus the maximum ‘‘0 level’’voltage is 6V. A bona fide signal must exist for at least150 msec. This buffer is described further in Chapter 4(see Fig. 4-15).

Output Buffer

The output buffer circuit is shown in Figure 3-40. Aninput greater than 2V turns on Q1 and Q2 to provide

Figure 3-37 Input isolator circuit.

62 Chapter 3

approximately 18V output. C1 provides a 75-msecdelay through the unit. High-voltage, high-frequencytransients on the output are limited and clipped by R6and the 24-V zener Z1. Should the output be shorted,Q2 is protected by the current-limiting action of R6.

Optical Isolator

The isolation of solid-state circuits and componentsfrom input and output signals is also accomplishedwith the use of optocoupler devices. This integratedcircuit component uses an internal light-emitting diode(LED) and a photon detector to transmit signals,providing optical isolation between inputs and out-puts.

6 INTEGRATED CIRCUITS

The next trend in solid-state relaying was toward theuse of linear and digital integrated circuits to replacethe discrete transistor circuits described previously. Anoverview of the linear integrated circuit operationalamplifier and its application to basic relay unitsfollows.

6.1 Operational Amplifier

Figure 3-41 shows the equivalent circuit of a basicoperational amplifier. The triangle symbol is used forthis device. The supply voltages +Vcc (generally +15Vdc) with a ‘‘common’’ of 0V are not shown. The

Figure 3-38 Output isolator circuit.

Figure 3-39 Input buffer circuit. Figure 3-40 Output buffer circuit.


input terminals are a and b; b is the noninverting inputsince a positive voltage produces a positive output. Apositive voltage on a, the inverting terminal, will yielda negative output.

The output e0 is amplified by the open-loop gain Aso that

e0 ¼ Aen ¼ Aðeb � eaÞ ð3-7ÞThe plot in Figure 3-42 shows that a small differentialchange drives the amplifier into saturation since theopen-loop gain A is very large.

Most applications use negative feedback. InFigure 3-43, where Zf is connected from the outputto the inverting input a, Iin can be determined from thedrops around the input loop,

� ein þ IinZin � en þ eref ¼ 0

Iin ¼ ein þ en � eref

Zin

ð3-8Þ

If Zf is much smaller than Ri, the input resistance, theassumption is that no current flows in the a or bterminals, so that

Iin ¼ If ð3-9Þ

The drops around the Zf feedback loop are

� eref þ en þ IfZf þ e0 ¼ 0

If ¼ eref � en � e0

Zf

ð3-10Þ

Equating Eqs. (3-8) and (3-10) with the assumption ofEq. (3-9) provides

en ¼ eref � Zf

Zf þ Zin

� �ein � Zin

Zf þ Zin

� �e0 ð3-11Þ

Substituting Eq. (3-7) and solving for e0, we obtain

e01

Aþ Zin

Zin þ Zf

� �¼ eref � Zf

Zf þ Zin

� �ein ð3-12Þ

If we assume that A is very large, 1/A approaches 0and is much less than

Zin

Zin þ Zf

Thus,

e0 ¼ 1þ Zf

Zin

� �eref � Zf

Zin

� �ein ð3-13Þ

Equation (3-13) is the general operational amplifierequation with negative feedback. If it is substitutedinto Eq. (3-11), the solution for en will equal 0. Thus, aand b terminals are of the same relative potential. Theinverting input terminal (a) is referred to as virtualground.

Figure 3-41 The equivalent circuit of an operational

amplifier.

Figure 3-42 Saturation characteristics for circuit in Figure

3-41.

Figure 3-43 An operational amplifier with negative feed-

back.

64 Chapter 3

6.2 Basic Operational Amplifier Units

A number of basic units are derived from a singleoperational amplifier to use in relay circuits. These aredescribed without the additional components requiredfor accuracy, stability, or compensation.

6.2.1 Inverting Amplifiers

The inverting amplifier of Figure 3-44 is the circuit ofFigure 3-43 with terminal b connected directly tocommon (0 V). From Eq. (3-13), we get

e0 ¼ � Zf

Zin

� �ein ð3-14Þ

If resistors are used as shown in Figure 3-44, theoutput e0 is the opposite of the input modified by thescale factor

Rf

Rin

6.2.2 Noninverting Amplifiers

If the a input is reduced to 0 through Rin (Fig. 3-45a)and the input applied to terminal b instead of eref , Eq.(3-13) reduces to

e0 ¼ 1þ Rf

Rin

� �ein ð3-15Þ

The input and output are in phase with a scale factorof

1þ Rf

Rin

If Rf is made very large compared to Rin, then for asine wave input, the output essentially will be aninphase square wave to provide a squaring circuit.

Another version is the voltage follower shown inFigure 3-45b. Rf approaches 0, a short circuit, and Rin

approaches infinity, an open circuit. The gain factor

Rf

Rin

approaches 0 so that the scale factor

1þ Rf

Rin

approaches unity. Thus, the output voltage e0 equalsor follows ein. In this circuit, the input impedance seenby ein essentially is infinite and no current flows intothe b terminal.

6.2.3 Adders

An adder unit (Fig. 3-46) has two separate inputsthrough Ra1 and Ra2 to the negative terminal a withterminal b at 0. Equation (3-13) reduces to

e0 ¼ � Rf

Ra1ea1 � Rf

Ra2ea2 ð3-16Þ

If Ra1 ¼ Ra2 ¼ Rf, then the output equals the negativeof ea1 þ ea2.Figure 3-44 An inverting amplifier unit.

Figure 3-45 A noninverting amplifier and voltage follower

unit.


6.2.4 Subtractors

The basic circuit is shown in Figure 3-47. The voltageat the plus terminal of the operational amplifier will be

Rf

Rf þRineb

Substituting this in Equation (3-13), we obtain

e0 ¼ Rf

Rinðeb � eaÞ ð3-17Þ

If Rf ¼ Rin, then e0 ¼ eb � ea.

6.2.5 Integrator and Simple Low-Pass Filter

With a capacitor as the feedback component, theinverting amplifier of Figure 3-44 becomes an inte-grator (Fig. 3-48):

e0 ¼ � 1

C

Zifdt ð3-18Þ

and since

if ¼ ein

Rin¼ iin

e0 ¼ � 1

RinC

Zeindt

ð3-19Þ

this circuit is a simple low-pass filter. Consideringmagnitudes only,

Zf ¼ 1

2pfC

so that Eq. (3-14) becomes

je0j ¼ � 1

2pfCRinjeinj ð3-20Þ

Thus, as frequency increases, the magnitude of e0decreases.

6.2.6 Differentiator and Simple High-Pass FilterUnit

This circuit is shown in Figure 3-49 and is the invertedamplifier circuit with a capacitor in the input circuit

iin ¼ if ¼ Cdein

dtð3-21Þ

so that

e0 ¼ �RfCdein

dtð3-22Þ

with magnitudes from Eq. (3-14), where

Zin ¼ 1

2pfCand Zf ¼ Rf :

Thus,

je0j ¼ �2pfCRf jeinj ð3-23Þ

Figure 3-46 An adder unit.

Figure 3-47 A subtractor unit.

Figure 3-48 An integrator and low-pass filter unit.

66 Chapter 3

This is a simple high-pass filter since as f decreases, je0jdecreases.

6.2.7 Phase-Shift Units

A variety of phase-shift units are obtained usingcapacitor and variable resistor combinations. Theseare illustrated in Figure 3-50. A phase-angle range of90 to 1808 is obtained with Zf adjustable from 0 to�908 (Fig. 3-50a), 1808 to 2708 with Zin adjustablefrom 08 to �908 (Fig. 3-50b). Inverting operationalamplifiers are used in both of these circuits.

Noninverting amplifiers with the RC networkconnected as a voltage divider (Figs. 3-50c and d)provide a phase-angle range of 0 to þ908 or 0 to �908,depending on the position of R and C.

6.2.8 Level Detectors

Figure 3-51a shows a level detector using the opera-tional amplifier in the differential mode. From Eq.(3-7), we have

e0 ¼ Aðeref � einÞ ð3-24Þand

eref ¼ R2

R1 þR2

� �Vcc ð3-25Þ

As illustrated in Figure 3-42, a change in the level ofeinðeaÞ slightly before or below erefðebÞ will cause theamplifier to go into either negative or positivesaturations, respectively. With eref formed by the R1-R2 voltage divider, e0 becomes low with ein above eref ,and high when ein is less than eref .

Hysteresis is obtained with positive feedbackthrough resistor R3 (see Fig. 3-51b). With e0 large,eref is determined by the voltage divider consisting ofR2 in series with the parallel combination of R1 and

R3. This voltage is higher than with just R1 and R2: e0approaches 0 when ein exceeds the eref . This causes erefto be lowered to a potential determined by the dividerrelationship of R1 in series with the parallel combina-tion of R2 and R3. Thus, the voltage at which e0switches from high to low is greater than that when itswitches from low to high. This is illustrated in theexample in the lower half of Figure 3-51b.

6.2.9 Active Filters

A typical active filter unit is shown in Figure 3-52.High ‘‘Q’’ circuits, different gains, and resonantfrequencies are easy to obtain. Inductance is not usedin these circuits. The filters can be cascaded and areunaffected by loading.

Figure 3-50 Phase shift units.

Figure 3-49 A differentiator and high-pass filter unit.


6.3 Relay Applications of Operational Amplifier

Three protective relay applications illustrate the use ofthe basic operational amplifier units described. Therelaying inputs from current and voltage transformersare converted to low-level signals by shunts orauxiliary transformers.

6.3.1 Instantaneous Overcurrent Unit

An operational amplifier instantaneous overcurrentunit is shown in Figure 3-53. Input current i isconverted to a proportional voltage through shunt Rand filtered by an active bandpass filter (OA1). Thepickup at other than the system frequency is signifi-cantly higher to minimize harmonic effects. OA2 is anadjustable gain inverting amplifier with a gain of

�K ¼ P1

R4

The amplified signal �Ki is rectified by OA3 and OA4.When the signal from OA2 is negative, OA3 forces itsoutput positive, while the input to the (þ) terminal ofOA4 will be negative through R7. This back-biasesdiode D1 to disconnect the OA3 output to OA4. OA4acts as a voltage follower with its output negative andfollowing the (þ) terminal input. When OA2 outputgoes positive, the output of OA3 goes negative andapplies a negative input to the (þ) terminal of OA4through D1. With R5 equal to R6, OA3 is a unity gaininverter with the input to OA4 negative when the OA3input is positive.

Figure 3-51 Level detector units.

Figure 3-53 An instantaneous overcurrent unit.Figure 3-52 A multiple-feedback band-pass filter unit.

68 Chapter 3

6.3.2 Sequence Networks

Sequence networks can be designed by using opera-tional amplifiers. A negative sequence circuit is shownin Figure 3-54. From Chapter 2, we have

I2 ¼ 1

3ðIa þ a2Ib þ aIcÞ

With phase-shift units (Fig. 3-50), Ib is shifted 2408 andIc 1208. With adder units, the final output is

e0 ¼ R

3ðIa þ Ibff240�� þ Icff120�Þ

and with R ¼ 1O,

e2 ¼ e0 ¼ 1

3ðIa þ a2Ib þ a1cÞ

Positive sequence and composite filters can be designedfollowing the same techniques.

6.3.3 Threshold Squarers and Square-WaveDetectors

These basic units are used in phase-comparison pilotsystems. A typical circuit is shown in Figure 3-55. Theoutputs provide square waves at a low level for keyingto a remote terminal, and at a high level for localcomparison.

The top circuit X has an adjustable noninvertingamplifier and a level detector. If the sine wave is ofsufficient magnitude to exceed the level detector setting(R4 and R5), the level detector output switches to 0during the positive half-cycle as shown in the wavetraces. P1 determines the magnitude at which theoutput switches. At low currents, the output remains 1,whereas at high currents, the output is a square wave.

The middle circuit Y is similar to X except that aninverting amplifier is used to provide a positive outputfrom a negative input current. At high currents, the

Figure 3-54 An operational amplifier negative sequence

network.

Figure 3-55 A threshold squarer and square-wave detector.


level detector switches from 1 to 0 during the negativehalf-cycle input.

The lower circuit Z generates a symmetrical squarewave at low currents. At no current, it has a 1 output.The circuit is similar to the X circuit. P3 is much largerthan R3, and R8 much larger than R9 to provide ahigh-gain and low-level detector-switching voltage.Square-wave detection is accomplished with theoperational amplifier timing circuit whose output isD. Resistor R11 should be greater than R10 so that thecapacitor C discharge rate is less than the charge rate.With no square wave (i.e., no current), then capacitorC will remain fully charged, causing the level detectorfollowing it to remain switched in the low state.However, if a square wave exists, then the zerotransitions allow capacitor C to discharge below the

level detector threshold, causing output D to become 1.When the square wave becomes 1 during the oppositehalf-cycle, the charge rate of R11, C, is high enough toprevent further switching of the level detector.

7 MICROPROCESSOR ARCHITECTURE

Relay design has evolved from electromechanical andsolid-state to microprocessor. The functions of electro-mechanical sensing units, sequence networks, andsolid-state logic units are performed through theprocessing of digital signals. The microprocessorcomponent, integrated with RAM and ROM devices,and software programs make up the basic unit inmicroprocessor relay design.

70 Chapter 3

4

Protection Against Transients and Surges

W. A. ELMORE

1 INTRODUCTION

The sporadic damped phenomena that occur inelectrical systems are generally described as transientsand surges. In this book, the two terms are consideredsynonymous and will be used interchangeably. In somereferences, however, transients refer to those phenom-ena related to lumped system parameters; surges referto those phenomena related to distributed parameters.For any disturbance in an electrical circuit, such as theopening or closing of a switch or breaker, theassociated damped transients may be either oscillatoryor unidirectional. Surges also appear as travelingwaves with a distinct propagation velocity. In suchcases, wave reflections may produce voltages substan-tially greater than the forcing voltage that initiated thephenomenon. Lightning surges must be considered aswell. With rare exception, however, experience indi-cates that only high-voltage systems need be protectedagainst lightning.

From a relaying standpoint, the effect of transientsand surges on secondary control circuits is of principalimportance. Primary transients affect secondary cir-cuits through common electrical connections, such as‘‘ground’’ circuits and electrostatic or electromagneticinduction, as well as current and voltage transformers.

1.1 Electrostatic Induction

A simplified version of electrostatic pickup is shown inFigure 4-1. An error signal is introduced into the‘‘signal lead’’ via the mutual coupling capacitance CM.

The magnitude of the coupled voltage VL is CM/(CMþCG) per unit of Vn, as long as RL and RS arevery high. Vn is the effective noise voltage and RR theeffective load resistance of the noisy lead. The lowerRL and RS are, the lower the transient voltage. If RL

and RS are so low that their effect predominates, thevoltage on the signal lead becomes approximatelyRTCM (dVn / dt), where RT is the parallel equivalent ofRL and RS, and dVn / dt the rate of change of the noisevoltage. The voltage on the signal lead cannot howeverexceed CM/(CMþCG) per unit, regardless of the rateof change of the noise voltage.

In some systems such as those used in solid-staterelaying, where negative—rather than ground—is the‘‘common,’’ the equivalent circuit is that shown inFigure 4-2a. The basic circuit is rearranged in Figure4-2b.

Figure 4-1 Equivalent circuit for electrostatic induction

with common ground return.

71

1.2 Electromagnetic Induction

Figure 4-3 illustrates electromagnetic pickup. Flux-linking of the signal pair, resulting from current flow inan adjacent circuit, induces a false signal voltage. The

total induced loop voltage is MdI/dt, whereM is theeffective mutual impedance between the two circuitsand dI/dt the rate of change of current I. Transposingthe signal circuit will reduce the induced voltage, asshown in Figure 4-4.

1.3 Differential- and Common-ModeClassifications

Surges can be classified into two modes: differential(also known as normal or transverse) and common(also known as longitudinal).

Differential-mode surges produce voltage on a pairof conductors in the same way as a legitimate signal.Differential-mode signals are illustrated in Figures 4-1,4-2, 4-3, 4-8, and 4-9.

Common-mode surges produce equal voltages on apair of conductors, with respect to some commonreferences. Common-mode surges are generated asshown in Figure 4-7. Common-mode voltage is alsoproduced by the circuit shown in Figure 4-2 if CM1

equals CM2 and CG1 equals CG2.Differential-mode surges are more likely to produce

misoperation of equipment, whereas common-modesurges are more likely to produce dielectric failure.(Note also that purely common-mode surges, whenapplied to unbalanced circuits, will produce a differ-ential-mode component and vice versa.)

Figure 4-2 Equivalent circuit for electrostatic induction

without common ground return.

Figure 4-3 Electromagnetic induction.

Figure 4-4 Transposing the signal circuit to minimize

electromagnetic induction.

72 Chapter 4

2 TRANSIENTS ORIGINATING IN THE HIGH-VOLTAGE SYSTEM

2.1 Capacitor Switching

Primary circuit transients are frequently generated bycapacitor switching and are substantially more severewhen interruption is accompanied by restriking.

2.1.1 Single-Bank Capacitor Switching

Figure 4-5 demonstrates what happens when acapacitor bank is energized by closing a switch. Ahigh-frequency, high-magnitude current I flows. Thecapacitance of the bus and connected apparatus causesthe same phenomenon to occur when a bus section isenergized. Unless precautions are taken to avoidtransients, such switching can cause 5- to 6-kV peaksin secondary circuits. Figure 4-6 illustrates whathappens if restriking occurs when a capacitive currentis interrupted.

At the instant of interruption (current zero), fullvoltage VC is trapped on the capacitor bank. Thisvoltage cannot change unless further current flows.The source voltage VB, however, continues to varysinusoidally. If the interruption cannot support therecovery voltage, a restrike occurs and current flowsagain. (The recovery voltage is VC�VB.) The mostunfavorable instant of restriking is shown in Figure 4-6.With continuity reestablished, VC equals VB. In the

process of equalization, considerable overshoot occurs,and both VB and VC approach three times normal line-to-neutral peak voltage. The current flow, immediatelyfollowing the restrike, is also very high. The currentoscillates at the natural frequency of the circuit anddecays with time, as governed by the circuit timeconstant.

2.1.2 Back-to-Back Capacitor Switching

Back-to-back capacitor switching consists of theenergization of one bank of capacitors adjacent to apreviously energized bank. Back-to-back capacitorswitching is much like the energization of a singlebank, except that the effective inductance is generallyvery much lower. The capacitance, on the other hand,is only somewhat lower, since it is the series combina-tion of the bank capacitance and capacitance of theunit or units that were energized before the switch wasclosed. For these reasons, the magnitude and fre-quency of the current are generally much higher forback-to-back energization than single-bank energiza-tion.

2.2 Bus Deenergization

Bus dropping is similar to capacitor bank deenergiza-tion, except the capacitance C is very much smaller.

Figure 4-5 Transients generated by energizing a capacitive

circuit.

Figure 4-6 Transients generated by opening a capacitive

circuit.

Protection Against Transients and Surges 73

Current magnitude is also generally smaller, and thefrequency is higher. When a simple disconnect is usedto drop the bus, the nonlinearity and prolongedexistence of the restriking arc cause significantelectrical noise. These characteristics together withthe large voltages and currents that accompanyrestrikes produce one of the most severe surgeinfluences in a substation. Surges of up to 8 kV havebeen measured in secondary circuits during disconnectarcing.

2.3 Transmission Line Switching

Transmission line switching is also similar to capacitorbank switching, except for the distributed nature of theinductance and capacitance of the line. The inrushcurrent tends to be substantially less than that forcapacitor bank switching. Frequency is inverselyproportional to the length of the transmission line.

2.4 Coupling Capacitor Voltage Transformer(CCVT) Switching

These transformers contain capacitance voltage-divid-ing networks. After energization, deenergization, andrestriking, they are subjected to the same high-frequency, high-current phenomenon experienced inthe other cases of lumped capacitance switching. Evenin a well-designed capacitance voltage transformer,there is perceptible capacitance between the high-voltage and low-voltage windings (Fig. 4-7). At thehigh frequencies associated with capacitor deviceswitching, the impedance of this capacitance will besmall. A surge voltage is developed during disconnectrestriking around the path g-g0-p-q-x or y and isroughly equivalent toL di/dt�M di/dt+Ri. (L andR are the inductance and resistance of the ground leadof the voltage device, and M is the mutual impedancebetween the ground lead and voltage leads.) If Mequals L, the total surge voltage reduces to Ri. Inpractice, M can never equal L, but it will approach it ifthe potential leads are placed as close as possible tothe ground lead. This arrangement will lessen thetransient voltage between the voltage leads andground.

Since voltage transformers are inductive devices,they are not subject to this phenomenon.

2.5 Other Transient Sources

Many other switching-type operations generate tran-sients: unequal pole-closing of a circuit breaker, faultoccurrence, fault clearing, load-tap changing, linereactor deenergization, series capacitor gap flashingand reinsertion, and so forth. In general, the peakmagnitude of such transients is substantially less thanfor the phenomenon described above.

3 TRANSIENTS ORIGINATING IN THE LOW-VOLTAGE SYSTEM

3.1 Direct Current Coil Interruption

During interruption of an inductive circuit, such as arelay coil, the L di/dt effect may produce a largevoltage across the coil (Fig. 4-8). In general, thevoltage will be greatest at the instant of interruption.Voltage magnitude will generally be independent of thesupply circuit characteristics and equal to the differ-ence between the extinction voltage of the interruptingcontact and the battery voltage. The surge voltage

Figure 4-7 Surge in secondary leads during disconnect

switch restriking on a capacitance voltage transformer.

74 Chapter 4

increases as a function of the speed with which theinterruptor forces current zero. Although voltages inexcess of 10 kV have been generated across 125-V coilsin laboratory tests. 2.5 kV is a more typical value.

3.2 Direct Current Circuit Energization

Energizing a circuit that is capacitively coupled toadjacent or nearby circuits can produce a transient inthe latter circuits (Fig. 4-9). When switch 1 is closed,VR appears as a false signal across the effectiveresistance of the adjacent circuit. Initially, full batteryvoltage appears across the coupled circuit. This voltagethe decays exponentially, in accordance with the RCtim constant.

3.3 Current Transformer Saturation

Current transformer saturation, which may producever high secondary voltage, is caused by high primarycurrent, poor current transformer quality, or excessive

burden. The surge repeats during each transition fromsaturation in one direction to saturation in the other.The voltage appearing at the secondary consists high-magnitude (possibly several kV) spikes with alternatingpolarity that persist for a few milliseconds eve half-cycle.

3.4 Grounding of Battery Circuit

When a ground occurs on the dc system, thedistributed and lumped capacitance of a system maycause sensitive devices to operate. Figures 4-10 and4-11 illustrate trip circuit behavior in the event of anaccidental ground. Comparable phenomena can causesensitive close circuits and tripping relays to malfunc-tion.

4 PROTECTIVE MEASURES

4.1 Separation

4.1.1 Physical Separation

Noise in critical circuits can be controlled effectively byphysically separating quiet and noisy circuits. Sincemutual capacitance and mutual inductance are inverselogarithmic functions of distance, small increases indistance produce substantial decreases in circuitinteraction.

Figure 4-9 Transients produced in adjacent circuits by dc

circuit energization.

Figure 4-8 Transients produced by interruption of an

induction circuit.


Similarly, control circuits should be routed perpen-dicular to noisy circuits. For example, a cable ductshould be run perpendicular to a high-voltage buswhen possible. Another way of effectively controllingsurges is to group circuits with comparable sensitiv-ities. Low-energy-level circuits, especially, should begrouped together and placed as far as possible frompower circuits.

4.1.2 Electrical Separation

Circuits can, of course, also be separated electrically.For example, surges can be controlled by thediscriminate application of inductance to block con-duction of high-frequency transients into protectedregions. This principle is illustrated by the filter circuitshown in Figure 4-12. High-frequency transients arediverted harmlessly to ground.

Transformer isolation (Fig. 4-13) puts an effectivecommon-mode barrier between segments of a system.High capacitance from each winding to ground andlow capacitance from winding to winding furtherreduce common-mode interaction between windings.

Figure 4-11 Accidental ground on trip lead.

Figure 4-12 Choke coil isolation.

Figure 4-13 Transformer isolation for common mode

voltage.

Figure 4-10 Accidental ground on battery positive.

76 Chapter 4

4.2 Suppression at the Source

4.2.1 Resistor Switching

Transient voltages can be kept comparatively low byequipping disconnects and circuit breakers withresistors that are inserted during operation of thedevice. For reasons of economy, this arrangement isoccasionally used to restrict the surge level in substa-tions.

4.2.2 Parallel Clamp

The surge associated with coil interruption can bevirtually eliminated by paralleling the coil with a zenerdiode. Where an extended dropout time is undesirable,a varistor may be substituted for the zener diodearrangement. Although the varistor allows a highersurge than the zener diode, its limiting action issatisfactory.

The zener diode Z1 in Figure 4-14 performs a dualsurge function. First, it minimizes the inductive ‘‘kick’’produced by the deenergization of the auxiliary coil.Also, when one of the contacts (Trip A in Figure 4-14)closes, the voltage induced in TC BKR A resultingfrom the interruption of current flow when 52a Aopens cannot cause the auxiliary relay to pick upundesirably. Current path I will have no detrimentaleffect. Zener Z1 will allow forward voltage of onlyapproximately 0.7V, which is insufficient to operatethe auxiliary relay. This scheme prevents undesirabletripping of breaker B.

The transient associated with extreme ac saturationof a current transformer can also be squelched byintroducing a voltage-limiting device across the

secondary. Silicon carbide devices can be used in thisprotective function.

4.2.3 Suppression by Termination

Figures 4-1, 4-2, 4-3, and 4-9 illustrate the value ofreduced input impedance RL in restricting the magni-tude and/or duration of transients. However, if RL isreduced, the energy requirement for operation isincreased, and more heat is generated when legitimateinputs are applied.

A small capacitor offers another method of redu-cing input impedance at high frequency, with littleeffect at 50 or 60Hz or on dc. This device neitherrequires a higher input energy for operation norgenerates heat. One such widely used capacitor is a0.01-mf ceramic capacitor. It limits a 2500-V, 1-MHzsurge, with a 150-O source to 350V peak to peak.Short leads to the capacitor are imperative.

When sensitive relays, trip circuits, and close circuitsexist in a substation, the capacitance on the dc must berestricted if false operations are to be avoided when aground occurs.

4.3 Suppression by Shielding

A signal lead that is shielded and has one or moregrounds will have increased capacitance to ground CG

(Fig. 4-1). For high RS and RL values, this increase incapacitance to ground reduces the ‘‘false’’ signalvoltage VL that results from the presence of anadjacent noisy lead. If a shield were used in Figure4-2, it would surround the signal lead and the commonnegative and would be grounded in one or morelocations. This arrangement tends to force CM1=CG1 toequal CM2=CG2, and any capacitively induced signalvoltage across RL to be 0.

Grounding a shield at both ends allows shieldcurrent to flow. Shield current resulting from magneticinduction will tend to cancel the flux that created it.The net effect of the shield on the signal lead is toreduce the noise level. Both ends should not begrounded if the shield is the signal return path.

4.4 Suppression by Twisting

Measures that cause the signal and return leads tooccupy essentially the same space minimize the effectof differential-mode coupling (Fig. 4-4). As shown bythe polarity marks, twisting a pair of leads cancels theeffect of adjacent circuit flux. Also, twisting the signalFigure 4-14 Zener Z1 applied for surge suppression.


lead and negative causes CM1 to equal CM2, and CG1 toequal CG2 (Fig. 4-2). This technique substantiallyreduces the influence of the adjacent noisy lead.

A combination of shielding and twisting effectivelyminimizes the influence of surges in adjacent circuits.For circuits properly treated with SPP capacitors at theterminal blocks, shielding is not required for staticrelaying circuits inside a panel or switchboard. TheseSPP capacitors are 750-V dc oil-filled capacitors.Shielded twisted pair conductors are required forlow-energy-level circuits routed outside a panel.

One lead of the shielded twisted pair is normally thesignal lead. The other lead (except where it is sensingcontact status) connects the negatives of the twodevices. Within a panel, electrostatic coupling is theonly significant intercircuit transient influence. A singleground on the shield, therefore, is sufficient. Forconsistency, ground should be at the input end.

4.5 Radial Routing of Control Cables

Circuits routed into the switchyard from the controlhouse should not be looped from one piece ofswitchyard apparatus to another with the returnconductor in another cable. Rather, all supply andreturn conductors should be in a common cable. Thisarrangement avoids the large EMI (electromagneticinduction) associated with the large flux loop thatwould otherwise be produced.

4.6 Buffers

Another effective method of delaying and desensitizinga circuit is to use a buffer (Fig. 4-15). Without causingthe transistor to conduct or damaging any element,this buffer can accommodate a test source operating at1 to 1.5MHz with a 150-O source impedance directlyacross the input (differential mode) and a 2500-V

(open circuit) first peak, which decays to 1250V in 6 ormore msec. The buffer can also withstand a sustained 7-Vdc input, or a high dc input voltage of sufficientduration to produce a minimum 4000-msec-V product(for example, 20V for 200 ms).

Buffering low-energy-level circuits greatly decreasesthe susceptibility of static relays to surge damage ormalfunction and, in general, eliminates the need forshielding circuits inside a relaying panel.

4.7 Optical Isolators

Optical isolators can provide excellent electricalseparation between two circuits. Figure 4-16 describesan input isolator. When the LED is conducting, as aresult of a voltage being applied at the input, base driveis provided for the phototransistor and it conducts,supplying a ‘‘logical’’ input to the protected equip-ment.

Similar isolation is accomplished for an output by acircuit such as that of Figure 4-17.

Figure 4-15 A standard input buffer circuit for solid-state

relays.

Figure 4-16 Optically isolated input.

78 Chapter 4

4.8 Increased Energy Requirement

Surges can also be endured by raising the thresholdvoltage or energy level at which operation occurs. Theequivalent circuits of Figures 4-10 and 4-11 show thathalf-maximum battery voltage, applied through theappropriate capacitance, must not be allowed to trip the

breaker or operate any other devices. An auxiliary relaydesigned to pick up at 71V or more will not respond toa single ground on a dc circuit with a maximumoperating voltage of 140V, regardless of the magnitudeof the capacitances on the system. (Note that thehigher-voltage design will not solve the problem shownin Figure 4-9; the higher-energy-level restriction will.)

Figure 4-17 Optically isolated output.


5

Instrument Transformers for Relaying

W. A. ELMORE

1 INTRODUCTION

Instrument transformers are used both to protectpersonnel and apparatus from high voltage and toallow reasonable insulation levels and current-carryingcapacity in relays, meters, and instruments. Instrumenttransformer performance is critical in protectiverelaying, since the relays can only be as accurate asthe information supplied them by the instrumenttransformers. Standard instrument transformers andrelays are normally rated at 5 or 1A; 100, 110, or120V; and 50 or 60Hz.

Where the relays operate only on current or voltagemagnitude, the relative direction of current flow in thetransformer windings is not important. Relativedirection (and, therefore, polarity) must be known,however, where the relays compare the sum ordifference of two currents or the interactions of severalcurrents or voltages. The polarity is usually marked onthe instrument transformer but can be determined ifnecessary.

2 CURRENT TRANSFORMERS

One major criterion for selecting a current transformerratio is the continuous current ratings of the connectedequipment (relays, auxiliary current transformers,instruments, etc.) and of the secondary winding ofthe current transformer itself. In practice, with loadcurrent normally flowing through the phase relays ordevices, the ratio is selected so that the secondaryoutput is around 5A (or 1A) at maximum primary

load current. When delta-connected current transfor-mers are used, the

ffiffiffi3

pfactor must be considered.

Although the performance required of currenttransformers varies with the relay application, high-quality transformers should always be used. Thebetter-quality transformers reduce application pro-blems, present fewer hazards, and generally providebetter relaying. The quality of the current transformersis most critical for differential schemes, where theperformance of all the transformers must match. Inthese schemes, relay performance is a function of theaccuracy of reproduction—not only at load currents,but also at all fault currents as well.

Some differences in performance can be accommo-dated in the relays. In general, the performance ofcurrent transformers is not so critical for transmissionline protection. The current transformers shouldreproduce reasonably faithfully for faults near theremote terminal, or at the balance point for coordina-tion or measurement.

2.1 Saturation

For large-magnitude, close-in faults, the currenttransformer may saturate; however, the magnitude offault current is not critical to many relays. Forexample, an induction overcurrent relay may beoperating on the flat part of the curve for a large-magnitude, close-in fault. Here it is relatively unim-portant whether the current transformer current isaccurate, since the timing is essentially identical. Thesame is true for instantaneous or distance-type relaying

81

for a heavy internal fault well inside the cut-off orbalance point. In all cases, however, the currenttransformer should provide sufficient current tooperate the relay positively.

2.2 Effect of dc Component

The presence of dc in the primary current can beparticularly detrimental to ct performance. Thisphenomenon is described in Section 6, ‘‘Direct CurrentSaturation.’’ Having a dc component in fault currenton an ac power system is a decaying phenomenon. If act is going to saturate due to the dc component of faultcurrent, it will do so in the first few cycles. Until thiseffect takes place, the fidelity of transformation isreasonably good, and instantaneous overcurrent anddistance relays may perform their task before the ctperformance collapses.

Following the disappearance of the dc componentof fault current, the behavior of the ct will againimprove. The error of transformation may then bepredicted by using one of the methods described inSection 4.

3 EQUIVALENT CIRCUIT

An approximate equivalent circuit for a currenttransformer is shown in Figure 5-1. Current is steppeddown in magnitude through the perfect (no-loss)transformation provided by windings ab and cd. Theprimary leakage impedance (ZH) is modified by n2 torefer it to the secondary. The secondary impedance isZL; Rm and Xm represent the core loss and excitingcomponents.

This generalized circuit can be reduced, as shown inFigure 5-1b. ZH can be neglected, since it influencesneither the perfectly transformed current IH/n nor thevoltage across Xm. The current through Xm, themagnetizing branch, is Ie, the exciting current. The Rmbranch produces a negligible influence.

The phasor diagram, with exaggerated voltagedrops, is shown in Figure 5-1c. In general, ZL isresistive and ZB is resistive or has a lagging angle. Ielags Vcd by 908 and is the prime source of error. Notethat the net effect of Ie is to cause IL to lead and besmaller than the perfectly transformed current IH/n.

Any simple equivalent diagram for a currenttransformer is, at best, crude. Exciting current isaccompanied by harmonics that, in turn, produceharmonic relay currents. An analysis for application

purposes is usually made on the basis of sinusoidalfundamental quantities. Although this approach ishighly simplified, the equivalent diagram is an excellenttool for picturing the phenomenon and estimating theapproximate performance to be expected.

4 ESTIMATION OF CURRENT TRANSFORMERPERFORMANCE

A current transformer’s performance is measured byits ability to reproduce the primary current in terms ofthe secondary; in particular, by the highest secondaryvoltage the transformer can produce without satura-tion and, consequently, large errors. Current trans-former performance with symmetrical (no dc) primarycurrent can be estimated by

FormulaThe current transformer excitation curvesThe ANSI transformer relaying accuracy classes

The first two methods provide accurate data foranalysis; the latter gives only a qualitative appraisal.All three methods require determining the secondary

Figure 5-1 The equivalent circuit and phasor diagram of a

current transformer.

82 Chapter 5

voltage Vcd that must be generated

Vcd ¼ VS ¼ ILðZL þ Zlead þ ZBÞ ð5-1Þwhere

VS = rms symmetrical secondary induced voltageFig. 5-1)

IL = maximum secondary current in amperes(symmetrical)

ZB = connected burden impedance in ohms

ZL = secondary winding impedance in ohms

Zlead = connecting lead impedance in ohms

4.1 Formula Method

The formula method uses the fundamental transformerequation

VS ¼ 4:44 fANBmax10�8ðvoltsÞ ð5-2Þ

where

f = frequency in hertz

A = cross-sectional area of the iron core insquare inches

N = number of turns

Bmax = flux density in lines per square inch

Both the cross-sectional area of the iron and itssaturation density are sometimes difficult to obtain.Current transformers generally use silicon steels, whichsaturate from 77,500 to 125,000 lines/in.2. The lowerfigure is typical for current transformers built before1947; a value of 100,000 is typical of most transformers.

The formula method consists of determining VS

using Eq. (5-1), then calculating Bmax using Eq. (5-2).If Bmax exceeds the saturation density, there will beappreciable error in the secondary current.

Assume, for example, that a 2000:5, high-perme-ability silicon steel transformer has 3.1 in.2 of iron anda secondary winding resistance of 0.31O. The max-imum current for which the current transformer mustoperate is 40,000A at 60Hz. The relay burden,including the secondary leads, is 2.0O. Will thiscurrent transformer saturate?

If the current transformer does not saturate, thesecondary current IS would be 40,000 divided by 400,or 100A, since N equals 400. Thus, the currenttransformer should be able to produce a secondary

voltage VS of 100(2.0þ 0.31), or 231V. Equation (5-2),solved for Bmax, will determine whether the currenttransformer can reproduce this current

Bmax ¼ 2316108

4:4466063:16400¼ 70;000 lines=in:2

Therefore, the current transformer should have ironthat will not saturate below 70,000 lines/in.2. Since thecurrent transformer in this example uses high-perme-ability silicon steel, it will not saturate with symme-trical primary current.

4.2 Excitation Curve Method

A typical excitation curve for a current transformer isshown in Figure 5-2. These data represent rms currentsobtained by applying rms voltage to the currenttransformer secondary, with the primary open-circuited. The curve gives the approximate excitingcurrent requirements for a given secondary voltage.

Figure 5-2 Exitation curves for a multiratio bushing current

transformer with an ANSI accuracy classification of C100.

Instrument Transformers for Relaying 83

With this method, a curve relating primary currentto secondary current can be developed for the tap, leadlength, and burden being used (Fig. 5-3). Any value ofprimary current can then be entered on the curve todetermine the expected value of secondary current.

The following examples will illustrate some of theproblems encountered in estimating current transfor-mer performance using the excitation curve method.

Example 1 Phase Relays

The breaker has a multiratio 600:5 bushing currenttransformer and the feeder is protected with over-current relays. The relays should operate for approxi-mately 60A rms symmetrical primary current. Thetotal burden on the current transformer, including thecurrent transformer secondary resistance, is 1.6O perphase when the relays are on the 6-A tap, and 3O perphase on the 3-A tap. The excitation curve for thetransformer is shown in Figure 5-2.

One approach would be to use a current transfor-mer ratio of 60:6, or 10 (the 50;5 tap), to takeadvantage of the lower burden on relay tap 6:

N ¼ 10 turns

IL ¼ 6A to operate the relay

VS ¼ ILZ total ¼ 661:6 ¼ 9:6V

From the excitation curve for VS of 9.6V, Ie wouldbe 6A, and NIe equals 60. Therefore, the primarypickup current is

IH ¼ NIS þNIe ¼ 60þ 60 ¼ 120A

This value is considerably higher than the 60A desired.

In theory, when making the phasor addition, the angleof the burden and the exciting branches should betaken into account. This refinement is not necessary,however, since it is obvious from the curve of Figure5-2 whether or not the current transformer would beoperating in the saturated region.

An alternative approach would be to use a ratio of60:3, or 20 (the 100:5 tap), with the higher burden ofthe 3-A relay tap. If we use this ratio, N equals 20, andIS equals 3A to operate this relay:

VS ¼ ILZ total ¼ 363 ¼ 9V

From the excitation curve (Fig. 5-2), Ie equals 0.5A,and NIe is 10. The primary pickup current IH would be

IH ¼ 60þ 10 ¼ 70:0A

This value is closer, but still too high.Now suppose that the breaker has two sets of

current transformers, with the secondaries connectedin series. Then each current transformer carries one-half the burden, or 1.54O on the 3-A tap. This value isslightly more than one-half of 3.0O because of thesecondary resistance of the added transformer. Usingthe 100:5 tap, we obtain

N ¼ 20 turns

IL ¼ 3A

VS ¼ 361:54 ¼ 4:62V per transformer

Then, from Figure 5-2, we have

Ie ¼ 0:33

NIe ¼ 6:6

IH ¼ 3620þ 6:6 ¼ 66:6A

Although this alternative offers some improvement, IHis not as close to the desired 60A as might have beenexpected. In both cases, the current transformer isoperating on the straight-line part of the characteristic,making significant improvement difficult. On the otherhand, two 50:5 current transformers in series wouldshow a marked improvement over the 50:5 ratio. HereIH, calculated by the above methods, is 71A. Whilemuch better than 120A, this value is still not as goodas the 66.6-A pickup obtained using the two currenttransformers with a 100:5 ratio.

Similar evaluations can be made for other config-urations.

Figure 5-3 Excitation curve method.

84 Chapter 5

Example 2 Phase and Ground Relays

The following example, shown in Figure 5-4, willdetermine the minimum primary current that willoperate the phase and ground relays.

PHASE RELAYS. For the phase relays, the total phaseburden Z equals 0.68 plus 0.08, or 0.76O, where 0.08 isthe current transformer secondary resistance on the100:5 tap (N¼ 20):

IL ¼ 5A(to operate the relay on the 5-A tap)

VS ¼ 0:7665 ¼ 3:8V

From Figure 5-2, we get

Ie ¼ 0:28A

IH ¼NðIL þ IeÞ ¼ 20ð5þ 0:28Þ ¼ 105:6A primary

If we neglect the exciting current (Ie), this valuewould become 20 times 5, or 100A primary, whenusing the 100:5 current transformer ratio.

GROUND RELAYS. If we assume the ground currentflows only in phase a, the equivalent circuit is shown inFigure 5-5. To obtain 0.5O through the ground relay,with its impedance assumed here of 22O, 11V must beproduced across the ground relay. If we neglect thesmall unknown voltage across the phase relays, thisground relay voltage will appear across the phase-band phase-c current transformers to excite them fromthe secondary side. From Figure 5-2, an Ie of 0.6Adevelops 11V across these current transformers. Theaccuracy required, generally, does not warrantcorrection for the small phase relay drop. However,such a correction could be made on a trial-and-errorbasis.

Thus, the phase-a relay circuit must supply 0.6 plus0.6 plus 0.5, or 1.7A. Given the phase relay impedanceof 0.68O and current transformer impedance of 0:08O,or 0:76O total, the phase-a current transformer mustsupply

VS ¼ 11þ ð1:760:76Þ ¼ 12:3V

Ie ¼ 0:8A (from Fig. 5-2)

IL ¼ 1:7þ 0:8 ¼ 2:5A

IH ¼ 2:5620 ¼ 50A primary

Thus, 50A is required to operate the ground relay. Ifthe exciting requirements of the three current trans-formers had been ignored, the current required tooperate the ground relay would have been estimated tobe 0.5 times 20, or 10A primary. From this, it isapparent that such may be a significant factor.

Using the 200:5 tap on the current transformercould improve sensitivity here. Dramatic improvementwould also be possible if a modern, low-impedanceground relay were substituted.

4.3 ANSI Standard: Current TransformerAccuracy Classes

The ANSI relaying accuracy class (ANSI C57-13) isdescribed by two symbols—letter designation andvoltage rating—that define the capability of thetransformer.

The letter designation code is as follows:

Figure 5-5 Equivalent circuits and distribution of currents

for a ground fault with the connections of Figure 5-4 in

Example 2.

Figure 5-4 Connections for Example 2 illustrating calcula-

tion of current transformer performance.


C: The transformer ratio can be calculated.T: The transformer ratio must be determined by

test.

The C classification covers bushing current transfor-mers with uniformly distributed windings, and anyother transformers whose core leakage flux has anegligible effect on the ratio within the defined limits.

The T classification covers most wound-type trans-formers and any others whose core leakage flux affectsthe ratio appreciably.

The secondary terminal voltage rating (Fig. 5-6) isthe voltage that the transformer will deliver to astandard burden at 20 times normal secondary current,without exceeding a 10% ratio error.

Figure 5-6 shows the secondary voltage capabilityfor various C-class current transformers, plottedagainst secondary current. With a transformer inthe C100 accuracy class, for example, the transformerratio can be calculated, and the ratio error will notexceed 10% between 1 and 20 times normal secondarycurrent if the burden does not exceed 1:0Oð1:0O65A620 ¼ 100VÞ.

ANSI accuracy class ratings apply only to the fullwinding. When there is a tapped secondary, aproportionately lower-voltage rating exists on the taps.

4.3.1 Current Transformer Data

The following current transformer data, required forrelaying service application, should be supplied by themanufacturer:

Relaying accuracy classification.Mechanical and thermal short-time (1-sec) ratings.

Both ratings define rms values that the transfor-mer is capable of withstanding. For mechanicalshort-time ratings, the rms value is that of the accomponent of a completely displaced primarycurrent wave. The thermal 1-sec rating is the rmsvalue of the primary current that the transformerwill withstand with the secondary winding short-circuited, without exceeding the limiting tem-perature of 250 8C for 55 8C-rise transformers, or350 8C for 80 8C-rise transformers. The short-time thermal current rating for any period of 1 to5 sec is determined by dividing the 1-sec currentrating by the square root of the required numberof seconds.

Resistance of the secondary winding between thewinding terminals. Data should be presented in aform that allows the value for each publishedratio to be determined.

For T-class transformers, the manufacturer shouldsupply typical overcurrent ratio curves onrectangular coordinate paper. The plot shouldbe between primary and secondary current, overthe range from 1 to 22 times normal current, forall standard burdens up to the one that causes aratio error of 50% (Fig. 5-7).

Figure 5-6 ANSI accuracy standard chart for class C

current transformers.

Figure 5-7 Typical overcurrent ratio curves for a T class

current transformer.

86 Chapter 5

For C-class transformers, the manufacturer shouldalso supply typical excitation curves on log-logcoordinate paper. The plot should show excita-tion current and secondary terminal voltage foreach published ratio from 1% of the accuracyclass secondary terminal voltage to a voltage (notto exceed 1600V) that will cause an excitationcurrent of five times normal secondary current(Fig. 5-2).

The ANSI standard burden is defined with a 50%power factor. These standard ohmic burden values areidentified in Figure 5-6. When fewer than the totalnumber of turns are in use on the C-class currenttransformer, only a portion of that burden can besupplied without exceeding the 10% error. Maximumpermissible burden is defined mathematically by

ZB ¼ NPVcl

100ð5-3Þ

where

ZB¼ permissible burden on the current transformerNP¼ turns in use divided by total turnsVcl¼ current transformer voltage class

Standard relaying burdens are listed in Table 5-1.B-0.1, B-0.2, B-0.5, B-0.9, and B-1.8 are standardmetering burdens. These burdens are defined with a 0.9power factor.

The following example shows current transformercalculations using ANSI classifications: the maximumcalculated fault current for a particular line is 12,000A.The current transformer is rated at 1200:5 and is to beused on the 800:5 tap. Its relaying accuracy class isC200 (full-rated winding); secondary resistance is0:2O. The total secondary circuit burden is 2:4O at a60% power factor. Excluding the effects of residualmagnetism and dc offset, will the error exceed 10%? If

so, what corrective action can be taken to reduce theerror to 10% or less?

The current transformer secondary winding resis-tance may be ignored because the C200 relayingaccuracy class designation indicates that the currenttransformer can support 200V plus the voltage dropcaused by ct internal secondary resistance at 20 timesrated current, for a 50% power-factor burden. The ctsecondary voltage drop may be ignored then if thesecondary current does not exceed 100A:

N ¼ 800

5¼ 160

IL ¼ 12; 000

160¼ 75A

The permissible burden is given by (from Eq. 5-3)

ZB ¼ NPVcl

100

NP ¼ 800

1200¼ 0:667

Thus

ZB ¼ 0:667ð200Þ100

¼ 1:334O

Since the circuit burden, 2:4O, is greater than thecalculated permissible burden, 1:334O, the error willexceed 10% at the maximum fault current level (75A).Consequently, it is necessary to reduce the burden, usea higher current transformer ratio, or use a currenttransformer with a higher relaying accuracy classvoltage.

5 EUROPEAN PRACTICE

In Europe, current transformers are described in termsof protection and measurement classes. The protectionclasses are those of most interest in relaying, and theycarry the designation P, a maximum error of 5 or 10%,a corresponding volt-ampere burden, a rated current,and an accuracy limit factor. For example, a 30-VAclass, 5P10, 5-A ct is compatible with a 30-VAcontinuous burden at 5A. This corresponds to a 6-Voutput. It produces no more than 5% error at1066 ¼ 60V. The permissible burden is 30=ð565Þ ¼ 1:2O.

Three types of ct’s are defined by the TPX, TPY,and TPZ designations.

Table 5-1 Standard Relay Burden Designations

Characteristics for 60-Hz and 5-A secondary circuit

Standard

burden

designation

Impedance

(r) VA Power factor

B-1 1.0 25 0.5

B-2 2.0 50 0.5

B-4 4.0 100 0.5

B-8 8.0 200 0.5

(proportion of totalturns in use)


5.1 TPX

The TPX is a nongapped core current transformer witha 0.5% ratio error and secondary time constant of 5 secor more. It may be used with other TPX or TPY ct’s inall types of protection applications.

5.2 TPY

This ct has a gapped core and secondary time constantof 0 to 10 sec. It has a ratio error of +1% and largercost than the TPX. Its transformation of the dccomponent of fault current is not as accurate as theTPX. It may be combined with other TPX or TPY ct’sin any relaying application. Its advantage is that itsremanent flux is quite small compared to that of anongapped core ct.

5.3 TPZ

TPZ ct’s have a linear core with a secondary timeconstant of 60+6msec for 50-Hz and 50+5msec for60-Hz applications. This provides a very short dccollapse time, making the ct suitable for breaker failureapplications in which the overcurrent supervision issusceptible to dc influence. When used in combina-tions, it should be used only with other TPZ ct’s. It hasa +1% ratio error at the rated primary current.

6 DIRECT CURRENT SATURATION

To this point, current transformer performance hasbeen discussed in terms of steady-state behavior only,without considering the dc component of the faultcurrent. Actually, the dc component has far moreinfluence in producing severe saturation than the acfault current. The dc component arises because (1) thecurrent in an inductance cannot change instanta-neously and (2) the steady-state current, before andafter a change, must lag (or lead) the voltage by theproper power-factor angle.

Figure 5-8 shows the current immediately followingfault inception for two cases: fully offset and with nooffset. In the fully offset case, the fault is assumed tooccur at the instant that produces the maximum dccomponent. In the second case, the fault occurs at atime that produces no dc offset.

Figure 5-9 shows an example of the distortion andreduction in the secondary current that occurs as a

result of dc saturation. Note the improvement inperformance as the dc diminishes.

If VK � 6:28 IRT, the dc component of a faultcurrent will not produce current transformer satura-tion. In this expression,

VK¼ voltage at the knee of the saturation curve,determined by extending the straight-line por-tions of the curve to find their intersection

I¼ symmetrical secondary current in amperes rms

Figure 5-8 Current immediately after fault inception.

Figure 5-9 Direct current saturation of current transformer.

88 Chapter 5

R¼ total secondary resistance in ohmsT¼ dc time constant of the primary circuit in cycles

Here, T¼ (LP/RP)f, where

LP¼ primary circuit inductance in henriesRP¼ primary circuit resistance in ohmsf¼ frequency

Direct current saturation is particularly significantin bus differential relaying systems, where highlydiffering currents flow to an external fault throughthe current transformers of the various circuits.Dissimilar saturation in any differential scheme willproduce operating current.

Figure 5-10 shows how current transformer satura-tion relates to time. Severe current transformersaturation will occur if the primary circuit dc timeconstant is sufficiently long and the dc componentsufficiently high. Curves d, e, and f of Figure 5-10 showthat the dc component requires substantially greaterflux than that needed to satisfy the ac component.

Time is required to reach saturation flux density.This time can be estimated from Figure 5-11, asfollows:

From the current transformer excitation curve forthe tap in use, determine VK from the intersec-tion formed by extending the two straight-linesegments of the curve. Note that both axes musthave the same logarithmic scales, as is illustratedin Figure 5-2.

Calculate VK/IRT.Obtain t/T from Figure 5-11.Calculate t, the time to saturate.

VK must be modified if residual flux is to beconsidered. For example, with a residual flux of 90%,the saturation voltage value must be multiplied by 1minus 0.9, or 0.1, to determine the earliest time tosaturation. This will give a conservative value for timeto saturation.

7 RESIDUAL FLUX

Any iron-core device will retain a flux level even afterthe exciting current falls to 0. Superimposed on thisresidual flux are variations in core flux, dictated by thecurrent transformer secondary current and secondaryburden.

Figure 5-12 shows the importance of previousloading history on current transformer residual fluxlevel. Suppose a current transformer has a residual flux

defined by point a. If a symmetrical sinusoidal primarycurrent starts to flow, requiring a flux variation asshown, the pattern between a and A will be traced out.

Figure 5-10 Current transformer flux during assymetrical

fault (one cycle time constant).


The average value of the dc exciting current with thispattern is Ia. This current flows in the secondary andhas no counterpart in the primary. It decays with thetime constant associated with the secondary circuit. Atthe completion of this transient, the pattern has movedto cC with an equal flux variation that is symmetricalaround the vertical axis. The pattern continues to betraced out. If the circuit were now interrupted, theresidual flux would have the value existing at themoment of interruption, which is quite different fromthe initial value assumed. In fact, any value of fluxbetween 0 and the saturation level may be retained inthe core, depending on previous events.

Reducing the residual flux to 0 requires theapplication of a secondary voltage high enough toproduce saturation, followed by a gradual reduction ofthe voltage to 0.

A current transformer with an air gap in the corehas a fairly low residual flux, approximately 10% ofsaturation density. Residual flux for current transfor-mers with no intentional air gap is approximately 90%of saturation density (max).

Air-gap current transformers do not saturate asrapidly as devices without air gaps subjected to equalcurrent and burden. However, the magnetizing currentis higher, resulting in greater ratio and phase-angleerrors.

After interruption of the primary current, residualflux decays very slowly (taking approximately 1 sec),and the secondary current collapses slowly. Also, air-gap units are more costly to manufacture, since thesmall air gap must be both accurate and maintainable.

Figure 5-11 (a) Current transformer time to saturate. (b)

Expanded scale of (a).

Figure 5-12 Residual flux in current transformer.

90 Chapter 5

Although in theory residual flux can cause relayingproblems, there have been very few documented casesin which the residual flux has caused a relaymisoperation.

8 MOCT

The MOCT (magneto-optic current transducer)relieves many of the problems associated with currenttransformation. This device utilizes the Faraday effectto produce a high-accuracy analog output that is notinfluenced by iron saturation. The Faraday effect is therotation of the plane of polarization when plane-polarized light is sent through glass in a directionparallel to an applied magnetic field. This is illustratedin Figure 5-13. The angle of rotation is directlyproportional to the strength of the magnetic field.

In the application of this principle to currentmeasurement, the transmission line current is thesource of the magnetic field (see Fig. 5-14).The strength of the field is directly proportional tothe instantaneous current magnitude. By placing therotator (Faraday-effect sensor) in proximity tothe transmission line conductor and comparing theangle of rotation of a light beam, a voltage is generatedthat is directly proportional to current.

This voltage is then used as an input to protectiverelays at a level and in a way that is virtually identicalto that which is used when current transformers supplythe relay through a current-to-voltage transformation.

9 VOLTAGE TRANSFORMERS ANDCOUPLING CAPACITANCE VOLTAGETRANSFORMERS

Voltage transformers (formerly called potential trans-formers) and coupling capacitance voltage transfor-mers are selected according to two criteria: the systemvoltage level and basic impulse insulation levelrequired by the system on which they are to be used.Under ANSI, two nominal secondary voltages, 115and 120V, are allowed for voltage transformers; thecorresponding line-to-neutral values are 115=

ffiffiffi3

pand

120=ffiffiffi3

p. The applicable voltage depends on the

primary voltage rating, as given in ANSI C57.13.The nominal secondary voltages for coupling capaci-tance voltage transformers are 115 and 66.4V.

Most protective relays applied in the United Stateshave standard voltage ratings of 120 or 69V, depend-ing on whether they are to be connected line to line orline to neutral.

9.1 Equivalent Circuit of a Voltage Transformer

The equivalent circuit of a voltage transformer (vt) isshown in Figure 5-15. Since regulation is critical toaccuracy, the circuit may be reduced to that shown inFigure 5-15b. The phasor diagram of Figure 5-15c hasgreatly exaggerated voltage drops to emphasize that,for typical transformers and burdens, the secondaryvoltage usually lags the ‘‘perfectly transformed’’primary voltage and is deficient in magnitude. Typicalrated maximum errors for these devices are 0.3, 0.6,and 1.2%. Voltage transformers have excellent tran-Figure 5-13 Faraday rotator concept.

Figure 5-14 Faraday effect current sensing.


sient performance, faithfully reproducing abruptchanges in the primary voltage.

9.2 Coupling Capacitor Voltage Transformers

Coupling capacitor voltage transformers (ccvt’s) andbushing capacitor voltage transformers are less expen-sive than voltage transformers at the higher voltageratings, but may be inferior in transient performance.With these voltage devices, a subsidence transientaccompanies a sudden reduction of voltage on theprimary. This voltage may be oscillatory at 60Hz orsome other frequency, or it may be unidirectional. Arepresentative severe secondary transient is shown inFigure 5-16.

Figures 5-17 and 5-18 illustrate the source of thesubsidence transient in the ccvt. In Figure 5-18,elements L and C generally contain stored energywhen a disturbance, such as a fault, occurs on theprimary. Because of the ‘‘ringing’’ tendency inherent inthe RLC circuit, a sudden short circuit on the primarydoes not produce an instantaneous collapse of thevoltage applied to the relays. The extent and durationof the deviation from the perfectly transformed voltagedepend on the values of R, L, and C. Other transientsare introduced by the presence of ferroresonantsuppression circuits and the relays themselves.

A voltage transformer is not significantly affectedby comparable transients and will reproduce primarytransients with excellent fidelity. Modern ccvt’s, suchas the PCA-7, have capabilities approaching those ofthe voltage transformer.

The subsidence transient of the ccvt may influencethe behavior of some relays. Solid-state phase andground distance relays, used in a zone 1 direct tripfunction, may be seriously affected by the temporaryexcessive reduction of voltage during the decay period.These relays either must be designed with a specialprovision that allows the subsidence transient to beignored, be time-delayed to override the transientperiod or they must have their reach shortenedsufficiently to avoid false tripping.

Figure 5-15 The equivalent circuit and phasor diagram of a

voltage transformer.

Figure 5-16 A typical subsidence transient of older type

coupling capacitor voltage transformers.

Figure 5-17 Simplified schematic of a coupling capacitor


Figure 5-18 Equivalent diagram of a coupling capacitor


92 Chapter 5

Table 5-2 describes a group of ABB ccvt’s withdiffering costs and performance. The transientresponse column indicates, in general, the degree ofcompatibility with different relaying systems. High-speed, direct trip, restricted reach relays require eitherthe use of the fast response ccvt’s or provision in therelaying system to accommodate or eliminate the effectof the error. The transient response data are based onthe percentage of voltage remaining at the ccvt outputterminals at the times indicated following suddenreduction of primary voltage from rated voltage to 0,with initiation at zero crossing. The burden is asdefined in ANSI C93.1.

The impedance of capacitance voltage transformersshould not be high enough to produce erroneousbehavior in the static compensator distance relays.Excessive impedance may cause false tripping for areverse fault. For this reason, bushing voltage devicesrated below 230 kV should not be used with solid-statedistance relays. Bushing voltage devices are, in general,seldom used. Their burden capability is limited andtransient performance poor.

9.3 MOVT/EOVT

Voltage sensing also may be accomplished using fiber-optic technology. The MOVT uses the Faraday effectdescribed above, sensing the current flowing through acapacitor stack connected from line to ground.Another voltage-sensing device, the EOVT, uses aPockel cell rather than the Faraday rotator. Itsprinciple uses light from an optical fiber, which ispassed through a special crystal that produces equalcomponents in the X and Y directions. An electric fieldcauses one of these components to be retarded, andthis results in a phase difference between the twocomponents. This, in turn, changes the light intensityat the sensor fiber in proportion to the electric field.With additional refinements, it produces an analogoutput proportional to the electric field present, andthis is, in turn, proportional to the instantaneousmagnitude of the voltage at the point of measurement.

10 NEUTRAL INVERSION

Neutral ‘‘inversion,’’ in which ground becomes exter-nal to the system voltage triangle, can occur onungrounded systems with a single potential transfor-mer connected line to ground. Figure 5-19 shows thepossible voltage across an unloaded voltage transfor-mer. Note that an Xc/Xm ratio of 3 would theoreticallycause an infinite voltage across the voltage transfor-mer. Such a situation never occurs, of course, becauseXm reduces as saturation occurs.

By loading the transformer carefully, this large,sustained overvoltage phenomenon can be avoided.Caution should always be exercised when the second-ary voltage of the transformer is used for synchronismcheck, since the loading will cause a phase shift.

Figure 5-19 Neutral inversion on an ungrounded power

system.

Table 5-2 ABB ccvt’s

Type

Cost

p.u.

Steady-

state

accuracy

(%)

Max.

burden

(VA)

Capacitance

p.u.

(1 p.u.¼0.006 at

115 kV)

Transient

response

(% voltage)

8msec 16msec

PCA-5 — 1.2 200 1.0 23.5 12

PCA-7 1.57 1.2 200 4.1 7 6.2

PCA-8 1.0 1.2 200 1.0 23.5 12

PCA-9 1.94 0.3 400 4.1 15 10.5

PCA-10 1.26 1.2 200 4.1 17.5 11

PCM-X 2.51 0.3 400 4.1 12 9.5


6

Microprocessor Relaying Fundamentals

W. A. ELMORE

1 INTRODUCTION

From the power-system viewpoint, microprocessor ornumerical relays are not unlike electromechanical,solid-state, or digital relays. Currents and voltagesmust be measured and compared with set points, orwith each other, and action must be deferred orinitiated. Other inputs such as received carrier, 52bswitch position, choice of pilot system, etc. temper theaction. Figure 6-1 shows a simplified block diagram ofa typical microprocessor relay.

Numerical relays must work within the frameworkof data sampled moment by moment. Currents, forexample, are not treated on a continuous basis, butthey, like all of the input quantities, are sampled one ata time at as fast a speed as the data handling andstorage hardware can accommodate.

The microprocessor has afforded us in protectiverelaying the remarkable capability of sampling vol-tages and currents at very high speed, manipulating thedata to accomplish a distance or overcurrent measure-ment, retaining fault information and performing self-checking functions. The utilization of this newtechnology has also presented us with new challengesregarding the manner in which information is handledand manipulated.

With multiple electromechanical or solid-statedevices operating concurrently, time coincidence is noproblem. However, a microprocessor literally canhandle only one task at a time. Multiplexors cansample only one quantity at a time so voltages andcurrents are not time-coincident. The awesome task ofthe programmer is to accommodate these peculiarities

and devise ways for the microprocessor to accomplishthe tasks in the right order and to cause comparisonsto be made based on the correct voltages and currentswithout the error associated with data skew. Dataskew is introduced by the comparison of quantitiestaken nonsimultaneously.

To allow the sampling of a fixed quantity ratherthan a rapidly changing quantity, the sample-and-hold(S/H) circuit is usually used in numerical relays. Anexample of this is shown in Figure 6-2. Since even thesimplest relaying function requires that multiple inputsbe read, a multiplexer is used. This is a device thatallows all the input quantities to be sampled (read) oneat a time.

The microprocessor requires that the informationbe presented to it in digital form, usually an 8- or 16-bitword. The conversion process from the analog signal(which is simply a scaled dc quantity that is repre-sentative of the sampled quantity) to the digital signalis accomplished with an A/D converter. Many varietiesof these devices have been used over the years. Therange and sampling rate required dictate the choice fora particular design for a protective relay.

The microprocessor accepts the sampled data andstores it for future use in RAM (random accessmemory). The data are acted on by algorithms orcomparisons defined by the program memory, which isstored in ROM (read-only memory) or more widely inEPROM (erasable programmable read-only memory).The program stored in ROM or EPROM is non-volatile.

Another vital element in the architecture requiredfor microprocessor relaying applications is the

95

NOVRAM (nonvolatile RAM) or EEPROM (electri-cally erasable programmable read-only memory). Datathat are stored in this type of memory are not lostwhen power is removed from the relay. Settings andtarget data are usually stored here.

Microprocessor-based algorithms typically requiretime-coincident sampling of the input quantities.Considerable ingenuity is used to address this in real-

time processes, particularly in relaying applicationswhere so much is dependent on the time relationshipbetween quantities. There are two basic methods ofsampling data, both of which use a sample-and-holdcircuit. An example of one method is shown inFigure 6-3. Using separate sample-and-hold circuitsfor each input, the microprocessor directs a ‘‘freeze’’ tooccur at each sampling point. The S/H circuit holds

Figure 6-1 Simplified diagram of a typical microprocessor relay.

Figure 6-2 Typical sample and hold.

96 Chapter 6

this sampled value until the microprocessor can read ineach value through the multiplexor and A/D circuit.The microprocessor then directs the S/H circuit toresume the sampling process until the next freezesignal.

An alternative sampling method that is less expen-sive is to use a single S/H circuit for all inputs. This isshown in Figure 6-4. A time correction factor isapplied to each sample after the first in a group.

It is known precisely what difference in samplingtimes exist and therefore all the samples in a time-sequence-sampled (multiplexed) group can be con-verted to coincident samples by applying an anglecorrection. Of course, other methods can be used.

2 SAMPLING PROBLEMS

Because of the practical limitation of sampling rates ina numerical relay, a varying input such as an ac current

or voltage will be perceived by the relay considerablydifferently from its actual continuous waveform. Highfrequencies in the waveform cannot only fail to beidentified due to inadequacies in the sampling process,but may indeed present themselves as a lower-frequencycomponent. Once this error intrudes into the process, itcannot be reconstituted and removed. Either the errormust not be allowed to occur in the first place byfiltering out the offending frequencies or a process suchas asynchronous sampling must be used. The mechan-ism of a high-frequency component in an input wave-form manifesting itself as a low-frequency signal iscalled aliasing. It will now be described using a phasortechnique that may clarify the concept for the reader.

3 ALIASING

Figure 6-5 depicts the representation of a simplesinewave by a phasor as it rotates. The projection of

Figure 6-3 Microprocessor relay with individual sample/hold.

Figure 6-4 Basic hardware for microprocessor relay.

Microprocessor Relaying Fundamentals 97

the phasor on the vertical axis, at a given time,represents the magnitude of the sine wave at that time.Note that to do this the phasor must be represented byits peak value, not its rms value. Phasors are generallymanipulated, however, using rms values. They aregenerally shown for a single frequency. Figure 6-6extends this to show a fifth harmonic phasor super-imposed on the fundamental and the distorted sinusoidthat is generated by the combination. The fifthharmonic phasor, of course, rotates through 4508 inthe time required for the 60-Hz fundamental phasor torotate through 908.

Now the effect of high-frequency distortion of thewaveform on the sampling process can be examined.Figure 6-7 uses an example of a seventh harmonic andan 8-per-cycle sampling rate.

It can be seen that the normal circle for thefundamental is distorted into an ellipse by the presenceof the harmonic. Thus, when the total waveform isconstructed by the projection on the vertical, it can beseen to be deficient in magnitude. This is thephenomenon known as, aliasing. It is the appearanceof a high-frequency signal as a lower-frequency signalthat distorts the desired signal.

4 HOW TO OVERCOME ALIASING

4.1 Antialiasing Filters

This effect may be removed by filtering the high-frequency components from the input. The elementthat accomplishes this function is called an antialiasingfilter. The ‘‘Nyquist criterion’’ states that in order toavoid the aliasing error, frequencies above one-half thesampling rate must be removed. Figure 6-8 shows atypical antialiasing filter.

4.2 Nonsynchronous Sampling

When the luxury of time exists in the relay response, analternative to the antialiasing filter can be used.Nothing is lost in shifting the sampling points for thefundamental frequency component of the quantitybeing measured. For example, in Figure 6-9, it is not

Figure 6-5 Generation of a sine wave by a phasor.

Figure 6-6 Phasor representation for fundamental and fifth harmonic.

98 Chapter 6

critical that the first sample be taken at the zerocrossing. It may be taken at any arbitrary point withthe following samples being equally spaced at thesampling rate. If then, after collecting eight samples inthis case, a jump is introduced to delay the beginningof the collection of the next eight samples, by a timecorresponding to 1808 of a particular high-frequencyquantity (25.718 on a 60-Hz basis for the seventhharmonic), an interesting effect takes place. If themeasured quantity appears to be too low, as a result of

the presence of the harmonic, for the first fundamentalcycle (see Fig. 6-7), it will for the second cycle(following the jump) appear to be too high by thesame amount. Thus, a comparison of information inthe adjacent cycles allows the effect of the seventhharmonic to be removed.

If the speed requirement of the device allows, then itis possible to eliminate the error associated with aparticular harmonic without an antialiasing filter.Note, however, while other harmonic frequencycomponents in the input signal are attenuated by thisasynchronous sampling procedure, the effect of onlyone frequency is eliminated. This concept has beenused successfully in the MCO and MMCO relays.

5 CHOICE OF MEASUREMENT PRINCIPLE

With electromechanical relays, the designer has littlechoice with regard to whether electrical quantities are

Figure 6-7 Aliasing effect of seventh harmonic.

Figure 6-8 Antialiasing filter. (a) Diagram; (b) frequency

response. Figure 6-9 Waveform samples taken from a sinusoid.


to be interpreted in terms of peak value, average value,rms value, or fundamental frequency value. With thepower of the microprocessor, the designer can applyany of these measurement techniques.

5.1 rms Calculation

The digital determination of root-mean-square valuesof waveforms is quite similar to the conventionalanalytical methods. Equations (6-1) and (6-2) illustratethis. By squaring the magnitude of each sample (In)over a cycle and summing that with other squares,dividing the sum by the number of samples, and takingthe square root, an rms value can be extracted from acomplex waveform.

Analog rms ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

2p

Z 2p

0

I2m sin2 ot dt

sð6-1Þ

Digital rms ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

8

X8

n¼1I2n

rð6-2Þ

Time overcurrent devices that are to be coordinatedwith other apparatus which experience I2R heatingeffects (such as fuses, conductors, and transformers)have been developed with an rms response. Forapplications in which harmonic effects are generatedby apparatus such as six-pulse rectifiers, it may berequired that the harmonics be ignored. Relays havebeen designed using microprocessor techniques thatare responsive only to the fundamental frequencycomponent of the input waveform.

5.2 Digital Filters

Any periodic waveform can be represented by afundamental and series of harmonic frequencies. Anyparticular frequency can be extracted by utilizingEqs. (6-3) through (6-5):

an ¼ 2

T0

Z T0=2

�T0=2

fðtÞ cos not dt ð6-3Þ

bn ¼ 2

T0

Z T0=2

�T0=2

fðtÞ sin not dt ð6-4Þ

fnðtÞ ¼ an cosontþ bn sin ont ð6-5ÞThe sum of the product of the function and the sine ofthe frequency that is to be extracted, taken over theperiod of the fundamental, produces a total thatcontains only the desired frequency. This is thefundamental premise of Fourier analysis. This calcula-

tion coupled with a similar one using a cosine functioninstead of a sine allows the complete magnitude andphase position of the desired frequency to be obtained.

With this integration, any desired frequency, such asa 60-Hz fundamental, can be extracted from adistorted periodic waveform, and all other frequencieswill be excluded.

5.3 Fourier-Notch Filter

The comparable digital process involves the multi-plication of individual samples by stored values from areference sine wave and summing the products over afull cycle

AC ¼XN�1

K¼0

fKðtÞCAK

AS ¼XN�1

K¼0

fKðtÞCBK

where

CAK ¼ 2

Ncos 2p

K

N

� �

CBK ¼ 2

Nsin 2p

K

N

� �K¼ number of sampleN¼ samples per cycle

where

f(t)¼ original functionT0¼ period of the waveformn¼ order of the harmonic

Consider, for example, an application in which thereare eight samples per (fundamental 60-Hz) cycle. Thecorresponding samples of a sine wave may be chosenas 0, 0.707, 1.0, 0.707, 0, � 0.707, � 1.0, and � 0.707.These are fixed values, being sinðK2p=8Þ, where K arethe samples 0 to 7 and 2p=8 corresponds to 458. Thesevalues are then multiplied by 2/N to obtain theconstants that are used.

If then the sampled values of the measured quantityover a full cycle are multiplied by these constants in theproper order and summed, the process of Eqs. (6-3)and (6-4) is duplicated. This provides information thatexcludes all frequencies except the fundamental. Bythis process, any frequency component can be isolatedand utilized to perform a desired function. From thisprocess results a function As ¼ Ksinot.

Similarly, by using a set of cosine functionconstants, the sample multiplication and summation

100 Chapter 6

can generate a function Ac ¼ Kcosot. Sincesin2 otþ cos2 ot ¼ 1;A2

s þA2c ¼ K2. Thus, the peak

value of a particular frequency component can befound by taking the square root of A2

s þA2c. Also since

sinot= cosot ¼ tanot;As=Ac ¼ ðKsinotÞ=ðKcosotÞ¼ tanot:

The angle of the function can be found by

y ¼ tan�1 As

Ac

� �

This algorithm is called a Fourier-notch filter.

5.4 Another Digital Filter

Other forms of digital filtering are used for specificapplications. In the IMPRS series of relays, foursamples per cycle are used. These samples may begin atany point in the cycle, such as at angle f in Figure 6-9.The values of the individual samples can be describedas

S1 ¼ sinðotþ fÞS2 ¼ sinðotþ fþ 90�ÞS3 ¼ sinðotþ fþ 180�ÞS4 ¼ sinðotþ fþ 270�ÞS5 ¼ sinðotþ fÞ

Digital filtering can be accomplished with the follow-ing procedure:

Ss ¼ S1 � S2 � S3 þ S4 ¼ 2ffiffiffi2

psinðotþ f� 45�Þ

Similarly,

SC ¼ S2 � S3 � S4 þ S5 ¼ 2ffiffiffi2

pcosðotþ f� 45�Þ

These values were related to the simple sine wave ofFigure 6-9 with a peak value of 1.0 for a sine wave,Im sinðotþ fÞ. The value of Im, the peak value, can beobtained by

Im ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiS2s þ S2C

q2

ffiffiffi2

p

This digital filter also has other useful qualities. If weconsider a dc current, it is obvious that S1 � S2 � S3 þS4 ¼ 0 because each sample is the same magnitude.Similarly, a linearly decaying ramp wave shape willalso produce a sum equal to 0. Since the dc componentof fault current has an exponential decay which isbetween the constant dc input and linearly decaying

input, it also is severely attenuated by this process ofsumming.

Using four samples per cycle, this digital filterremoves all of the even harmonics. With 60-Hzwaveform samples taken at 908 intervals, the samplesof 120Hz would occur at 1808 intervals. Every othersample will be equal, so S1 � S3 ¼ 0 and S2 � S4 ¼ 0.The second harmonic is eliminated by the summationSs ¼ S1 � S2 � S3 þ S4. With the fourth harmonic, thesamples are taken at 3608 intervals and being equalalso produce Ss ¼ 0. All even harmonics are elimi-nated.

5.5 dc Offset Compensation

Direct current offset in the fault current occurs as aresult of two natural laws: (1) Current cannot changeinstantaneously in an inductance and (2) current mustlag the applied voltage by the natural power-factorangle of the system. dc offset produces no desirableeffects in overcurrent or distance relays. To make thesedevices responsive only to the ac component of faultcurrent, it is necessary to remove the dc by someexpedient.

The maximum dc component of the fault current isImð1� e�t=TÞ, where Im is the peak value of thesymmetrical ac fault current, t the time in cycles, andT the dc time constant of the circuit that limits the faultcurrent. The dc removal algorithm can be exact if T isknown. Unfortunately, for a given system it is likely tovary considerably.

Many algorithms have been used. One uses theconcept that a sample of the fundamental componentof current has the same magnitude as, and the oppositesign to, a sample taken 1808 later. The dc componentsfor each of these samples are the same (if we assumethis component is truly dc). Thus, for a 480-Hzsampling rate

Offset ¼ IK þ IK�4

2

where IK is the value of a sample of current and IK�4

the value taken four samples previously. With eightsamples per cycle, these samples would be 1808 apartand the effect of the sinusoidal component would benullified in the summation. The offset may then beused as a correction factor for the samples taken in thisinterval.

With a decaying dc as opposed to an unvaryingvalue, some error is introduced in this process,depending on the dc time constant.


5.6 Symmetrical Component Filter

Another interesting digital filter, utilizing three samplesper cycle, is embodied in the MPR relay. In mostapplications, time-coincident quantities are necessary,but as the following symmetrical component defini-tions suggest, quantities that are 120 or 2408 displacedin time are useful

IA1 ¼ 1

3ðIA þ aIB þ a2ICÞ

IA2 ¼ 1

3ðIA þ a2IB þ aICÞ

IA0 ¼ 1

3ðIA þ IB þ ICÞ

where

a ¼ 1ff120�

The normal analog process for extracting IA2, forexample, from the three-phase currents is to rotate ICby 1208 and IB by 2408, and add both to IA. Withdigital techniques, an alternative procedure can beused. A sample of IA is added to a sample of IC that istaken 1208 later, then the sum is added to IB taken 2408later, giving the instantaneous value of 3IA2 thatexisted at the time of the IA sample. Figure 6-10illustrates this process. A similar procedure is used toextract 3IA1 from the individual samples of IA, IB andIC that are taken at 5.55-ms (1208) intervals. Note the

508 shift in the sampling interval at S1 and S4. Thisassures over several cycles that a reasonable distribu-tion of samples is obtained and an accurate measure-ment of negative sequence current is achieved. Thisalgorithm is useful for long-term effects such as motorheating, but is unsatisfactory for fault detection. Theaccuracy of this method is dependent on the nature ofthe harmonics.

5.7 Leading-Phase Identification

An interesting task for a microprocessor is to take acollection of nonsimultaneous samples of voltages andcurrents, determine the related phasors, modify themas the algorithm dictates, and compare the resultingtime-coincident phasors to establish which leads theother

A ¼ ax þ jay

B ¼ bx þ jby

sin g ¼ aybx

jAjjBj �axby

jAjjBj

ð6-6Þ

Equation (6-6) shows the concept that has beendeveloped for digital relays following many years ofexperience with various analog devices with similarfunctions. It states simply and remarkably that phasorA leads phasor B if the product aybx is more positivethan axby.

Only the difference in products, aybx � axby, isrequired to determine which phasor leads the other.No divide function, no sine or tangent calculation, andno table lookup are required, thereby providing veryefficient use of the microprocessor. If g, the angle bywhich phasor A leads phasor B, is between 0 and 1808,the sine of g is positive. With jAjjBj being alwayspositive, sin g is positive if aybx is more positive thanaxby.

5.8 Fault Detectors

Fault detectors provide an additional level of securityin a relaying application. They have been traditionallyof the simple-phase overcurrent variety with thefrequent addition of ground overcurrent. When thefault current/load current ratio is small, compromisesmust be made in the settings. Modern technologyallows more refined ‘‘fault detection.’’

Phase current change, DI, and phase-to-groundvoltage change, DV, can be implemented simply by

Figure 6-10 Sum of samples taken 1208 apart equivalent toIA þ a IC þ a2IB taken at S1.

102 Chapter 6

using a comparison of samples at similar points inadjacent cycles as shown in Figure 6-11. Thesechanges, along with a change in zero sequence current,provide a clear indication of the need for a distanceor directional unit to make a decision, and thereforethey are used to shift some relaying systems (MDAR,REL-301) to ‘‘fault mode.’’

By using DI (phase or I0 current change), a relayingsystem can discriminate between loss of voltage causedby a fault and that caused by potential circuitproblems. DI is present for faults and not loss ofpotential.

6 SELF-TESTING

The ability to monitor much of the hardware is aninherent part of any relaying system equipped with amicroprocessor. Some contain a provision for identify-ing specific types of failure, whereas others indicateonly the general condition of check-failure, requiring amore detailed examination to pinpoint the nature ofthe failure.

6.1 Dead-Man Timer

As part of the housekeeping tasks that are generallyperformed, a dead-man timer (also called watchdogtimer) supervises the fact that the microprocessor iscycling. If the microprocessor fails to perform a givenfunction within a predetermined band of time limits,an alarm output is produced.

6.2 Analog Test

Periodically, a known value of voltage is substitutedfor the normal inputs to the multiplexer. The output ofthe A/D converter is then checked for agreement withthe known input. If there is disagreement, an alarmoutput results. If there is agreement, the multiplexer,sample-and-hold, and A/D converter are proven to bein good working order.

6.3 Check-Sum

Any memory segment that is unchanging such as theROM can be checked through the process of adding upthe contents of the memory and periodically verifyingthat the sum is fixed. Any change in the contents of theROM following power-up constitutes a failure and willproduce an alarm.

6.4 RAM Test

Random access memory is completely checked duringthe initializing process when power is applied to therelay. Word patterns are written and read. Anyinconsistencies are identified.

6.5 Nonvolatile Memory Test

Some relays utilize nonvolatile memory for storingdetails that are pertinent to the operation of the relay,but will be changed by the user from time to time. Anexample of this is the settings. By storing the settings inthree locations when they are first entered andcomparing these three periodically, assurance isobtained that they are correct. Inconsistency producesan alarm.

The ability to test themselves is one of the principaladvantages of microprocessor relays. It relieves theneed to apply external quantities to them periodicallyto verify their capability to perform their intendedfunction. At the same time, it should be recognizedthat no relay is able to completely test itself in allrespects. Backup relays are still required even thoughthose failures that do occur in microprocessor relayshave a high probability of being identified immedi-ately.

Figure 6-11 D1 fault detector.


7 CONCLUSIONS

The introduction of microprocessor technology intoprotective relaying has afforded us the ability toachieve new functions and self-checking provisionsnot previously possible. At the same time, it has causeda reevaluation of long-established practices, resulting

in new approaches to old techniques, as well asencouraging new innovative methods of solvingpersistent protection problems.

The microprocessor has established its place inprotective relaying and will occupy a position ofprominence in future designs.

104 Chapter 6

7

System Grounding and Protective Relaying


1 INTRODUCTION

Ground fault protection is dependent on the power-system grounding, which can vary from solidlygrounded (no intentional impedance from the systemneutrals to ground) to ‘‘ungrounded’’ (systemgrounded only through the capacitance of the system).Ground relaying for effectively grounded systems isdiscussed in Chapter 12. In these systems, the X0/X1

ratio is 3.0 or less, and the R0/X1 ratio is 1.0 or less atall points and under all operating conditions. Witheffective grounding, the line-to-ground fault current isequal to or greater than 0.6 times the three-phase faultcurrent.

Solid grounding is necessary to meet these standardcriteria, particularly with overhead lines where theX0/X1 ratio averages between 1.6 to 3.5. In solidlygrounded systems, the neutrals of the wye-delta powertransformers are directly connected to earth throughthe station ground mat. Considerable design effort isexpended to keep the resistance in this connection to aminimum: Typical values of ground mat resistance toearth are on the order of 0.1O or less in areas of lowground resistivity. Typical values are higher in highground resistivity areas, resulting in a large stationground mat rise (voltage gradient) between the stationarea and remote grounds during ground faults.

Earth, remote ground, and true earth are difficultterms to define precisely, since the earth is a veryheterogeneous mass. The terms represent a mathema-tical fiction needed to identify the zero potential earthplane. In practice, they are considered to exist withinthe earth at any point remote from the influence of the

power system or where current can reasonably flow inthe earth structure.

This chapter will cover protective relaying schemesfor noneffectively grounded systems. These systems fallinto one of three categories:

UngroundedReactance-groundedResistance-grounded

In addition, this chapter will discuss the specialproblems of sensitive ground relaying on distributioncircuits and ground fault protection for bothungrounded and multigrounded three-phase, four-wire systems.

2 UNGROUNDED SYSTEMS

2.1 Ground Faults on Ungrounded Systems

The term ungrounded is strictly one of definition,indicating no physical connection of any kind betweenthe system and ground. Since, however, there is alwaysdistributed capacitance between the three phases of thesystem and ground, the system is grounded throughthis capacitance. On such systems, current flowsbetween each conductor and ground under normalconditions. In the event of a single line-ground fault,the corresponding line-to-ground capacitance isshunted out.

Using symmetrical components, Figure 7-1 showsthe networks and fault representation. Here, X1C, X2C,and X0C are the total distributed capacitances of each

105

system. Although they are shown here as a lumpedquantity, they are actually distributed parameters. X1C

and X2C can be neglected because their effect isinsignificant compared to that of X0C. X0C predomi-nates so that approximately

Ia ¼ 3I0 ¼ 3VG

X0Cð7-1Þ

Since the ground fault current returns through theshunt capacitance, the unfaulted phase currents are not0 (Fig. 7-2). The phase-b and -c voltages are shown asthe prefault line-to-line voltages or

ffiffiffi3

pVLG. This

relation holds true only for the steady-state conditionwith zero fault resistance; transient voltages can beconsiderably higher as shown in Figure 7-3.

When the circuit breaker opens and extinguishes thearc at or near current zero, the voltage is near its peakvalue. This voltage, shown in Figure 7-3 as 1.0 per unit(

ffiffiffi2

ptimes the rms value), remains on the line (or right-

hand) side of the breaker, while the generator voltagegoes to the maximum negative value one half-cyclelater. At that time, the voltage across the breakercontacts is essentially 2.0 per unit, the crest value. Avoltage of this value can cause the arc to restrike acrossthe breaker contacts, sending the line voltage from þ1per unit to �1 per unit.

The result is a high-frequency transient voltage,whose first peak overshoots the �1.0 value by �2.0(the difference between �1 and þ1), giving a peakvoltage value by �3.0. If the arc is again extinguished

at a high-frequency current zero, the trapped charge onthe line produces a voltage of �3.0 per unit. If a secondrestrike occurs at the next voltage positive maximum,the peak voltage will overshoot to þ5 per unit as it

Figure 7-1 Sequence network interconnection for ‘‘a’’

phase-to-ground fault on an ungrounded system.

Figure 7-2 Ground fault on ungrounded system.

Figure 7-3 Overvoltage due to reignitions and restrikes.

106 Chapter 7

goes from �3 to þ1 per unit. Theoretically, furthercycles of extinguishing and restriking of the arc wouldbuild higher and higher voltage values. In practice,however, flash-overs usually occur before these highvalues are reached.

The peak voltage values shown in Figure 7-3 aremaximum theoretical values based on arc extinction atzero current, no damping, and arc restrike at the crestvalue of the source voltage. In fact, circuit resistancewill introduce damping of the transient, reducing thepeak value of the first half-cycle overshoot. Further,restrikes may occur before the voltage reaches crestvalue voltage, which will reduce the value of peakovervoltages. Nevertheless, overvoltages can be veryhigh and represent the major disadvantage ofungrounded systems. An initial fault can cause asecond ground fault to occur on a different phasepossibly on a different feeder, producing a phase-to-phase-to-ground fault with its associated high currentand damage.

2.2 Ground Fault Detection on UngroundedSystems

Since the fault current for a single line-ground fault onan ungrounded system is very small, overcurrent relayscannot be used for fault detection. Voltage relays willdetect the presence of the voltage unbalance producedby the fault, but will not selectively determine itslocation in the system. The unbalanced phase and zerosequence voltages that occur during ground faults areessentially the same throughout the system. Sinceselective isolation of the fault is not possible, relayschemes are only useful for providing an alarm.

Figure 7-4a shows the preferred ground faultdetection system. The voltage transformers must havea primary voltage rating equal to the line-line voltage,since this is the voltage that will be impressed on thetwo unfaulted phases during a line-ground fault.Under normal conditions, the voltage across the relayis approximately 0. When a single line-ground faultoccurs, the voltage becomes 3V0, or approximately200V with 69-V secondary windings. The electro-mechanical type CV-8 relay or solid-state type 59Grelay shown, with their 200-V continuous rating, willdetect fairly high-resistance faults.

Wye-connected transformers with grounded neutralon an ungrounded system may be subject to ferror-esonance during switching or arcing ground faults. Toavoid this, care must be exercised in planning therelationship between the magnetizing impedance of thetransformer, its knee point voltage, and its load.

Users have successfully applied resistors across thebreak in the broken delta configuration (as in Fig. 7-4)having a value that will limit current in the delta loopto the rated current of the voltage transformersecondaries. Much higher values of resistance havealso been used successfully. Karlicek and Taylor intheir important paper, ‘‘Ferroresonance of GroundedPotential Transformers on Ungrounded Power Sys-tems’’ (AIEE Power Apparatus & Systems, August1959, pp. 607–618), concluded that the appropriatevalue of the resistor (called 3R in Fig. 7-4) is 100La/N2, where La is the voltage transformer primaryinductance in millihenries and N the transformer turnsratio.

Typical resistor values in use are as follows:

Voltage transformer ratio Resistor (O)

2400–120 250

4200–120 125

7200–120 90

14,400–120 60

Although the primary fault current may be low,high secondary currents can flow. This should bechecked with the short time or continuous rating of thevoltage transformer and resistor.

Applying a grounded-wye-broken-delta transformerwith only a relay connected across the break and noshunt resistor is equivalent to very-high-impedancegrounding. Any shunt resistor, even as high as 20Xc, isbetter than none. It will damp any high transientvoltage oscillations and probably hold the peak valuesto less than twice the normal crest voltage to ground.

The alternative ground fault detection scheme (Fig.7-4b) is not recommended and should only be appliedafter careful study. The CVD electromechanical relayor the type 27/59 solid-state relay in this system hasseparate contacts for operation on either over- orundervoltage. With the vt connected to phase c, line-to-ground faults on phases a and b produce anovervoltage on the relay; faults on phase c producean undervoltage. For the scheme to work, thecapacitance to ground of the lines must be fairlyclosely balanced and high enough to keep the neutralof the system at close to ground potential.

This scheme can also produce ferroresonance orneutral inversion. When XC/XL is 3.0, VT theoreticallywould be infinite. Even without faults on the system,the high magnetizing impedance of the voltagetransformer can approach resonance with the line

System Grounding and Protective Relaying 107

capacitance to neutral, causing a high overvoltageacross the secondary. Neutral inversion can occurduring a line-ground fault on a phase other than c.Such a fault produces unbalanced impedances toground; the resultant current flows can drive thesystem ground point outside the delta. A loadingresistor across the relay or, less desirably, in series withthe transformer primary may prevent these problems.

3 REACTANCE GROUNDING

There are three different types of reactance grounding:

High-reactance groundingResonant groundingLow-reactance grounding

3.1 High-Reactance Grounding

Until the early 1940s, some utilities operated their unit-connected generators with the neutral ungrounded.Their purpose was to keep the internal line-to-groundfault current in the generator very low and prevent theiron from being damaged by arcing. Unfortunately, theresult was a high insulation failure rate in machinewindings.

These failures were caused by high-voltage transi-ents, similar to those discussed earlier. This problemwas compounded by an inability to detect single line-

ground faults in the generator. As a result, the faultspersisted, causing undue damage.

The initial solution was to connect the generatorneutral to ground through the primary of a voltagetransformer and put an overvoltage relay across thesecondary. In theory, a single line-ground fault wouldsimply cause the generator neutral voltage to shift withrespect to ground, activating the relay and tripping themachine or sounding an alarm. In practice, however,this system actually increased the machine failure rate.The cause was arcing grounds—a phenomenon similarto the restrikes that can occur when switching acapacitive reactance.

The arcing ground phenomenon can be explainedusing Figure 7-5. The equivalent single-line diagramshown in Figure 7-5a is for a generator groundedthrough a high reactance Xn, with a line-to-groundfault near one terminal. Xc is the distributed capacitivereactance of the windings to ground, connected half-way between the generator reactance Xg. If the arc isextinguished when the small fault current passesthrough 0, the voltage across the arc path must gofrom nearly 0 to the normal crest value. In doing so, itmust oscillate around the steady-state normal value.

As shown in Figure 7-5b, the resultant voltagetransient will reach a peak value of twice the normalcrest line-to-neutral voltage, one-half cycle of the high-frequency transient after the arc is extinguished. If thearc restrikes at this point, the fault voltage is drivenback to 0. When the arc is initially extinguished, thereactor voltage has to go from the positive maximum

Figure 7-4 Ground fault detection on ungrounded systems.

108 Chapter 7

to 0. As a result, it has a transient oscillating periodfrom the positive maximum to negative maximum.

The first half-cycle of this oscillation is shown inFigure 7-5c. If the arc restrikes at the instant when thefault voltage is twice the normal crest value, as wasassumed in Figure 7-5b, the reactor voltage has to gofrom the negative minimum to positive maximum. Theresult is another transient oscillation, with a peak valueof three times the normal maximum line-groundvoltage.

Note that in high-reactance grounding, the reactorvoltage is applied between the generator neutral andground. Since the BIL of the reactor is higher than thatof the generator windings, insulation failures are morelikely to occur in the generator windings.

The switching surges that result from clearing line-to-ground faults for ungrounded systems also occur inhigh-reactance grounded systems. In the latter case, theresulting transient overvoltages will be even higher.The source voltage for an ungrounded system is thenormal line-to-neutral voltage that, theoretically,produces successive line-side voltage peaks of 1.0,3.0, 5.0, . . . of normal crest voltage to neutral. For thehigh-reactance grounded system shown in Figure 7-5,with the reactor between the neutral and ground, thesource voltage is the normal line-line voltage. The

corresponding theoretical transient peaks areffiffiffi3

p; 3

ffiffiffi3

p;

5ffiffiffi3

p, and so on. For these reasons, high-reactance

grounding was discontinued many years ago.

3.2 Resonant Grounding (Ground FaultNeutralizer)

In certain sections of the United States, resonantgrounding has been applied successfully in unit-connected generator grounding applications. It is notapplied in transmission line applications in the UnitedStates, but other countries use it. In this scheme, thetotal system capacitance to ground is compensated foror cancelled by an inductance in the grounded neutralof the power transformers. The grounding reactor,equipped with taps that permit it to be tuned to systemcapacitance, was first called a Petersen coil. It is nowmore commonly designated a ground fault neutralizer.

Theoretically, if the reactor perfectly matches thesystem capacitance, a line-to-ground fault will producezero current, the transient fault arc will be extin-guished, and the arc path deionized, without the needfor deenergizing the circuit.

In this system, approximately 75% of line-groundfaults are self-extinguishing. The remaining faults mustbe cleared by a line breaker.

In theory, resonant grounding should reduce lineoutages considerably. This system does, however, havea number of disadvantages:

Transformers connected to the system must havefull line-line insulation even when wye-con-nected.

The entire system must be fully insulated for line-line voltage.

The ground-fault neutralizer must be retuned toaccommodate any changes in system configura-tion: additions, extensions, line removals, orswitching.

System effectiveness will be reduced considerably ifa substantial number of lines are of wood poleconstruction. The high insulation to ground willresult in a larger portion of line-line faults(conductor swing caused by wind).

A high incidence of faults will occur essentiallysimultaneously in different parts of the system.

3.3 Low-Reactance Grounding

Low-reactance grounding used to be applied tosystems fed at generator voltage. The generator neutral

Figure 7-5 Overvoltages on reactance grounded system due

to arcing fault.


was grounded through a reactor. The reactor was sizedto keep the magnitude of a single line-to-ground faulton the machine terminals equal to a three-phase fault.[A reactor value of (2X1�X0�X2)/3 was used.]

In general, low-reactance grounding was applied tolarge industrial plant systems with radial distributionfeeders, and ground fault protection consisted simplyof overcurrent relays. Gradually, this type of generatorgrounding has been replaced by low-resistance ground-ing.

Another type of low-reactance grounding providesground fault current relaying for systems suppliedfrom a delta source. The reactance grounding schemeshould

Supply sufficient ground fault current to operaterelays for a fault when the line value (X0þ 2X1) isthe highest.

Limit the transient overvoltages attributable toground faults to a value of 2.5 times normalline-to-neutral crest value, with two restrikes.

In this scheme, either a grounded wye-delta or zig-zagtransformer can be used, although the zig-zag (Fig.7-6) is more common because of its economy. Thewindings shown parallel are on the same core leg. Withthis connection, the positive sequence impedance of thebank is very high and equal to the magnetizingimpedance. When zero sequence current passesthrough the bank as shown, the impedance is equalto the leakage reactance.

The rating of the transformer is chosen so that themaximum X0/X1 value is 4. When X0/X1 equals 4, theline-ground fault current is half the three-phase, short-

circuit value, if we assume that X2¼X1. Thus, if aground relay is used in the common neutral connectionof the line current transformers, the current level is amaximum of half that of the phase current for a three-phase fault.

4 RESISTANCE GROUNDING

Resistance grounding is applied in systems with three-wire distribution at the generator voltage and for unit-connected generators. The two general types ofresistance grounding are low- and high-resistancegrounding.

4.1 Low-Resistance Grounding

Whenever low impedance grounding is desired, resist-ance grounding is generally preferred to the low-reactance systems described above. Specifically, low-resistance grounding is used for systems fed directly atthe generator voltage (Fig. 7-7a) or through a delta-wye transformer (Fig. 7-7b). When a line-to-groundfault occurs in the system, the current flowing in theground resistor results in a sudden change in generator

Figure 7-6 Reactance grounding.

Figure 7-7 Low-resistance grounding of systems fed

through Delta-Wye transformer.

110 Chapter 7

load, causing severe generator angular swings andhigh-peak shaft torques. To keep the ground resistorloss low, the resistor is generally sized to limit thesingle line-to-ground bus fault to around 100 to 400A.Ground relaying using residual overcurrent relays maybe applied, but zero-sequence-type current transfor-mers will provide greater sensitivity.

For distribution systems, transient overvoltagesmay be limited to less than 2.5 times the normal crestvalue to ground, if

R0

X0� 2:0 and

X0

X1� 20 ð7-2Þ

For a neutral resistor R,

R0 ¼ 3R

Assuming that X0 ¼ 20X1 and R0 ¼ 2X0, then we have

Z0 ¼ ð40þ j 20ÞX1

For a line-to-ground fault,

Ig ¼ 3I0 ¼ 3

X1 þX2 þ Z0

¼ 3:0

ðj 1þ j 1þ 40þ j 20ÞX1

¼ 3:0

ð40þ j 22ÞX1¼ 3:0

45:65 X1ff � 28:8 �

¼ 0:066

X1ff � 28:0 � per unit

ð7-3Þ

The three-phase fault current would be

I3f ¼ 1:0

X1per unit ð7-4Þ

Thus, if we use Eq. (7-4), the line-to-ground faultcurrent magnitude is

Ig ¼ 0:066 I3f ð7-5Þ

The resistor can be in the neutral of the transformer(Fig. 7-7), or a resistor can be inserted in the neutral ofthe grounded zig-zag transformer (Fig. 7-6). In eithercase, the reactance component of the resistor must beconsidered. Cast-iron grid-type grounding resistorshave a power factor of approximately 0.98, stainlesssteel types one of approximately 0.92. The reactance,while small in itself, is tripled in the zero sequencecircuit.

4.2 High-Resistance Grounding

High-resistance grounding is applied to a generator-transformer unit system by connecting a resistor acrossthe secondary of a distribution transformer in thegrounded generator neutral (Fig. 7-8). The resistorvalue is selected so that its KW loss for a solid line-to-ground fault at the machine terminal is equal to orgreater than the charging kVA of the low-voltagesystem. To do this, 3R is chosen to be equal to or lessthan Xco. Xco is the combined zero sequencecapacitive of the generator windings, cable connectionsto the transformer, low-voltage transformer winding,and station service transformer plus any surge protec-tive capacitors that are applied at the generatorterminals and connected phase to ground. This resistorvalue will limit generator iron burning from groundfaults, damp out oscillations, and limit the peaktransient voltage to 2.5 times normal line-to-neutralvoltage, or less. The rating required for the resistor is

Resistor KWR ¼ V2

1000Rð7-6Þ

where V is generator-rated phase-to-neutral voltage (involts).

The transformer kVA requirement is the same asthis resistor KW value, of course, but the kVA rating ishigher because of the choice of a higher-voltage ratingfor the transformer to avoid possible ferroresonance.

Figure 7-8 High-resistance grounding of the unit connected

generator-transformer.


The magnitude of primary fault currents in theseapplications is around 8 to 10A.

Sensitive protection is provided by an overvoltagerelay across the resistor. This application is detailedunder generator protection (Chap. 8).

5 SENSITIVE GROUND RELAYING

Ground relaying on distribution circuits can bedifficult. The range of fault currents can vary fromnegligible, for a conductor lying on or near the groundwith minimum electrical contact, to substantial, for aconductor making good contact with ground. Unfor-tunately, there is no practical way of distinguishing anintolerable situation from a tolerable one at a breakeror disconnection location.

Some years ago, a utility conducted tests on a 10-ftlength of no. 4 bare copper wire energized at 12 kV andlaid on a variety of surfaces such as dry grass, greenvegetation, dry base soil, and asphalt. Of 128 tests, 7%showed currents of less than 7A, 7% over 1000A, and55% had currents in the range of 150 to 600A.

Ground fault protection is dictated by the amountof ground fault currents available from the system tooperate relays and the ratio of this current to normalsystem residual unbalance. Load management mayhelp to reduce normal unbalance in some cases. Theminimum ground fault current must balance servicecontinuity with equipment protection. That is, it mustbe low enough to minimize equipment damage, buthigh enough to be recognizable and allow the faultedarea to be selectively isolated without nuisancetripping.

The design of the system grounding should becompatible with the sensitivity of the relaying that is tobe used. Three commonly used ground relay schemes,in order of increasing sensitivity, are

Ground relay in the common neutral connection ofthe line current transformers and/or groundedsource (Fig. 7-9).

Ground relay in the common neutral connection ofthe line current transformers, with a product-type relay to avoid operation on false residualcurrents (Fig. 7-10). The CWP scheme (Fig.7-10b) provides increased sensitivity, whereas theCWC scheme will not (see also Fig. 7-11).

Ground relay with a zero sequence (ring) type ofcurrent transformer (Fig. 7-12).

5.1 Ground Overcurrent Relay withConventional Current Transformers

In the scheme shown in Figure 7-9, the relays areusually set on the 0.5-A tap. Because of the largeburdens of electromechanical ground relays on theminimum tap, the relay pickup current multiplied bythe current transformer ratio will not be the primaryampere pickup when using lower-quality currenttransformers (see Chap. 5). To hold the excitingcurrent to a reasonable minimum, it may be necessaryto use a higher tap setting than would otherwise bedesired, a solid-state or numerical relay, or a bettercurrent transformer.

The unequal performance of current transformersduring heavy phase faults or initial asymmetricalmotor starting currents may produce false residualcurrents with the scheme shown in Figure 7-9a. Whenthese currents cause relay operation, an instantaneousrelay with a higher pickup should be substituted, or thetime overcurrent relay should have a larger time dialand/or pickup setting. Increasing the burden on thecurrent transformers in these cases causes them tosaturate more uniformly, reducing the false residualcurrent.

Higher burdens, however, may also decrease therelay sensitivity on light ground faults, depending onthe quality of the current transformers. False residualcurrents do not occur in the scheme shown in Figure7-9b or 7-12 and do not cause relay operations inFigure 7-10a or b.

With the application of a ground relay set on the0.5-A tap, the fault current in the relay should not beless than twice pickup, or 1.0A secondary.

To be able to detect a ground fault that produces acurrent having a value of 10% of the current producedby a bolted phase-to-ground fault is a reasonablecriterion. The maximum load on any circuit off a bus

Figure 7-9 Ground protection with conventional current

transformers and protective relays.

112 Chapter 7

dictates the critical ct ratio. With a primary current of0.10 IG, a ratio of K:5 and minimum secondary currentof 1A produce the following requirement:

IG � 2K ð7-7ÞThat is, the current permitted by the grounding devicefor a bolted phase-to-ground fault should equal orexceed twice the primary rating of the largest ct usedon a circuit off of the bus. If the use of this criterionproduces an excessively high ground fault current, alower value of current can be chosen by using a higherresistance value. This will then require a more sensitive

relaying scheme to be used, such as a zero sequence(ring) type of current transformer with its low ratio(typically 50:5) and an instantaneous overcurrentrelay, such as an IT or 50D.

5.2 Ground Product Relay with ConventionalCurrent Transformers

Better security is possible using the schemes ofFigure 7-10, and increased sensitivity may be provided

Figure 7-10 Sensitivity ground protection utilizing product-type relays.

Figure 7-11 Phasors for Figure 7-10b for a phase ‘‘a’’-to-

ground fault on a high-resistance grounded system.

Figure 7-12 Ground protection utilizing the zero sequence

(ring) type current transformers.


by the scheme of Figure 7-10b. These schemes will notoperate on false residual currents; the relays requirecurrent in both windings to operate. No system groundor zero sequence current or voltage will exist for phasefaults or motor starting currents.

Two product-type devices are used in the scheme ofFigure 7-10b: the CWP and CWP-1. The CWP isapplicable for reactance grounded systems and theCWP-1 specifically tailored to high-resistancegrounded systems with its 45 8 lead characteristic.

The greatest ground fault sensitivity is provided bythe CWP-1 (32N) relay. The relay pickup is adjustablebetween 5 and 40mA with 100V across the potentialcoil at maximum torque.

The phasors for a high-resistance grounded systemwhere the CWP-1 relay is applicable are shown inFigure 7-11.

5.3 Ground Overcurrent Relay with ZeroSequence Current Transformers

The scheme shown in Figure 7-12 provides maximumsensitivity. There are no false residual currents. Thezero sequence type of current transformer has theconductors passed through the center hole, and the ctratio is not dictated by the load current. Secondarycurrent is the transformed system zero sequencecurrent 3I0.

The standard ratio for the zero sequence type oftransformer (type BYZ) is 50/5; 100/5 ratios wereoriginally used. Various nondirectional relays can beapplied, as outlined in Table 7-1.

The IT relay has a large burden when the 0.15-A tapis used (19.6O). Saturation of the 50/5 ct will occur atroughly 5A primary. Field tests indicate that thesecondary pulse width at 1800A primary is onlyapproximately 30 8 following each zero crossing, butthis relay operates satisfactorily with this extreme

degree of saturation. This relay and ct combination isintended to provide sensitive detection of ground faultsand is not expected to perform adequately in thepresence of fault currents beyond the moderate (up to1800A rms symmetrical primary) range. Where largerfault currents are expected, the system of Figure 7-9aand a larger ct ratio should be used.

When the maximum fault current exceeds themaximum values shown, the output waveform isnonsinusoidal. Relay timing will tend to becomevariable and longer than indicated in the publishedliterature.

The above schemes are for feeder circuit protection.For the ground protection of equipment, a grounddifferential scheme can be used with a differential-typerelay or product-type (CWC) relay as shown inFigure 7-13. This is also applicable to short-run feederswith three conventional ct’s or a zero sequence type ctat each end of the protected zone. The CWC relay isrecommended as it provides high sensitivity and isrelatively independent of the current transformerperformance.

6 GROUND FAULT PROTECTION FOR THREE-PHASE, FOUR-WIRE SYSTEMS

6.1 Unigrounded Four-Wire Systems

Unigrounded, four-wire systems have insulated neu-trals; the only ground connection is at the substation.Loads generally are connected phase to neutral, andthe net load unbalance returns through the neutral as aresidual current. For faults from phase to ground, thecurrent returns through the earth to the substationneutral.

There are three different relay schemes for groundfault protection for unigrounded systems, as shown inFigure 7-14. Figure 7-14a illustrates the conventional

Table 7-1 Relay Settings and Sensitivities Using the 50/5 BYZ Zero Sequence Current Transformers

Minimum sensitivity in primary 3I0amperes

Maximum primary 3I0 amperes for

accurate timing and coordination

Relay type Relay setting (43=4 ID) (73=4 ID) (43=4 ID) (73=4 ID)

IT 0.15 5.0 5.0 — —

CO-8 or 9 0.5 9.0 10.0 25 112

CO-8 or 9 2.5 24.0 24.0 540 1215

CO-11 0.5 6.0 7.0 70 150

CO-11 2.5 24.0 24.0 700 900

114 Chapter 7

scheme used on three-phase, three-wire systems. For afour-wire system, the load unbalance current wouldflow through the ground relay, requiring a setting toavoid operation on the maximum load unbalance. Thisscheme is generally not recommended for the uni-grounded system. The four-current transformerscheme shown in Figure 7-14b provides much highersensitivity, since it does not measure the loadunbalance residual current. Even greater sensitivity isprovided by the zero sequence type of currenttransformer depicted in Figure 7-14c. Comparativesensitivities for various relays in this scheme are listedin Table 7-1.

If a line-to-neutral fault occurs on the system, onlythe conventional scheme (Fig. 7-14a) will respond. Theconnection of the current transformers in the other twoschemes results in cancellation of the fault current,unless it involves ground. The phase relays will provideprotection, however, since phase-to-neutral fault cur-rent in one phase will be of the same order ofmagnitude as a three-phase fault.

A fault between neutral and ground is possible eventhough the neutral is nearly at ground potential,probably as the result of a broken neutral conductor.The schemes of Figures 7-14b and c will measure anycurrent returning through the earth. Because the

ground return may be a high impedance path, causinglow voltage at the load points, the more sensitivewindow-type current transformer scheme is recom-mended.

6.2 Multigrounded Four-Wire Systems

Many three-phase, four-wire distribution systems aresolidly grounded at the substation, with the neutralwire also grounded at each distribution transformerlocation. Such systems are difficult to protect againstground faults. The scheme of Figure 7-9 is used mostoften with the sacrifice of sensitivity dictated bymaximum load unbalance.

Figure 7-13 Ground differential for wye winding using

CWC (product-type relay).

Figure 7-14 Methods of ground protection on unigrounded

systems.


8

Generator Protection

Revised by: C. L. DOWNS

1 INTRODUCTION

The frequency of failure in rotating machines is lowwith modern design practices and improved materials,yet failures will occur and delayed tripping orinsensitivity of protection may result in severe damageand long outages for repairs. For these reasons,abnormal conditions must be recognized promptlyand quickly isolated to avoid extending the damage orcompounding the problem.

Abnormal conditions that may occur with rotatingequipment include the following:

Faults in the windingsOverloadOverheating of windings or bearingsOverspeedLoss of excitationMotoringInadvertent energizationSingle-phase or unbalanced current operationOut of stepSubsynchronous oscillations

Some of these conditions do not require that the unitbe tripped automatically, since in a properly attendedstation, they can be corrected while the machineremains in service. These conditions are signaled byalarms. However, most require prompt removal of themachine from service.

For any particular hazard, the initial, operating,and maintenance costs of protective schemes and thedegree of protection they afford must be carefullyweighed against the risk encountered if no protection

were applied. The amount of protection that should beapplied will, of course, vary according to the size andimportance of the machine.

2 CHOICE OF TECHNOLOGY

In the choice of relays to be applied for the variousfunctions described here, it will be recognized that theyare available as discrete functions in their individualhousing or as a complete complement containing allthe pertinent protection plus data acquisition.

Electromechanical, solid-state, and microprocessor-based devices are used depending on personal choiceand whether or not a new installation, an upgrade, or afunctional addition is involved.

Although electromechanical and single-functionsolid-state relays have proven their reliability, flex-ibility, and effectiveness, the trend is toward micro-processor-based integrated packages. Many of theseprovide event recording, oscillography, self-monitor-ing, communications, adaptive characteristics, andother features that only a microprocessor-orientedsystem can provide.

3 PHASE FAULT DETECTION

Internal faults in equipment generally start as a groundin one of the stator windings and may occasionallydevelop into a fault involving more than one phase.Differential protection is the most effective schemeagainst multiple-phase faults. In differential protec-

117

tion, the currents in each phase, on each side of themachine, are compared in a differential circuit. Any‘‘difference’’ current is used to operate a relay.

Figure 8-1 shows the relay circuits for one phaseonly. For normal operation or a fault outside the twosets of current transformers, Ip entering the machineequals Ip leaving the machine in all phases, neglectingthe small internal leakage current. The secondarycurrent of each of the ct’s is the perfectly transformedprimary current minus the magnetizing current Ie.

The relay current Ie1 � Ie2 is the difference in theexciting or magnetizing currents. With the same type ofcurrent transformers, this current will be small atnormal load. If a fault occurs between the two sets ofcurrent transformers, one or more of the left-handcurrents will suddenly increase, whereas currents onthe right side will either decrease or increase and flowin the opposite direction. Either way, the total faultcurrent will now flow through the relay, causing it tooperate.

If perfect current transformers were available, anovercurrent relay in the ‘‘difference’’ circuit could beset to respond very sensitively and quickly. Inpractice, however, no two current transformers willgive exactly the same secondary current for the sameprimary current. Discrepancies can be traced tomanufacturing variations and differences in secondaryloading caused by unequal length of relay leads andunequal burdens of meters or instruments connectedin one or both secondaries. The differential currentproduced flows through the relay. Although normallysmall, the differential current can become appreciablewhen short-circuit current flows to an external fault.An overcurrent relay would have to be set above themaximum error current that could be expected during

an external fault. On a symmetrical basis (no dc offsetin the primary current), this would not exceed 10A ifa C class ct were used within rated burden and theratio were chosen such that the secondary current didnot exceed 100A for the maximum ‘‘through’’ phasefault. To avoid operation on an asymmetrical fault,the trip time would have to exceed three dc timeconstants.

3.1 Percentage Differential Relays (Device 87)

The percentage differential relay (Fig. 8-2) solves theproblems of poor sensitivity and slow operation. Theinputs from the two sets of current transformers areused to generate a restraint quantity. This is thencompared to the difference of these two currents.Operation (or restraint) is produced as a result ofthe comparison of the difference to the restraint.This desensitizes the relay for high external faultcurrents.

The current required for relay operation increaseswith the magnitude of the through fault current. Thepercentage of increase may be constant, as in the CA(87) generator percentage differential relay. Alterna-tively, the percentage of increase may vary with theexternal fault current, as in the high-speed SA-1 (87)generator relay. The effect of the restraint on internalfaults is negligible, because the operating quantity isweighted and responds to the total secondary faultcurrent.

Generator differential relays are available withvarious percentage differential characteristics. Theyare typically 10%, 25%, and variable-percentagedifferential types. The percentage indicates the differ-

Figure 8-1 The basic differential connection.

Figure 8-2 Schematic connections of the percentage differ-

ential type relay. (Only one-phase connections are shown.)

118 Chapter 8

ence current as a percentage of the smallest restraintcurrent required to operate the relay. The pickup (thevalue of current into one restraint winding and out theoperating winding) is the current required to barelymake the relay operate. Its value tends to be smallerfor the lower-percentage differential relays and is aslow as 0.10A for some. Operating time, in general, issmaller for solid-state relays, being 25msec for thesolid-state SA-1 compared to 80 to 165msec for theelectromechanical CA relay.

Multifunction microprocessor relays do not have aphysical operating winding, the difference of therestraint currents being computed mathematically bythe protection algorithm.

In all differential schemes, it is good practice to usecurrent transformers with the same characteristicswhenever possible and avoid connecting any otherequipment in these circuits.

3.2 High Impedance Differential Relays(Device 87)

High impedance differential relaying is based on theconservative premise that the ct’s on one side of thegenerator perform perfectly for an external fault andthe other set of ct’s saturate completely. It takesadvantage of the fact that the voltage appearing acrossthe relay is limited for an external fault to the voltagedrop produced by the maximum secondary currentflowing through the leads from the relay to thesaturated ct and through its internal resistance. Foran internal fault, the voltage will approach the open-circuited ct voltage (usually limited by a varistorinternal to the relay). In general, this scheme is not assensitive as the percentage differential scheme but ismore secure.

3.3 Machine Connections

Most generators have wye-connected windings. Asshown in Figure 8-2, three relays connected to wye-connected current transformers provide phase and, insome cases, (depending on the type of neutral andsystem grounding) ground fault protection. Figure 8-3illustrates a similar protective scheme for deltagenerators. In this scheme, the delta windings mustbe brought out so current transformers can be installedinside the delta.

3.4 Split-Phase

Generators with split-phase windings can be protectedby two sets of differential relays: one connected as inFigure 8-2 and the other as in Figure 8-4. Thisarrangement protects against all types of internalphase faults, including short-circuited turns or open-circuited windings. This scheme may be extended toaccommodate other winding arrangements involvingmore than two equal windings per phase. Unless theratios of the current transformers produce an exactmatch, the scheme of Figure 8-4 must be equipped withauxiliary transformers to provide a balance duringnormal operation.

Figure 8-3 Percentage differential relay schematic for a

delta-connected machine. (Only one-phase connections are

shown.)

Figure 8-4 Schematic connections for one-phase only for

the protection of a machine with split phase windings.

Generator Protection 119

4 STATOR GROUND FAULT PROTECTION

The method of grounding affects the degree ofprotection afforded by differential relays. The higherthe grounding impedance, the less the fault currentmagnitude and the more difficult it is to detect highimpedance faults. With high impedance grounding, thedifferential relays will not respond to single-phase-to-ground faults. A separate relay in the grounded neutralwill provide sensitive protection, since it can be setwithout regard to load current.

The ground relay may also operate for groundfaults beyond the generator. For this reason, a timedelay may be necessary to coordinate with anyoverlapped relays. A typical case is a generatorconnected directly to a bus with other circuits. A faulton one of these circuits should not trip the machine;the relays in the faulted circuit will clear such faults. Awye-delta transformer bank will block the flow ofground current, preventing faults on the opposite sideof the banks from operating ground relays. In the unit-connected scheme, the transformer bank limits theground relay operation to faults in the generator, theleads up to the transformer bank, and the deltawinding.

4.1 Unit-Connected Schemes

The unit-connected system is the most commonarrangement for all but small generators. For unitsystems high-resistance grounding is used, and themachine is generally grounded through a distributiontransformer and resistor combination, as shown inFigure 8-5. Since the secondary is rated at 120 or240V, the physical size of the resistor can be somewhatsmaller than if it were connected to the primary.

4.1.1 ‘‘95% Scheme’’

The unit system responds to the voltage shift of thegenerator neutral with respect to ground that occursfor a ground fault in the machine, bus, or low-voltagewinding of the transformer. The relay used must beinsensitive to the substantial normal third harmonicvoltage that may be present between neutral andground, and yet sensitive to the fundamental frequencyvoltage that accompanies a fault.

Since the magnitude of the neutral shift is dependenton the location in the winding of the ground fault(neutral-to-ground fault produces no neutral shift) andthe usual choice of relay sensitivity and distributiontransformer voltage ratio provides roughly 95% cover-

age of the winding, this relay is often referred to as a95% relay.

4.1.2 Neutral Third Harmonic Undervoltage

Other schemes take advantage of the presence of thethird harmonic voltage between neutral and groundand respond to undervoltage for a neutral-to-groundfault.

4.1.3 100% Winding Protection

Other ground relaying schemes provide completeprotection of the generator stator by injecting a signalinto the stator and monitoring it for change. Thisconcept allows 100% coverage even though themachine is at standstill, whereas the 95% and neutralthird harmonic schemes depend on the machineoperating at rated speed and voltage.

4.2 95% Ground Relays

The CV-8 or solid-state 59G low-pickup overvoltagerelay can be used for unit-generator applications asshown in Figure 8-5. Provided that a full-rated primarywinding is used, the maximum voltage for a solidground fault is 120=

ffiffiffi3

p(69.3V with a 120-V distribu-

tion transformer secondary), or 240=ffiffiffi3

p(138.6 V with

a 240-V secondary).The scheme has good sensitivity for internal ground

faults while being very insensitive to third harmonicvoltages. Various provisions are used to make the relayinsensitive to the third harmonic. The third harmonicpickup of the CV-8 relay, for example, is approxi-mately 8 times the pickup at rated frequency.

The voltage appearing from the neutral of thegenerator to ground is dependent on the location of theground fault. The more sensitive the ground relay, thegreater the percentage of the winding protected.Obviously, a neutral-to-ground fault goes undetectedby this relay and other devices must be considered. Insome relays, the sensitivity is proportional to its rating,and the best protection is obtained by using the relayhaving the lowest voltage rating. For a ground fault atthe line terminal of the machine, full line-to-neutralvoltage will exist from neutral to ground. The voltageon the relay is, of course, dependent on the ratiochosen for the distribution transformer. If this voltageexceeds the rating of the relay and is not removed bytripping the field circuit within the short-time cap-ability of the relay, the SV scheme shown in Figure 8-5may be used to protect the relay for those fault cases

120 Chapter 8

producing this high voltage. The SV relay is set to openits contacts at a value somewhat lower than thecontinuous rating of the protected relay, inserting R tolimit the voltage. In general, the 95% type relay isallowed to trip immediately to remove voltage prior tothe occurrence of damage to the relay, making the SVunnecessary. Other solid-state relays such as the type59G have a 208-V continuous rating and 60-Hzsensitivity as low as 1V.

Time-delay settings of 25msec to 4 sec are used forthis function. These longer delays allow for coordina-tion with voltage transformer fuses, if required.

Operation of the ground relay can be avoided forfaults on the main voltage transformer secondary bygrounding one phase of the secondary rather than theneutral. Then a ground fault on the voltage transfor-mer secondary will not produce a machine neutralvoltage shift and the ground relay will not operate.This is the recommended grounding practice forgenerator voltage transformers.

Another scheme that is often used in industrialapplications uses a product-type relay for grounddetection. This arrangement is described in Chapter 7,Figure 7-13. Good sensitivity is achievable because ofthe ability to use a low-ratio ct in the neutralgrounding resistor path that is not related to themachine full-load current.

4.3 Neutral-to-Ground Fault Detection (Device87N3)

Figure 8-6 describes a scheme for detecting a neutral-to-ground fault on the generator. This fault is, in itself,not hazardous. A second ground fault at the machineterminal, however, causes a line-to-ground fault that isnot limited by any neutral impedance. This faultcurrent magnitude will quite likely exceed the currentmagnitude for which the machine is designed. Machine

Figure 8-5 Schematic connections for ground fault protection of a unit type machine resistance grounded through a

distribution transformer.


destruction may result. Early detection, then, isimperative.

The scheme of Figure 8-6 compares the thirdharmonic voltage present between the machine neutraland ground with that at the line terminals. The relativevalues of these voltages are established by thedistributed capacitances of the generator, phase leads,and transformer low-voltage winding plus the ground-ing system. Though the third harmonic voltagechanges with machine loading, the ratio of these valuesremains relatively constant.

When a ground fault occurs, this relationshipchanges and therefore allows detection of faults at allpoints not covered by the ‘‘neutral shift’’ relay. Thesetwo relays, 59N and 59D, provide ‘‘100% ground faultcoverage.’’ Very high impedance faults in the protectedzone, of course, cannot be detected. Also for the unit-connected system neither relay will provide any backupfor faults on the transmission system.

4.4 100% Winding Protection

Another important ground fault detection scheme(GIX-104) utilizes an injected current that is monitoredfor magnitude. In the configuration shown in Figure8-7, the small injection voltage is applied across thelower part of the grounding resistor. As a result of thedistributed capacitance of the generator and connectedapparatus, a current flows and produces a voltage dropacross the measuring transformer. A ground faultoccurring anywhere in the protected winding will causethe current to rise and the voltage to increase acrossthe grounding resistor. To differentiate from otherconditions that could produce a similar voltage, theinjection voltage is commutated, producing a codedcharacter that is readily identifiable.

The GIX-104 may also be equipped to detect,through the injection principle, a ground fault in thegenerator field.

Figure 8-6 180-Hz voltage comparator.

122 Chapter 8

This scheme has the important advantage that it isable to detect a stator or field circuit groundirrespective of the status of the generator. It may bedead, spinning without field, near synchronizing, on-line, loaded, or unloaded.

5 BACKUP PROTECTION

5.1 Unbalanced Faults

Unsymmetrical faults produce more severe heating inmachines than symmetrical faults. The negativesequence currents that flow during these unbalancedfaults induce currents in the rotor having twice systemfrequency. These currents tend to flow in the surface ofthe solid rotor forging and the nonmagnetic rotorwedges and retaining rings. The resulting I22R lossquickly raises the temperature. If the fault persists, themetal will melt, damaging the rotor structure.

Such faults result from failure of a protectivescheme or equipment external to the machine. Therelative magnitudes of negative sequence currents forline-to-line faults on a typical turbine generator underdifferent operating conditions are shown in Figure 8-8.The effect of the higher excitation during the fault isincluded for short circuits with load on the system.

According to ANSI standards, the permissibleintegrated product I22t, (where I2 is negative sequencecurrent in per unit of machine rated current and t isseconds duration) that ‘‘indirectly cooled’’ turbinegenerators, synchronous condensers, and frequency-changer sets can tolerate is 30. The standard forhydraulic turbines or engine-driven generators is 40.Standard ‘‘directly cooled’’ machines up to 800MVAare capable of withstanding a permissible integratedproduct of 10, whereas some very large machines(1600MVA) can only tolerate 5. Early inspection todetect damage is recommended for machines subject tofaults between the above limits and 200% of the limit.Serious damage can be expected for faults above 200%.

With so many influences on the I22t as shown inFigure 8-8, it is evident that a relay designed torespond in a way similar to the manner in which heat isgenerated in the machine is mandatory. Many varia-tions are available in generator negative sequenceovercurrent relays (see Fig. 8-9), but each is tailored tomatch the I22t ¼ K characteristic for K between 10 and40, with some spanning an even greater range. TheANSI standard explicitly restricts this limit to I2 valuesabove the full-load level. Earlier relays covered this,but provided no I2 protection below the full-loadvalue. Later versions of the ANSI standard introducedan additional limit, unrelated to time. More recentlydesigned negative sequence overcurrent relays providean alarm level and trip capability for currents betweena pickup setting and full load. This is pertinent toincreased heating effects caused by such factors asunbalanced load or faulty circuit breakers.

The negative sequence protective function is recom-mended for all machines rated 1000 kVA or larger,though it can be justified for important smallermachines. Examples of schematic connections forunbalanced fault protection are shown in Figure 8-10.

The filter output that is applied across the operatingcoil of the COQ relay is

VF a 2K2Ia2 ð8-1Þwhen the connection shown in Figure 8-10a is used. If,however, the auxiliary current transformer is not used,

VF aK2ð2Ia2 þ Ia0Þ ð8-2Þwhere K2 is the filter constant.

If Ia0 is small, its effect can be ignored. Otherwise, itwill be necessary to use either the auxiliary currenttransformer to remove it or the relay with neutralmade up inside, where its effect can be nullified. Theauxiliary current transformer is not normally required

Figure 8-7 Neutral voltage injection.


in unit-connected applications. For relays such as theSOQ and 46Q that use the equivalent of delta currents,any zero sequence current is ignored by the relay.

When a continuous load unbalance in excess of the5 or 10% of the capability of the particular machinemay occur, the SOQ relay, or an additional relay 46Q,set for the desired alarm level may be used to alert anoperator. With the SOQ, instrumentation can identifythe level of negative sequence current to permit adecision between tripping or decreasing the machineloading.

An excellent backup for ground fault detectingrelays for a unit-connected generator utilizes a relaysupplied by a current transformer (rated 100:5 orthereabouts) in the secondary leads of the neutraldistribution transformer. This is used by some utilitiesas the primary ground relaying.

Integrated protection relays such as the REG-100 orREG-216 include the negative sequence protective andalarm functions.

5.2 Balanced Faults

5.2.1 Distance Relay (21)

A generator should be protected also against damagethat will result from prolonged contribution to abalanced fault. A distance relay such as a KD-11, fedfrom current transformers in the neutral of thegenerator and a voltage supply connected at thegenerator voltage level, provides such protection. Asingle relay of this type complements the COQ, 46Q, orSOQ in recognizing balanced faults internal andexternal to the generator. It also supplements these

Figure 8-8 Relative magnitudes or negative sequence currents for line-to-line faults on a typical machine under different

operating conditions. (From AIEE Transactions, Volume 72, 1953, Part iii, Page 283, Figure 1.)

124 Chapter 8

relays by sensitively recognizing unbalanced faults.The connection described above makes the relaydirectional from the neutral, but gives it a reach inboth directions from the voltage transformer location.As a result, it will sense some generator as well astransformer faults.

The distance relay is usually set to reach through theunit transformer. Unlike single-phase distance relays,the reach of the KD-type relays is not affected by thephase shift through the bank. When set for animpedance greater than the transformer impedance,the KD-11 relay will operate for both generator andline-side phase faults that manifest an impedancewithin this reach. A timer must be used to ensureonly the minimum equipment outage necessary to cleara fault. The timer must be set to coordinate with thehigh-voltage transmission line relays and all otherrelays it overreaches.

5.2.2 Voltage-Controlled OvercurrentRelay (51V)

If a negative sequence overcurrent relay is used, one 2-to 6-A 51-V relay also may be used to provide thebalanced fault backup function. A simple overcurrentunit is unsuitable for preventing a sustained machinecontribution to a fault because, with a regulator out ofservice, the bolted sustained (or synchronous) three-phase fault contribution is less than machine full-loadcurrent. The 51-V relay, on the other hand, can be setwell below full-load current and not operate on load.Its overcurrent unit is supervised or torque-controlledby an undervoltage unit, and therefore voltage must bebelow the voltage setting to permit the overcurrent unitto function. Both the voltage and current units areindependently adjustable, making coordination withother overcurrent devices simpler than if the currentunit response were a function of voltage level.

Figure 8-9 Comparison of relay and generator character-

istics.Figure 8-10 Schematic connections of the COQ relay for

unbalanced fault protection.


6 OVERLOAD PROTECTION

6.1 RTD Schemes (Device 49)

Most large generators are equipped with resistancetemperature detectors (RTDs), which may be used in abridge circuit to provide sensing intelligence to anindicator or a relay such as the DT-3 or 49 T.

This relay is restrained when the resistance is low,indicating low machine temperature. When the tem-perature of the machine exceeds some preset level suchas 120 8C for class B insulated machines, the bridgebecomes unbalanced and the contacts close.

6.2 Thermal Replicas (Device 49)

A thermal replica relay utilizes stator current toapproximate the heating effects in the generator. Themachine thermal time constants on heating and cool-ing are represented to take cognizance of previous andpresent loading effects. When the replica indicates thattemperature in excess of the allowable value for themachine insulation has been reached, tripping takesplace.

7 VOLTS PER HERTZ PROTECTION

From the fundamental expression for induced voltagein a coil,

E ¼ 4:44 fANBm 10�8

where

E¼ induced rms voltage in volts

f¼ frequency in hertz

A¼ cross-sectional area of the core in square inches

N¼ number of turns

Bm¼ flux density in maxwells per square inch

Since all the elements of the equation are constantsexcept E and f, it can be seen that

Bm aE

f

Flux density is an excellent indicator of no loadheating effect. Hysteresis and eddy current losses areeach proportional to a power of the flux density.Therefore, impending overheating can be recognizedby measuring volts per hertz.

Overexcitation can be caused by an attempt by agenerator voltage regulator to maintain rated voltageduring coast-down or holding manual excitation at afixed level during acceleration. Since the limits ongenerators and transformers are inverse-time-related(i.e., a higher volts per hertz value is permitted for ashorter time to stay within the bounds of acceptableheating), an inverse-time, volts-per-hertz relay such asthe MVH should be used to protect these devices whenoverexcitation is likely. Figure 8-11 shows an exampleof the coordination that can be achieved.

8 OVERSPEED PROTECTION

A generator accelerates when it becomes separatedfrom its load. The acceleration depends on the inertia

Figure 8-11 Example of volts/hertz protection using MVH

relay.

126 Chapter 8

ðWR2Þ, load loss, and governor response. To recognizeoverspeed, a permanent magnet generator is oftenconnected to the machine shaft to supply a voltage tothe governor that is proportional to speed. Thegovernor may also be equipped with a speed-respon-sive flyball mechanism. Either the permanent magnetgenerator or flyball mechanism can initiate prime-mover control to remove power input and alleviateoverspeed. An overfrequency relay, such as an MDF(device 81), can be used to supplement this overspeedequipment.

9 LOSS-OF-EXCITATION PROTECTION

9.1 Causes of Machine Loss of Field

Loss of excitation can occur as a result of

Loss of field to the main exciterAccidental tripping of the field breakerShort circuits in the field circuitsPoor brush contact in the exciterField circuit-breaker latch failureLoss of ac supply to the excitation systemReduced-frequency operation when the regulator is

out of service

Some relays such as the KLF and KLF-1 containmultiple operating units. The KLF and KLF-1 contain(1) a directional unit, (2) an offset mho unit, and (3) aninstantaneous undervoltage unit.

Loss of excitation produces voltage and currentvariations as shown in Figure 8-12. This causes atrajectory on the R-X diagram such as that depicted inFigure 8-13b. The initial operating point is dependenton load level and power factor angle prior to the loss ofexcitation. When reactive power begins to flow into themachine, the locus moves into the �X region. As thetransient continues, with the field collapsing, the locusmoves into the characteristic circle of the loss-of-fieldrelay. If the relay is equipped with an undervoltageunit and at this time the voltage is sufficiently belownormal to operate it, tripping takes place after a shortdelay (X dropout). (See Fig. 8-13a.)

If reduction in terminal voltage is not appreciable,the undervoltage unit does not drop out. This isindicative of a loss of excitation condition that is notlikely to affect system stability nor influence adjacentmachines significantly. For this series of conditions, itwill suffice to sound an alarm to alert the operator toallow him or her to take action to restore the field oranticipate shutdown. A timer is often run by Z and D

and V to initiate partial shutdown if the operator isunable to correct the problem quickly.

These relays can also be used to detect loss of fieldin a synchronous condenser or motor. External faultswill cause the D and/or Z functions to restrain to avoidundesirable tripping of the machine.

Other relays such as the solid-state type 40 relayperform the function with only the reduced-diameterimpedance measurement.

9.2 Hazard

The generator must be kept on line, supplying poweras long as possible, particularly when the machinerepresents a sizable portion of the system capacity. Tothis end, an early warning of low excitation would givethe operator an opportunity to restore the field ifpossible and avoid tripping. Unnecessary tripping andthe resultant loss of kW output can precipitate systembreakup and a major outage.

Figure 8-12 Underexcited operation of generator.


9.3 Loss-of-Field Relays

A relay designed to detect low excitation shouldperform the following functions:

Alert the operator to any low excitation that coulddamage the machine or result in instability

Alert the operator to a loss-of-field condition asearly as possible, giving him or her time tocorrect the condition

Trip the machine automatically in the case ofimpending system instability

The KLF-1 differs from the KLF relay in that it hasa separate phase voltage supply for each of threedifferent measuring elements. As a result, loss of anyone phase voltage to the KLF-1 relay cannot causeincorrect tripping. With the KLF, some combinationsof load and phase voltage loss can operate the relay.The KLF-1 must have a wye-wye voltage supply, withthe neutral brought to the relay. The KLF may be usedwith a wye-wye, delta-delta, or open-delta-open-deltasupply.

When partial or complete loss of excitation occurson a synchronous machine, reactive power flows fromthe system into the machine and the apparentimpedance as viewed from the machine terminals(Vt/I) goes into the negative X region (Fig. 8-12).The kW output is controlled by the prime-mover input,whereas kV Ar output is controlled by the fieldexcitation. If the system is large enough to supply thedeficiency in excitation through the armature, thesynchronous machine will operate as an induction

generator, supplying essentially the same kW to thesystem as before the loss of excitation.

Loss of synchronism does not require immediatetripping unless there is an accompanying decrease inthe terminal voltage that threatens the stability ofnearby machines. Generally, it takes at least 2 to 6 secto lose synchronism. Many instances have beenreported in which machines have run out of synchron-ism for several minutes because of loss of excitationwithout damage to the machines.

9.4 KLF and KLF-1 Curves

Figure 8-14 shows how a generator capability curvecan be transformed into an R-X diagram. For eachpoint on the curve, an angle b can be measured fromthe horizontal and the value of three-phase MVA read.Knowing the line-to-line voltage at which the cap-ability curve applies, a value of Z can be calculatedusing

Zp ¼ kV2

MVAprimary ohms ð8-3Þ

Z ¼ kV2ðRcÞMVAðRvÞ secondary ohms ð8-4Þ

where Rc and Rv are the current and voltagetransformer ratios, respectively.

Point Zp or Z can then be plotted, at angle b, on theR-X diagram. Other key points on the circle arcs can

Figure 8-13 Trip circuits and R-X diagram showing operation of the KLF (40) loss-of-field relay.

128 Chapter 8

be obtained in the same way until the entire capabilitycurve is transformed.

The steady-state stability curve is another signifi-cant limit that can be related to a loss-of-field relaywith impedance-measuring qualities. The MW-MVAR curve can be developed as shown inFigure 8-15a. In this figure, V is the per unit terminalvoltage, Xs the equivalent per unit system impedanceas viewed from the generator terminals, and Xd theper unit unsaturated synchronous reactance of thegenerator. Both Xs and Xd are measured on themachine MVA base.

Figure 8-15b converts the machine’s steady-statestability curve to an R-X diagram. Note that the curveof Figure 8-15b can be plotted directly from aknowledge of Xs and Xd without the intermediate

step of Figure 8-15a. To be useful in setting a loss-of-field relay, these per unit values must be converted tosecondary ohms.

Figure 8-16 relates KLF (or KLF-1) setting tocapability and minimum excitation limiter (MEL)curves. Assume a given kW load on the machine andthat the vars into the machine are gradually beingincreased by decreasing machine field current,producing a trajectory, as in curve A of Figure 8-16. If the regulator is in service, the MEL preventsoperation at a level that would jeopardize themachine thermally. If the regulator is out of service,Z continues to decrease until the KLF impedanceunit operates. An alarm indicates a hazardousoperating condition if the voltage is high. A lowvoltage, which may seriously jeopardize systemstability, trips the machine after 0.25 sec (Fig. 8-13a). The loss-of-field relay must reach into the plusX area if its locus is to follow closely the machinecharacteristic. A directional unit is included in therelay to avoid tripping for close-in faults beyond theunit transformer.

9.5 Two-Zone KLF Scheme

Like all other elements in generator protection, theloss-of-field relay should be supported by backup toprevent catastrophic failure if a device is out of serviceor a component should fail.

Two loss-of-field relays provide better protectionthan one. The first, or zone 1, relay is set to berestrictive (Fig. 8-17) and typically trips through a0.25-sec timer. It provides fast clearing on loss of field,yet is secure against swings such as that shown passingthrough points CDEF in Figure 8-17. The zone 2 relayis set wider and typically drives a 1-sec timer to detectpartial loss of field, provide an alarm function, andback up the zone 1 relay. A total functionalcharacteristic similar to the KLF type is required todo this. Other setting data are given in Tables 8-1 and8-2.

Figure 8-18 shows the dc schematic for the KLF orKLF-1 zone 2 relay. Zone 1 may be used without aTD-A timer unless extreme emphasis on security ismade. No TD-2 relay is required if the undervoltagecontacts of the zone 1 relay are shorted or the type 40,solid-state relay is used. Without TD-2 quick trippingoccurs when Z and D operate irrespective of thevoltage level.

Figure 8-14 Transformation from KW-KVAR plot to R-X

plot.


10 PROTECTION AGAINST GENERATORMOTORING

Generator motoring (sometimes referred to as reverseactive power) protection is designed for the primemover rather than the generator. With steamturbines, for example, turbines will overheat on lowsteam flow, but will be protected by steam tempera-ture devices. With hydroturbines, hydraulic flowindicators protect against blade cavitation on lowwater flow. Similar devices are used to protect gasturbines.

Generator motoring protection can be provided bydevices such as limit switches or exhaust-hoodtemperature detectors. However, a reverse-powerprotective relay is recommended for added safety.

Due to the extreme nature of this hazard, twomotoring relays are recommended to operate in anOR mode and respond to different current andvoltage. The reverse-power relay is commonly usedwith diesel engine generating units, particularly whenthe danger of explosion and fire from unburned fuelexists.

Motoring results from a low prime-mover input tothe ac generator. When this input cannot satisfy all thelosses, the deficiency is supplied by the generatorabsorbing real power from the system. Since fieldexcitation should remain the same, the same reactivepower would flow as before motoring. Thus, onmotoring, the real power will flow into the machine,whereas the reactive power may be either flowing outof or into the machine. Usually, the reactive power will

Figure 8-15 Conversion of steady-state stability curve to R-X.

130 Chapter 8

be supplied to the system as machines are not generallyoperated underexcited.

A relay designed to detect motoring must beextremely sensitive and even then cannot detect allconditions of reverse power. For example, suppose a

turbine had its valves closed to slightly less than theno-load steam requirements. The turbine wouldsupply, say, 99% of the losses, and the generator (asa motor) would supply 1%. If the total losses were3.0% of the kW rating, the kW drawn by the generator(as a motor) from the power system would be only1.0% of 3.0%, or 0.03%, of the nameplate rating. Thisis a challenge beyond the capabilities of most motoringdetection relays. Solid-state and microprocessor tech-nology allows such low sensitivity while, at the otherend of the scale, providing full-load current capability.The SRW relay may be set as low as 1mA.

When the prime mover is spun at synchronousspeed with no power input, the approximate reversepower required to motor a generator, as a percentageof the nameplate rating in kW, is as follows:

Condensing steam turbine 1 to 3%Noncondensing steam turbine 3þ%

Diesel engine 25%Hydraulic turbine 0.2 to 2þ%

10.1 Steam Turbines

When operating under full vacuum and zero steaminput, condensing turbines require about 3.0% of thekW rating to motor. Noncondensing turbines require3.0% or more of the rated kW to motor whenoperating against atmospheric or higher exhaustpressures at zero steam flow.

10.2 Diesel Engines

If no cylinders are firing, diesel engines require about25% of the rated kW figure. If one or more cylindersare firing at no load, the reverse power will be lowerthan this depending on the governor action and effecton the system frequency.

10.3 Gas Turbines

The large compressor load of gas turbines represents asubstantial power requirement from the system whenmotoring. Consequently, the sensitivity of the anti-motoring device is not critical.

10.4 Hydraulic Turbines

When the blades are under the tail-race water level,the percent of kW rating required for motoring is

Figure 8-16 KLF setting related to capability and MEL

curves.

Figure 8-17 KLF relay stable swing following clearing of

nearby three-phase fault.


probably well over 2.0%. From 0.2 to 2.0% kW isrequired for the turbine to motor when the bladesare above the tail-race level. For turbines using aKaplan adjustable-blade propeller, the flat-bladecondition probably requires less than 0.2% kW tomotor.

Most commonly, a single-phase relay is used formotoring detection, with sensitivities ranging from1mA to 5A. All have adjustable operating times.When the more sensitive relays are used, care should beused in selecting a current transformer for them.Indeed, a metering accuracy-class ct is preferred to arelaying accuracy class. A typical schematic is shown inFigure 8-19.

11 INADVERTENT ENERGIZATION

Many instances of inadvertent energization haveoccurred over the years. A few of the causes were

Closing the generator breaker with the machine atstandstill

Closing a station service breaker with the machineat standstill

High-voltage breaker flashover near synchronismClosing of generator disconnect with unit breaker

closed

Although careful operating procedures and the judi-cious use of interlocks can usually prevent these

Table 8-1 Recommended Settings for KLF Relay

Setting Zone 1 (alone) Zone 2 (alone) Both zone 1 and zone 2

Impedance setting See Figure 8-17. See Figure 8-17. See Figure 8-17.

Voltage setting (a) Undervoltage contact

shorted or

(b) set at 80% for security.

80% Zone 1 voltage contact

shorted.

Zone 2 dropout voltage set at

80%.

TD-1 (see Fig. 8-18) 1/4 to 1 sec (1/4 see adequate). 1/4 to 1 sec (1 sec preferred). Zone 1 timer¼ 1/4 sec.

Zone 2 timer¼ 1 sec.

TD-2 (see Fig. 8-18) Not required for (a) above. 1min. None for zone 1. zone 2

For (b) above use 1min. timer¼ 1min.

Advantages Less sensitive to stable system

swings.

(1) More sensitive to LOF

condition.

(1) Same as (1), (2), and (3) at

left.

(2) Can operate on partial

LOF.

(2) Provides backup

protection.

(3) Provide alarm features for

manual operation.

Table 8-2 Special Settings for Multi-machines Bussed at Machine Terminals

Setting Zone 1 (alone) Zone 2 (alone) Both zone 1 and zone 2

Impedance setting See Figure 8-17. See Figure 8-17. See Figure 8-17.

Voltage setting (a) Undervoltage contact

shorted or

(b) set at 87% for security.

87%. Zone 1 voltage contact

shorted with zone 2 set at

87%.

TD-1 (see Fig. 8-18) 1/4 to 1 sec (1/4 sec adequate). 1/4 to 1 sec (1 sec preferred). Zone 1 timer¼ 1/4 sec.

Zone 2 timer¼ 1 sec.

TD-2 (see Fig. 8-18) Not required for (a) above. 10 sec for directly cooled. None for zone 1.

For (b) above use 10 sec for

directly cooled, 25 see for

indirectly cooled.

25 sec for indirectly cooled. Zone 2 timers: 10 sec for

directly cooled, 25 sec for

indirectly cooled.

132 Chapter 8

occurrences, the ingenuity of humans to circumventthese procedures and interlocks is legendary.

Full-voltage energization of a machine at standstilldoes not produce an enormous magnitude of current,but it does supply an extreme impact of torque, andmechanical damage to the shaft or bearings may occur.The resulting current is of sufficient magnitude thatfast removal is necessary if thermal damage to thegenerator is to be avoided. Instantaneous separationoffers no guarantee that no damage will occur.

Various relays applied for other functions maydetect inadvertent energization. Loss-of-field relaysmay respond, but they usually use single-phase current,and therefore, complete sensing is not afforded for allpossible combinations of this phenomenon. Relaysapplied to detect motoring will operate, but theiroperating time is set to other criteria and the long timecustomarily used is unsuitable for this additionalfunction. Distance- and voltage-controlled overcurrentrelays applied for generator phase backup protection

may operate, but again the time delay for tripping maybe unsuitable.

To further compound the difficulties associated withdetecting inadvertent energization is the fact thatgenerator potential circuits are often disconnected inthe interests of safety when a machine is shut down.Any of the ‘‘normal’’ relays that are dependent on thisvoltage supply will be unable to respond at the verytime when they are needed. Also, it must beremembered that the flashover of an open breakercannot be cleared by energizing its trip coil.

Several protective schemes have been used success-fully to detect three- or single-phase inadvertentenergization. Among them are

Directional overcurrent relaysPole disagreement relaysRelays containing logic to detect overcurrent for a

short time following 0VFrequency-supervised overcurrent relays

Figure 8-18 Type KLF or KLF-1 dc schematic for zone 2 loss of excitation protection. (Timer settings are given in Tables 8.1

and 8.2.)


Voltage-supervised overcurrent relaysDistance relaysUnit transformer neutral overcurrent in special

breaker-failure relaying scheme.

Care must be exercised to assure that either a voltagesupply is available when operation is required orundervoltage allows operation of the scheme. Further,the circumstances associated with inadvertent energi-zation itself must not be able to circumvent a necessarypart of the logic for tripping, such as the requirementfor the presence of reduced frequency.

12 FIELD GROUND DETECTION

A single ground on the field of a synchronous machineproduces no immediate damaging effect. It must bedetected and removed because of the possibility of asecond ground that could short part of the fieldwinding and cause damaging vibration. Care should beused in establishing any field ground detecting schemeto assure that any bearing current that is allowed willnot cause bearing deterioration. One scheme that hasbeen used is shown in Figure 8-20. A small leakagecurrent flows through the field-to-ground capacitance,

Figure 8-19 Typical schematic for antimonitoring protection using CRN-1 (32) relay.

134 Chapter 8

which, on a large turbogenerator, can be between 0.3and 0.5 mf. The relay detects an increase or decrease inthe magnitude of this current.

12.1 Brush-Type Machine

One recommended field ground protection scheme fora generator with brushes (i.e., stationary field leadsaccessible) is illustrated in Figure 8-21. This scheme,which does not require any external source, uses thevery sensitive d’Arsonval dc relay, type DGF.

The DGF relay uses a voltage divider circuit,consisting of two linear resistors (R1 and R2) and anonlinear resistor whose resistance varies with theapplied voltage. If the field becomes grounded, avoltage will develop between point ‘‘M’’ and ground.

The magnitude of this voltage will vary according tothe exciter voltage and point at which the field isgrounded. The voltage will be at maximum if the fieldis grounded at either end of the winding.

A null point will exist in the field winding where aground will produce no voltage betweenM andground. This null point will be located at a point onthe field winding from which there is balance betweenthe two field winding resistances and two relayresistances to positive and negative. A ground at thenull point will go unrecognized until the field voltage ischanged as a result of daily reactive (or voltage-level)scheduling. Faults having impedance of up to300,000O can be recognized with this scheme.

A pushbutton, connected across a portion of the R2

resistor, permits a manual check for possible groundfaults at the center of the winding. This provision is

Figure 8-20 Path of the currents in a machine when using an ac field ground relay.


desirable when the generator is to be ‘‘base-loaded’’and will not experience periodic excitation variations.

Another scheme that has been used successfullyutilizes a Wheatstone bridge with the field circuit toground forming one leg of the bridge. A solid groundfault anywhere in the field circuit can be detectedimmediately through the recognition of the resultingunbalance.

12.2 Brushless Machines

For a ‘‘brushless’’ type of machine, no normal access isavailable to a stationary part of the generator fieldcircuit, and no continuous monitoring to detect fieldgrounds is possible. One widely employed scheme usesa 60-Hz tuned overvoltage relay connected between theneutral of the three-phase ac exciter and ground. Aground on the exciter, in the three-phase rectifierbridge, in the field, or on the dc leads will be detected.

This requires a pilot brush connected to the neutral ofthe exciter that is periodically dropped.

Another variation of this form of detection is usedby the solid-state type 64F relay. It impresses a dcvoltage from the negative dc lead to ground andcontinuously monitors current flow. An increase incurrent accompanies a field ground.

The pilot brush arrangement can also be used withthe Wheatstone bridge scheme (YWX111) described inSection 12.1.

12.3 Injection Scheme for Field GroundDetection

The scheme described in Section 4.4, GIX-104, isequally applicable to field ground detection, irrespec-tive of the type of excitation system in use. This isshown in Figure 8-22 for a ‘‘rotating rectifier’’excitation system. As with the stator winding grounddetection application, this system is able to detect fieldcircuit grounds even though the machine is at standstillor running, excited or not.

13 ALTERNATING-CURRENT OVERVOLTAGEPROTECTION FOR HYDROELECTRICGENERATORS

Alternating-current overvoltage protection is recom-mended for hydroelectric generators subject to over-speed and consequent overvoltage on loss of load.Some hydroelectric generators can go up to 140% ormore of rated speed when full load is dropped. Thevoltage may reach 200% or more.

The ac overvoltage protective scheme is shown inFigure 8-23. The relay, which changes the excitation toreduce the output voltage, can also provide backupprotection for the voltage regulator.

14 GENERATOR PROTECTION AT REDUCEDFREQUENCIES

Many turbine generators are started on turning gear,which rotates the shaft at about 3 rpm. For a cross-compound machine, the field must be applied beforethe machine is removed from the turning gear. At thispoint, excitation should be limited to rated volts perhertz to avoid overexciting the unit or station servicetransformer. A tandem unit need not have field applieduntil it is up to speed and ready to synchronize.

Figure 8-21 Field ground protection scheme for a gen-

erator.

136 Chapter 8

Cross-compound generators may be operated forseveral hours during warmup at frequencies well belowtheir rating. Current transformer and relay perfor-mance must be considered at these reduced frequenciesbecause fault magnitudes are approximately the sameas at rated frequency. Current transformer perfor-mance can be expected to deteriorate badly at lowfrequency. There may be a small compensating effect,however, in the reduction of burden impedance.

The performances of some electromechanical relaysassociated with the generator or a generator-transfor-mer unit at 15 and 30Hz are summarized in Table 8-3in terms of 60-Hz performance.

For cross-compound turbine generators, the low-and high-pressure units should have their fields appliedand be synchronized while on turning gear, so thatthey are brought together up to rated speed. Synchro-nizing surges may occur, at these low speeds, that willoperate the loss of field relays. These synchronizingsurges can reach 60% of the full-load current, since theimpedance in the generators is very low. Loss-of-fieldrelays applied to a cross-compound unit should there-fore be disabled during startup.

The SC relay (and/or the SV) is recommendedas supplementary protection when reduced fre-quency protection is required. The SC currentrelay has a flat characteristic and increases slightlyin sensitivity as the operating frequency drops.When an SC relay is operated on dc, it picks upat approximately 15% below its normal 60-Hzpickup. The pickup of the SV voltage relay isalmost directly proportional to frequency; itssensitivity at 15Hz is thus 4 times the sensitivityat 60Hz. For this reason, the SV relay providesexcellent reduced-frequency protection.

In this application, an SC relay is frequently locatedin the differential circuit of each phase of the generatordifferential relay. The SV relay is connected across thesecondary resistor in the generator neutral circuit.

None of the relays listed in Table 8-3 will overheatnor operate incorrectly if left in the circuits when thegenerator is operated at reduced frequencies. The KLFand KLF-1 relays, when used in a cross-compoundconfiguration, must have their trip incapacitatedduring startup. The SC and SV relays used for low-frequency protection must be removed from service fornormal operation. This may be done by a frequencyrelay set for approximately 55Hz (for a 60-Hz system)and an auxiliary relay.

The microprocessor-based multifunction relay TypeGPU2000R has frequency-tracking algorithms. Theseinsure that its protective elements which are most

Figure 8-22 Rotating rectifier excitation system.

Figure 8-23 Overvoltage protection for generator.


needed during low-frequency operation maintain theircharacteristics.

15 OFF-FREQUENCY OPERATION

Turbine blades are carefully designed to have nomechanical resonant conditions when rotating at ratedspeed. If this were not the case, mechanical deteriora-tion would occur as the blades flex under the stress ofloaded operation. At elevated or reduced speed, thereare resonant points where prolonged operation pro-duces blade fatigue damage and ultimate failure.

If operating frequency (speed) deviates from therated value, corrective action must be initiated ortripping must result. Since mechanical fatigue is acumulative phenomenon, the time of loaded operationat reduced frequency must be monitored and accumu-lated over the life of the turbine. Figure 8-24 shows thelimits that one manufacturer imposes for machines in

two categories: (1) those having the longest stageblading of length 18 to 25 in. and (2) those having thelongest stage blading of length 28 1

2 to 44 in. Thesecurves are not a national or international standard.The specific manufacturer of the turbine should becontacted to ascertain specific recommendations forlimits.

Under- (or over-) frequency relays as described inthe chapter on ‘‘load-shedding’’ may be used to detectfrequency excursions, but this must be monitored bywatt level because off-frequency operation at a lightload (above the level that produces adequate steamflow to remove turbine friction and windage losses) isnot hazardous. The combination of these two sensingelements is used to drive an accumulation timer toallow an estimate of the extent of life remaining to bemade. The need for blade examination and possiblereplacement can then be evaluated.

The microprocessor-based system REG-216 has aprovision for integrating low-frequency overtime and

Table 8-3 Performance of ABB Protective Relays at Reduced Frequencies

Pickup in percent of 60-Hz pickup

15 Hertz 30 Hertz Classificationa

Overcurrent (51) CO-2 165 115 A

CO-5 b 150 B

CO-6 b 143 B

CO-7 b 140 B

CO-8 262 138 A

CO-9 260 140 A

SC 85 93 A

COV Performance same as CO Unit used in relay

Voltage (59) SV 26 50 A

CV contact-making

voltmeter

122 120 A

CV-8 b b C

Differential (87) CA generator 255 123 A

CA transformer b 149 B

SA-1 370 175 A

HU-1 250 130 A

Negative sequence COQ O

(46) SOQ

Loss of field (40) KLF (or KLF-1) D

a(A) Protection available at both 15 and 30 cycles. (B) Protection available at 30 cycles only. (C) Additional protective relays required for start-up

or low-frequency operation. (O) The sensitivity of the COQ & SOQ relays to negative sequence currents is a direct function of frequency, while its

sensitivity to positive sequence currents is an inverse function of frequency. This relay will operate for heavy three-phase and phase-to-phase faults

at reduced frequencies, but should not be relied upon for primary protection during warm-up. (D) Since the KLF or KLF-1 relays operate on

lagging reactive power into the machine, the relay will neither operate falsely nor provide loss-of-field protection during the warm-up period.bVery insensitive or nonoperable at the frequency indicated.

138 Chapter 8

alarming and tripping to prevent damage to theturbine. Also, to avoid any detrimental effects ofthe front-end filters in detecting overcurrent at off-frequency levels, the 50, instantaneous-trip functionis sensed ahead of these filters. Overcurrent protec-tion, then, is provided at a frequency as low as2Hz.

16 RECOMMENDED PROTECTION

Figure 8-25 shows the recommended protection forlarge tandem-compound unit-connected turbine gen-erators. Figures 8-26 and 8-27 illustrate the recom-mended protection for machines that are not unit-connected. Generally, such generators are used inindustrial applications.

A wide variety of implementations of the numberedfunctions are in use and some are described in Table8-4. The REG-100 and REG-216 are complete multi-function microprocessor-based packages.

17 OUT-OF-STEP PROTECTION

As generator impedances become larger in proportionto the system impedance, the electrical center will becloser to the generator. This condition intensifies theneed for out-of-step detection as part of the generatorrelaying complement. Such relaying schemes are

described in Chapter 14, ‘‘System Stability and Out-of-Step Relaying.’’

18 BUS TRANSFER SYSTEMS FOR STATIONAUXILIARIES

The automatic transfer of highly essential stationauxiliary loads such as reactor coolant pumps, boilerfeed pumps, and induced draft fans is commonpractice. Paralleling the normal and emergency sourceson a continuous basis is not generally recommendedbecause the higher breaker-interrupting dutiesinvolved can cause problems, as can circulatingcurrents between systems. Transfers should not bemade if voltage of the alternate supply is notsatisfactory or the load circuits are faulted. Also,supply breaker tripping should be delayed long enoughto permit fault sectionalizing in the load circuits.

18.1 Fast Transfer

This is a term applied to the connection of a bus to asecond power source with little or no time delay. Thisprocess is accomplished with an ‘‘open’’ transfer or‘‘closed’’ transfer. The open transfer disconnects thenormal source before connecting the second. Theclosed transfer allows the two systems to operate tiedtogether momentarily, and then to have the originalsource breaker tripped. For all the fast transfer

Figure 8-24 One manufacturer’s limits for off-frequency operation of combustion turbines.


schemes, it is recommended, if a synchronism checkrelay is required, that it be preenergized so that it maydetermine the existence (or lack of) synchronism priorto the need to transfer.

Figure 8-28a shows an example of the fast open-transfer scheme. 43 is a switch having automatic andmanual positions. With 43A closed, CVX synchronismcheck relay contact closed due to the two buses beingin synchronism prior to the need for transfer, breakerA having been tripped for any reason (as evidenced byits 52b being closed), and no fault having occurred onthe bus to produce the opening of 86B, then the closecircuit of breaker C becomes energized.

The simple process of tripping breaker A, when inthe automatic mode in the absence of a bus fault,produces the closure of breaker C.

18.2 Choice of Fast Transfer Scheme

Open transfer would be selected in a system in which thesupportiveWR2 (moment of inertia) is sufficient to keepthe switched bus close enough in frequency during theopen period to allow reconnection without an excessiveshock to rotating machinery connected to that bus.

Figure 8-25 Overall protection for a tandem-compound unit-connected generator.

140 Chapter 8

Figure 8-26 Recommended protection for large generators as used in industrial plants.

Figure 8-27 Recommended protection for small generators as used in industrial plants.


The ‘‘simultaneous’’ scheme is a variation of theopen transfer scheme. With it, the trip-coil of thenormal supply breaker and the close-coil of theincoming supply breaker are energized simultaneously.Since the close function is somewhat slower than thetrip function of a given type of breaker, the openingbreaker wins the race. Figure 8-28c shows the contactarray for this sequence.

Closed transfer would be the proper scheme to usewhen WR2 associated with the rotating equipment isinadequate to allow separation for even a short time.The impact of reenergization would be too great.Figure 8-28c shows the addition of a ‘‘52a’’ contact ofbreaker C to delay the tripping of breaker A on amanual transfer. It is not considered wise to delaytripping for a fault that operates 86S, the lockout relayassociated with a source circuit fault.

Synchronism check is a satisfactory means forsupervising the closing of the bus tie breaker. It isnot suitable, however, for a fault-forced transfer unlessthe system is supervised by an 86B lockout relay thatprevents transferring to a bus fault. Otherwise, itwould be possible for both sources to be connected to afault, compounding the damage and producing nouseful result.

18.3 Slow Transfer

Slow transfer carries with it the connotation ofdelaying the reenergization of a bus until the voltageon motors connected to the bus has decayed to the

Table 8-4 Generator Relaying

ANSIType

device no. Description Traditional Solid-state or numerical GPU-2000R REG 216 REG 316

2 Timer — 62T X X X

TD-5

21 Phase backup KD-11 a X X X

24 Volts/hertz — 59F X X X

MVH

32 Motoring CRN-1 32R X X X

SRW

40 Loss of field KLF 40 X X X

46 Negative sequence COQ 46Q X X X

SOQ

49 Thermal DT-3 49T — X X

50/51 Stator overcurrent CO-ITH Micro 51 X X X

59 Overvoltage CV-5 59 X X X

59F Field ground CV-8 59G — X X

64F

59N Stator ground (95%) CV-8 59G X X X

64S Current injection — GIX-104 — X

64R (100% ground stator/rotor)

67 Inadvertent-energization CRG-9 32D X X X

76 Field dc overcurrent D-3 76H — — —

78 Out-of-step KST-KD-3 GZX-104 — X X

MDAR

81 Underfrequency — 81 X X X

MDF

86G Generator lockout LOR LOR LOR LOR LOR

87G Generator differential CA 87M X X X

SA-1

87N3 Stator-neutral-ground DGSH 27G X — X

87T Overall differential HU-1 87T — X X

aAlternate 51Lþ 47H or 51V function.

142 Chapter 8

point where no damage can be expected with out-of-phase energization. This level is generally considered tobe 25% of rated. Figure 8-28b describes this arrange-ment.

19 MICROPROCESSOR-BASED GENERATORPROTECTION

Important features related to generator protection canonly be achieved by a coordinated generator protectionpackage, utilizing microprocessors. Self-monitoring,communications, oscillography, and adaptive settingsare accomplished straightforwardly and reliably. Flex-ibility in the application is important to allowindividual user selections. This is accomplished by aman-machine interface that allows ease of settings andin some cases the choice of software modules.

In the use of multifunction coordinated packages,sight should not be lost of the need for adequateredundancy to provide backup and cover the failure ofany element. The REG 216 and REG 100 seriesaddress these needs and provide extensive generatorprotection utilizing proven concepts.

Figure 8-28 Variations of transfer schemes.


9

Motor Protection

Revised by: C. L. DOWNS

1 INTRODUCTION

1.1 General Requirements

Motor protection is far less standardized than gen-erator protection. Although the National Electric Codeand NEMA standards specify basic protection require-ments, they do not fully cover the many different typesand sizes of motors and their varied applications. Thereare many other schemes, all of which offer differentdegrees of protection. As with generator protection, thecost and extent of the protective system must beweighed against the potential hazards. The size of themotor and type of service will also influence the type ofprotection required. Electromechanical, solid-state, ormicroprocessor-based relays can be used stand-alone orin combination with one another to achieve the desireddegree of security and dependability.

Motor protection should involve the detection ofthe following hazards:

1. Faults in the windings or associated feedercircuits, including both phase and ground faultdetection.

2. Excessive overloads. Overloads result in ther-mal damage to the insulation and can be causedby continuous or intermittent overload, or alocked rotor condition (failure to start or a jamcondition).

3. Reduction or loss of supply voltage. Any re-duction of supply voltage directly affects theapplied torque to the connected mechanical load.

4. Phase reversal. Starting a motor in reverse canbe hazardous to the load.

5. Phase unbalance. A small amount of unbalancecan result in a significant increase in the motortemperature.

6. Out-of-step operation for synchronousmotors.

7. Loss of excitation for synchronous motors.

Protective relays applied for one hazard may operatefor others. For example, a relay designed to operate onan excessive overload could also protect against a faultin the windings.

Protective devices may be installed on the motorcontrollers or directly on the motors. The protection isusually included as part of the controller, except forvery small motors, which have various types of built-inthermal protection.

Motors rated 600V or below are generallyswitched by contactors and protected by fuses orlow-voltage circuit breakers equipped with magnetictrips. Motors rated from 600 to 4800V are usuallyswitched by a power circuit breaker or contactor(often supplemented by current-limiting fuses toaccommodate higher interrupting requirements.Motors rated from 2400 to 13,800V are switchedby power circuit breakers.

Although protective relays may be applied to amotor of any size or voltage rating, in practice they areusually applied only to the larger or higher-voltagemotors.

145

1.2 Induction Motor Equivalent Circuit

Excessive heat in the motor can be caused duringstarting, a locked rotor condition, load requirements,voltage unbalance, or an open-phase condition andcan cause degradation of the mechanical and dielectricstrength of the insulation. This thermal deteriorationof the insulation sets up the possibility of future faults.

The equivalent circuit of the motor helps invisualizing what occurs in the motor during the aboveconditions. The motor impedance during these condi-tions is directly influenced by the slip of the motor:

slip s ¼ ns � nr

nsper unit

where ns and nr are the stator field and rotor speeds. Atstart, the slip is 1.00, or 100%. During a runningcondition, the slip is approximately 0.01 to 0.08, or 1 to8%. The equivalent circuit during a starting conditionis as shown in Figure 9-1. The input impedance at startcan be approximated by

Zstart ¼ Rs þRr þ jXs þ jXr

where Rs, Rr, jXs, jXr are the stator and rotorresistances and reactances. During a start condition,the impedance is predominately reactive. As the motorgains speed, the impedance becomes more resistive,and the power factor increases.

The negative sequence impedance of a motor,excluding wound rotor motors, is very nearly equalto Zstart above, and from Figure 9-2, can beapproximated by

Z2 ¼ Rs þ RR

2� sþ jXs þ jXR

With Z2 and Zstart approximately equal, the negativesequence current can be calculated for a particularnegative sequence voltage unbalance. If V2¼ 5% and

the starting current Is of the motor is 8IFL, then I2would be 0.40, or 40%. This will result in an increase instator heating and, in particular, rotor heating. Therotor heating results from the combined effect of thecounterrotating flux that causes large currents to beinduced at fs(2� s) Hz, where fs is the systemfrequency, and the increased skin effect in the rotorthat can cause its resistance to be 5 to 10 times normal.

1.3 Motor Thermal Capability Curves

Protecting a motor for a variety of hazards requires theprotection engineer to know full-load current, permis-sible continuous allowable temperature rise, locked-rotor current and permissible maximum time at thatcurrent, and accelerating time, which is a function ofthe load characteristics and starting voltage. A typicalmotor thermal capability curve, which is shown inFigure 9-3, is helpful in determining the temperatureendurance of the insulation. The lower part of thecurve is usually rotor-limited. The limit arises becauseof the I2R heating effect during a locked rotor

Figure 9-1 Induction motor equivalent circuit at start.

Figure 9-2 Induction motor negative sequence network.

Figure 9-3 Motor thermal capability curve.

146 Chapter 9

condition. This represents the time a motor can remainstalled after being energized before thermal damageoccurs in the rotor. This is an I2t limit, which can alsobe expressed as a (V2/R)t limit. The middle portion isthe acceleration thermal limit part of the curve. This isfrom the locked-rotor current to the motor breakdowntorque current portion of the curve. The upper sectionof the curve is the running or operating thermal limitportion. This represents the motor overload capacity.

2 PHASE-FAULT PROTECTION

The phase-fault current at the terminals of a motorusually is considerably larger than any normal current,such as starting current or the motor contribution to afault. For this reason, a high-set instantaneous-tripunit is recommended for fast, reliable, inexpensive,simple protection. When the starting current valueapproaches the fault current, however, some form ofdifferential relaying becomes necessary. The sensitivityof the differential relay is independent of startingcurrent, whereas instantaneous-trip units, whichrespond to phase current, must be set above thestarting current (including any dc offset due toasymmetrical transients that may be caused by voltageswitching). This difference is shown in Figure 9-4.

To allow for fault resistance and different types offaults and to assure twice pickup on the unit forminimum fault, the instantaneous phase-relay pickupshould be set at less than one-third of I3ph, where I3ph isthe system contribution, excluding the motor contribu-tion, to a symmetrical three-phase fault on the motorfeeder. Also, pickup should be set at 1.6 times ILR ormore, where ILR is the actual symmetrical startingcurrent, as limited by source impedance. The ratio I3ph/ILR should thus be greater than approximately 5.0.

In general, then, instantaneous-trip units can beused for phase protection if the motor KVA (orapproximately the horsepower) is less than one-halfthe supply transformer KVA. If not, differentialprotection, such as that obtained by the CA or 87M(Fig. 9-4), is required for sensitive fault detection.

The logic for this criterion comes from the follow-ing. Assume a motor is connected to a supplytransformer with 8% impedance. The maximum faultcurrent at the transformer secondary with an infinitesource is

I3ph ¼ 1=0:08

¼ 12:5 per unit on the transformer base

The maximum motor starting current in this case is

ILR ¼ 1

ð0:08þXMÞwhere XM is the motor impedance. In order thatI3ph/ILR> 5, XM must be greater than 0.32 per unit onthe transformer-rated KVA base.

If the motor has a full-voltage starting current of sixtimes full load, then XM¼ 1/6¼ 0.167 on the motor-rated KVA base. With a motor KVA of one-half thetransformer KVA, an XM of 0.167 would be 0.333 onthe transformer base, and greater than 0.32. Clearly,this rule of thumb should only be applied when there isno appreciable deviation from the parameters assumedabove.

3 GROUND-FAULT PROTECTION

A solidly grounded system may be protected by aninverse, very inverse, or short-time induction ormicroprocessor-based relay connected in the currenttransformer residual circuit. For a solid fault at themachine terminals, a typical setting is one-fifth of theminimum fault current. Time dial settings of around 1

Figure 9-4 Comparison of sensitivities of type CA differ-

ential relay and IT instantaneous trip unit.

Motor Protection 147

give operations of four to five cycles at 500% pickupwhen the CO-2 relay is used.

During across-the-line starting of large motors, caremust be taken to prevent the high in-rush current fromoperating the ground relays. Unequal saturation of thecurrent transformers produces a false residual currentin the secondary or relay circuits. Using two- ratherthan three-phase relays or three-phase relays withdifferent impedances will tend to increase the effects offalse residual currents.

False relay operation is unlikely if the phaseburdens are limited so that the voltage developed bythe current transformer during starting is less than 75%of the relaying accuracy voltage rating of the currenttransformer for the particular CT tap being used. Iffalse relay operation is a problem, the ground relayburden of an electromechanical relay should beincreased by using a lower tap. All three transformerswill then be forced to saturate more uniformly,effectively reducing the false residual current. Thisincreased saturation may reduce the sensitivity tolegitimate ground faults and this should be checked.Alternatively, a resistor or reactor can be connected inseries with the ground relay.

The common practice in 2400- to 14,400-V stationservice, and industrial power systems, is to use low-resistance grounding. By using the ‘‘doughnut currenttransformer’’ scheme, such systems offer all theadvantages of instantaneous trip units—speed, relia-bility, simplicity, low cost—without any concern forstarting current, fault contributions by the motor, falseresidual current, or high sensitivity.

Figure 9-5 shows how the BYZ zero-sequence-typecurrent transformer can be used as a supply for the 50instantaneous-trip (IT) unit or 51 time overcurrent(CO) relay. Typical sensitivities obtainable with theseground-fault protection systems are shown inTable 9-1. A voltage is generated in the secondary

winding only when zero sequence current is flowing inthe primary leads. Since virtually all motors have theirneutrals ungrounded, no zero sequence current canflow in the motor leads unless there is a ground faulton the load side of the BYZ. If surge-protectiveequipment is connected at the motor terminals,however, current may be conducted to earth by thisequipment. To date, there has been no reported case ofan instantaneous relay connected to a BYZ currenttransformer tripping because of surge-protectiveequipment. The presence of such equipment may safelybe ignored in choosing a relay.

Solid-state relays such as the type 50D when usedwith the BYZ give good performance in this applica-tion due to a lower burden characteristic. Also,specialized solid-state ground-fault relay systems suchas the Ground-Shield2 series provide a variety ofdoughnut CT window sizes, both toroidal andrectangular. As a system, the relay and CT character-istics are properly prematched by design, and thus neednot be further considered by the user. These specializedsystems usually have relay pickup settings marked interms of primary amperes.

The BYZ current transformer is also used in the fluxbalancing differential scheme, in which each phase is

Figure 9-5 BYZ ground relaying scheme.

Table 9-1 Relay Settings and Sensitivities Using the 50/5 BYZ Zero Sequence Current Transformers

Relay type Relay setting

Minimum sensitivity in primary 310amperes

Maximum primary 310 amperes for

accurate timing and coordination

43=4 IDa 73=4 ID

a 43=4 IDa 73=4 ID

a

IT 0.15 5.0 5.0 — —

CO-8 or 9 0.5 9.0 10.0 25 112

CO-8 or 9 2.5 24.0 24.0 540 1215

CO-11 0.5 6.0 7.0 70 150

CO-11 2.5 24.0 24.0 700 900

a43=4 ID and 73=4 ID are the inside diameter of the window in inches.

148 Chapter 9

equipped as shown in Figure 9-6. This schemecombines excellent phase- and ground-fault sensitivitywith freedom from load current and starting currentproblems.

For high-resistance grounded systems, where veryhigh sensitivity is required, the CWP-1 directionalground relay should be considered. The voltage acrossthe transformer grounding resistor may be used as avoltage polarizing source (Fig. 9-7). The relay has asensitivity of 7mA at 69V.

The maximum torque angle occurs when the currentleads the polarizing voltage by 458. It is interesting tonote that the maximum sensitivity angle is leading thereference voltage � 3V0 by 458. In high-resistancegrounded systems, the predominant impedance forground faults is the zero sequence network containingthe grounding resistor as 3R and the zero sequencedistributed capacitance. The resulting fault currentcalculated is influenced by this RC circuit and thusleads the applied voltage instead of lagging as wouldnormally occur for a ground fault on a low-resistancegrounded system. The CWP-1 directional ground relayis intended for use only on high-resistance groundedsystems.

4 LOCKED-ROTOR PROTECTION

A rotating motor dissipates far more heat than a motorat standstill, since the cooling medium flows moreefficiently. During a failure to start or accelerate afterbeing energized, a motor is subject to extreme heating(approximately 10 to 50 times more than for rated

conditions) in both the stator windings and rotor. Theequivalent circuit during a locked-rotor condition issimilar to a transformer-equivalent circuit with aresistively loaded secondary. The heat distributionbetween the stator and rotor is contingent on therelative stator resistance and 60-Hz rotor resistance.Unlike an overload condition, in which heat can beabsorbed over time by the conductors, core, andstructural members, a locked-rotor condition producessignificant heat in the conductors, which has little timeto be transferred to other sections of the motor.Extreme heating takes place and can be tolerated bythe motor for a very limited time. The time that amotor can remain at standstill after being energizedvaries with the applied voltage and is an I2t limit. Arelay with an I2t characteristic that could be set for anypermissible locked-rotor times and locked-rotor cur-rents would naturally be the best choice for protectingthe motor.

The heat generated within the motor can beapproximated by

I2H ¼ I21 þKI22

where

I1¼ per unit stator positive sequence currentK¼weighting factor to describe the increased

rotor resistance due to skin effect in the rotorbars at (2fs� fslip)

I2¼ per unit stator negative sequence currentfs¼ system frequency

fslip¼ slip frequency

Utilizing both the positive and negative sequencecurrents in an equation relating to IH allows the motorto be protected throughout the full range of statorcurrent with and without unbalance. The time-currentcharacteristic is as shown in Figure 9-8 for the MPRrelay. To insure adequate locked-rotor protection, thecurve position can be set slightly below the full-voltagelocked-rotor time. Depending on the availability ofRTDs, the cutoff can be set to protect the motorduring an overload condition. A typical cutoff settingwithout RTDs would be 115 to 125% of full-loadcurrent.

If the time required for the motor to accelerate theload is significantly less than the permissible locked-motor time, the motor can be effectively protectedusing conventional time-overcurrent relays (Fig. 9-10)or microprocessor relays (Fig. 9-8). If, however, thereis little difference in the two time periods, or theFigure 9-6 Flux balancing differential scheme.


starting time exceeds the locked-rotor time (Fig. 9-9),other considerations must be taken into account.

For the case shown in Figure 9-9, it is tempting totry to fit an overcurrent relay characteristic betweenthe two curves. It should be remembered, however,that the conventional overcurrent relay characteristic isa plot of operating time against sustained current (Fig.9-10), whereas the starting characteristic is a trace ofcurrent against time (Fig. 9-9). If ILR is applied to theCO relay for time ta, the contacts are very nearlyclosed. Current does not drop below the CO pickupvalue until time tb. Contact closure occurs at somepoint tc even though the CO relay characteristic isalways above the current trace.

Over a narrow range, such as that between two andthree times pickup, a CO relay can be assumed tooperate if the integral of (I� 1)n dt exceeds K, where Iis the multiple of pickup and n and K are constantsthat depend on the relay type and time dial setting. If alinear-linear plot of (I� 1)n and t is used, a varyingcurrent and time can be compared with the relaycharacteristic on an area basis. In Figure 9-11, forexample, the CO relay contact will not close if the

current drops below the CO pickup before area Aequals area B.

When both voltage and current are available, analternative solution to locked-rotor problems forlarge motors is to use a distance relay and timer.A motor during start behaves like a three-phasebalanced fault so a distance relay that responds tothree-phase balanced faults should be used. Theimpedance of the motor will remain fixed (largelyreactive at a low power factor) if the motor does notaccelerate. If the motor accelerates, both the impe-dance and power factor will increase (Fig. 9-12). Theimpedance of a motor with a locked rotor isessentially independent of terminal voltage and, asthe motor accelerates, its impedance changes asindicated. This change of impedance with motoracceleration makes the distance relay particularlywell-suited to this application. It also affords backupprotection for phase faults and some ground faults.The timer can be obtained from one of two types ofrelays: a time-overvoltage relay (59) or time-over-current relay (50), each supervised by the 52a breakercontact.

Figure 9-7 Typical connections of the product type CWP-1 (32N) for high-resistance grounded systems.

150 Chapter 9

An alternative to the distance relay technique is thePRO*STAR2 motor protection relay, which deter-mines the speed-dependent heating in the rotor duringa start. The relay uses an impedance measurement to

estimate the speed of the motor, and the rotorthermal model then accounts for the declining rotorresistance during acceleration. Thus a high-inertiamotor where the allowable locked-rotor time is less

Figure 9-8 Typical MPR characteristics.


than the normal starting time can be properlyprotected by this relay.

Some applications use a mechanical zero-speedswitch to supervise an overcurrent unit, preventingoperation of a timer once motor rotation is detected.This scheme will not detect a failure to accelerate tofull speed nor pullout with continued rotation, as theabove schemes will.

Figure 9-9 Motor starting time exceeding permissible

locked rotor time.

Figure 9-10 CO characteristic compared to current trace.

Figure 9-11 Area comparison.

Figure 9-12 KD-10 distance relay (21) used for locked rotor

and backup protection for large motor.

152 Chapter 9

5 OVERLOAD PROTECTION

Heating curves are difficult to obtain and varyconsiderably with motor size and design. Further, thesecurves are an approximate average of an imprecisethermal zone, in which varying degrees of damage orshortened insulation life may occur. It is difficult, then,for any relay design to approximate these variablecurves adequately over the range from light sustainedoverloads to severe locked-rotor overload.

Thermal overload relays offer good protection forlight and medium (long-duration) overloads, but maynot for heavy overloads (Fig. 9-13a). The long-timeinduction overcurrent relay offers good protection forthe heavy overloads, but overprotects for light andmedium overloads (Fig. 9-13b). A combination of thetwo devices provides complete thermal protection (Fig.9-13c).

The National Electric Code requires that an over-load device be used in each phase of a motor ‘‘unlessprotected by other approved means.’’ This requirementis necessary because single phasing (opening onesupply lead) in the primary of a delta-wye transformerthat supplies a motor will produce three-phase motorcurrents in a 2:1:1 relationship. If the two units ofcurrent appeared in a phase with no overload device,the motor would be unprotected. Thus, the NECrequires three overload devices, or two overloaddevices and another to detect unbalance such as aCM or 46D relay.

6 THERMAL RELAYS

There are two types of thermal relays. The DT-3 andtype 49T operate from a resistance temperaturedetector (RTD) that monitors the temperature in themachine windings, motor or load bearings, or loadcase. They are typically applied only to large motors,usually 1500 hp and above where RTDs are available.The RTD is an excellent indicator of average windingtemperature. It is influenced by the effects of ambienttemperature, ventilation variations, and recent loadinghistory.

The second type comprises replica relays such as theBL-1, type 49, and IMPRS, which use an overcurrentelement with very slow reset characteristics to replicatethe thermal condition of the motor and are appliedwhere RTDs are not available.

Multifunction motor protection relays such as thePRO*STAR, MPR, and REM-543 include multipleRTD inputs. When no RTDs are available, these relays

use thermal replica algorithms. When RTDs areavailable, the direct temperature measurement is alsoused, and alarm and trip set points are established foreach of the RTDs.

Figure 9-13 Typical motor and relay time current char-

acteristics.


No current-responsive relay can protect a motorsubjected to blocked ventilation. Relays using RTDinputs for thermal protection overcome this short-coming by responding to temperature alone.

6.1 RTD-Input-Type Relays

Several RTD types are available for use in temperaturemonitoring:

10O copper100O nickel120O nickel100O platinum

Microprocessor-based motor protection relays typi-cally use three-wire input RTDs. The RTD has a well-defined ohmic characteristic vs. temperature. Accuratedetection of the resistance of an RTD requires that thelead resistance be subtracted from the total resistancemeasured by the relay. One scheme used by the MPRcirculates a precision current out the X terminal,through the RTD, and returns to terminal Z via thereturn lead (Fig. 9-14). The following equation results:

VXZ ¼ RLEADICC þRRTDICC þRLEADICC

where

RLEAD¼ resistance of the leads to the RTDICC¼ constant current source

RRTD¼ resistance of the RTD

It next circulates this same current through the Yterminal and return path to Z to obtain

VYZ ¼ RLEADICC þRLEADICC

By subtracting VYZ from VXZ, the relay obtains

VDELTA ¼ VXZ � VYZ ¼ RRTDICC

RRTD ¼ VDELTA=ICC

The error associated with the resistance of the leads isremoved and translations from resistance to tempera-ture can be performed by the microprocessor relayusing setting data stored in nonvolatile RAM orEEPROM that relates to the physical properties ofthe particular material used by the RTD.

The DT-3 is a bridge-type relay. The exploring coilsform part of a Wheatstone bridge circuit, which isbalanced at a given temperature. As the motortemperature increases above the balance temperature,operating torque is produced (Fig. 9-15). With the DT-3 relay, only one resistance temperature detector (10,100, or 120O) or exploring coil is required.

The DT-3 relay is a d’Arsonval-type dc contact-making milliammeter that is connected across thebridge. The bridge is energized by either 125 or250Vdc or supplied with 120Vac through a transfor-mer and full-wave bridge rectifier in the relay. Therelay scale is calibrated from either 50 to 190 8C (or 100to 1608). The right- or left-hand contacts close whenthe temperature rises or falls to the preset valuebetween 50 and 190 8C (or 100 to 160 8C). The normalsetting for class B machines is 120 8C.

6.2 Thermal Replica Relays

Replica-type relays (BL-1 and, additionally, theIMPRS, MPR, and PRO*STAR) are designed toreplicate, within the relay operating unit, the heatingcharacteristics of the machine. Thus, when currentfrom the current transformer secondary passes throughthe relay, its time-overcurrent characteristic approxi-

Figure 9-14 MPR three-wire RTD input.

Figure 9-15 Typical schematic of the type DT-3 relay (49)

for motor overload protection. (Its advantages are good

protection for overload, blocked ventilation, and high

ambient temperature operation.)

154 Chapter 9

mately parallels that of the machine capability curve atmoderate overload.

Extreme variations in load, such as jogging, producea difficult relaying problem. In general, electromecha-nical thermal replica relays cool at a different rate fromthe motor they protect. Variations in load mayproduce a ratcheting effect on the relay and causepremature tripping. Microprocessor motor protectionrelays typically acknowledge previous loading historythrough the use of an RTD input, which establishes astarting level.

The thermal replica relay is recommended whenembedded temperature detectors are not available,although RTD-input-type relays are recommendedwhen they are. Replica-type relays are typicallytemperature-compensated and operate in a fixed timeat a given current, regardless of relay ambientvariations. Although this characteristic is desirablefor the stated conditions, it produces underprotectionfor high motor ambient and overprotection for lowmotor ambient.

7 LOW-VOLTAGE PROTECTION

Low voltage prevents motors from reaching ratedspeed on starting, or causes them to lose speed anddraw heavy overloads. An equation of the averageaccelerating motor torque is directly related to thevoltage present:

TA ¼ E

ER

� �2

Tm � TL

where

ER¼ rated motor voltageE¼ voltage available at motor bus

TL¼ load torqueTm¼ rated voltage motor torqueTA¼ average accelerating torque

It can be seen that the voltage available to the motorsignificantly affects the accelerating torque of themotor.

Motors should be disconnected when severe low-voltage conditions persist for more than a few seconds.Use of ac contractors, which generally release at 50 to70% of rated voltage, provides some low-voltageprotection. However, time-delayed undervoltage pro-tection is preferred, since it delays contactor release onmomentary voltage dips. For switchgear applications,the electromechanical CV (27), CP (27/47), and CVQ

(27/47), or solid-state Types 27/27D and Types 47/47Drelays will accurately detect undervoltage and initiate atrip or alarm as required.

8 PHASE-ROTATION PROTECTION

When starting in reverse can be a serious hazard, areverse-phase relay should be applied. The relay, suchas types CP, CVQ, 47, or 60Q, monitors the busvoltage and is wired to supervise motor starting.Another technique, used by the MPR multifunctionrelay, is to trip instantaneously when reverse-phasecurrents are detected. The voltage relay methodprevents motor energization, whereas the currentmethod requires that the motor be energized andthen tripped when reverse-phase sequence exists.

9 NEGATIVE SEQUENCE VOLTAGEPROTECTION

The CVQ (27/47) relay contains a negative sequencevoltage unit that operates as shown in Figure 9-16. Anegative sequence voltage network, as described inChapter 3, energizes an induction-disk voltage unit V2.If a three-phase voltage applied to the relay contains5% (adjustable to 10%) negative sequence content ormore, the negative sequence unit (V2) operates. A back

Figure 9-16 Simplified schematic diagram of the CVQ (27/

47) negative sequence voltage relay.


contact of the negative sequence unit opens a CV-7undervoltage unit coil circuit, and after a time delay,the contacts of the undervoltage unit initiate trippingor sound an alarm. This relay operates for

Reverse-phase rotation (100% negative sequence)Unbalanced voltage (partial negative sequence)Undervoltage (no negative sequence)

The CVQ relay is recommended for all importantbuses supplying motor loads.

Although the CVQ relay can detect single phasingof the supply to even a single, lightly loaded largemotor if its magnetizing impedance is low enough, itdoes not respond to single phasing between the pointof application of the CVQ and the motor. Figure 9-17displays two cases of an open phasing condition. Thefirst case is an open phase at A. The resulting sequencenetwork interconnections are shown.

In this first case, the negative sequence voltage relaymeasures the voltage across the negative sequenceimpedance of the motor or motors. In the second case,the open phase occurs at B. Figure 9-17 shows thesequence network interconnections. When the open isat location B, the relay now measures the negativesequence voltage across the source.

Very low negative sequence voltage is produced onthe source side of the open phase, which makes itextremely difficult for the negative sequence voltage

relay to detect. For practical purposes, the voltageunbalance occurs only on the load side of an openphase. In general, a phase-unbalance current relay ispreferred for detecting a feeder circuit open phase.

Figure 9-18 illustrates one type of CVQ relay thatresponds only to negative sequence voltage (not under-voltage). The six-cycle timer prevents operation for non-simultaneous pole closure of the supply breaker, 52–1.

The solid-state relay types 47 and 47D are thefunctional equivalents of the CVQ providing reverse-phase unbalanced voltage and undervoltage protec-tion. The type 60Q responds only to negative sequencevoltage, and will operate only for unbalanced condi-tions. The 60Q includes an adjustable time delay.

10 PHASE-UNBALANCE PROTECTION

Phase-unbalance protection is applied to a feedersupplying a large motor or group of small motorswhere there is a possibility of one of the feeder phasesopening as a result of a connector failure, fuse failure,or similar cause. The electromechanical type CMrelay (device 46) contains two induction-disk units(Fig. 9-19). One unit in the CM relay balances IaFigure 9-17 Motor single phasing.

Figure 9-18 External schematic of the CVQ relay used for

tripping on negative sequence voltage only.

156 Chapter 9

against Ib, and the other balances Ib against Ic. Whenthe currents become sufficiently unbalanced, torque isproduced in one or both of the units, closing theircontacts (which are connected in parallel in the tripcircuit). The solid-state type 46D determines thenegative sequence content of the three-phase currentsand includes a built-in timer.

One relay can protect many motors subject tocollective single phasing. In addition, a phase-unbalancerelay may protect up to five motors subject to individualsingle phasing, depending on how the motors areoperated and their relative sizes. For example, the relaywill not operate if a motor with a rating of one-fifth ofthe total feeder load is subject to single phasing whileunloaded and while the remaining motors are fullyloaded. The CM relay has 7-A continuous capabilityand operates when the unbalance exceeds approxi-mately 10 to 15% between 2 and 7 A. With no currentpresent in one current circuit and with 1 A in the other,the relay operates (on the 1-A tap).

Multifunction motor protection relays usuallyinclude an element for phase unbalance protection. Ifthe unbalance exceeds a set threshold for the set timedelay period, tripping is initiated (PRO*STAR,IMPRS, MPR, REM543). In addition, lower levels

of unbalance are used to increase the estimatedtemperature condition of the thermal replica (PRO*STAR, MPR, REM543).

11 NEGATIVE SEQUENCE CURRENT RELAYS

Nostandardshavebeenestablished for the I22t short-timecapability for amotor, although I22t ¼ 40 is regarded as aconservative value. (I2 is the per unit machine negativesequence current, t the time in seconds.)

The electromechanical negative sequence time-over-current relay COQ does not have the sensitivitynecessary to properly protect a motor against theoverheating caused by a prolonged load currentunbalance. The CM and type 46D are the preferredsingle-function relays. The phase-unbalance protectionof the multifunction motor protection relays alsoprovide the required sensitivity. The MPR can be setto pick upwhen I2 ¼ 5 to 30% of the full load tap setting,and the IMPRS and PRO*STAR when I2 ¼ 10 to 50%.Settings in the range 10–20% would be most typical.

12 JAM PROTECTION

A motor can experience excessive torque and over-current in response to a jam condition that can becaused by a binding action of the motor, bearings, ordriven load. To detect a jam condition, the relay has toscreen out other possibilities. The motor contributionto a nearby fault, which can last for a few cycles, canbe screened out by setting the jam time delay greaterthan the motor fault contribution time. High current isnot recognized as a jam condition unless the motor hasbeen through a start and is in a normal running state.Multifunction relays usually include this element, asthey are capable of keeping track of the state of themotor. The advantage of the jam function is muchfaster tripping than would be provided by the lockedrotor protection.

13 LOAD LOSS PROTECTION

The sudden reduction of shaft load is referred to asload loss, which can be caused by a shaft breakage,loss of prime on a pump, or the shearing of a driveFigure 9-19 The type CM phase unbalance relay (46).


pin. To minimize any damage to the driven load, themotor should immediately be taken offline. Detectionof load loss requires the recognition of the differencebetween no load preceding the application of load andno load following the application of load. The loadloss element is commonly supplied in multifunctionmotor protection relays (MPR, PRO*STAR,REM543) that have the ability to track the state ofthe motor.

14 OUT-OF-STEP PROTECTION

Out-of-step protection is applied to synchronousmotors and synchronous condensers to detect pulloutresulting from excessive shaft load or too low supplyvoltage. Other causes of pullout result from a faultoccurring on the supply system, whereby the fault type,clearing time, and location are factors relating to thestability of the motor. Underexcitation caused byincorrect field breaker trip or a short or open in thefield circuit can also result in loss of synchronism of themotor. For a discussion of the out-of-step protectionof large motors, refer to Chapter 8 on generatorprotection.

Small synchronous motors with brush-type excitersare often protected against out-of-step (or loss-of-excitation) operation by ac voltage detection devicesconnected in the field. No ac voltage is present whenthe motor is operating synchronously.

15 LOSS OF EXCITATION

Synchronous motors can be protected against loss ofexcitation by a low-set undercurrent relay connected inthe field. This relay should have a time delay ondropout to trip or alarm the operator. The KLF (orKLF-1) relay (40) (described in Chap. 8) can also beused to protect large motors against loss of field. Theunder-voltage units of these relays should have theircontacts shorted. Loss of excitation of a synchronousmotor does not usually depress the voltage enough tooperate reliably an undervoltage unit.

Unlike undercurrent relays, the KLF (or KLF-1)relay can detect both partial and complete loss of field,and some out-of-step conditions as well (Fig. 9-20).

Both out-of-step and loss-of-excitation conditionscan be detected with a CW watt-type relay (55), Figure 9-21 CW watt relay used for out-of-step detection.

Figure 9-20 KLF used for motor loss of field detection.

158 Chapter 9

connected for 0 torque when the current lags thevoltage by an appropriate power factor angle, suchas 308. Used in this way, the CW is referred to asa power-factor relay. The connection shown inFigure 9-21 gives maximum contact closing torquewhen the current lags its unity power-factor positionby 1208.

16 TYPICAL APPLICATION COMBINATIONS

Table 9-2 and the associated Figures 9-22 and 9-23show typical application combinations for motorprotection. Table 9-3 lists the combined protectionfunctions that are available by using a microprocessor-based motor protection relay.

Table 9-2 Typical Protection for Motors Below 1500 hp (Fig. 9-22)

Device number Quantity Description Typical setting Remarks

49/50 1 BL-1,2 unit with 2 ITs Set at full load for

motor with 1.15

service factor and

90% of full load for

1.00 service factor

motor. IT set 2 times

locked rotor.

Good overload

protection.

51/50 1 CO-5, 1–12 A; with

IIT 10–40 A

Current setting 1/2

locked rotor. Time

delay set to give

operating time>starting time. IT set

2 times locked rotor.

Locked-rotor

protection when

starting time 20 to

70 sec.CO-11, 4–12 A with

IIT 10–40 A Locked-rotor

protection when

starting time

�20 sec.

50G 1 IT, 0.15–0.3 A single

unit

0.15 A. For use with 50/5

BYZ.

47/27 1 CVQ, 5 to 10% V2

sensitivity, 55- to

140-V range

Low voltage 75 to

80%. V2 ¼ 5%.

Undervoltage, phase

sequence, and

unbalanced voltage

protection.

51N/50N (alternative

to 50G where BYZ

cannot be applied)

1 CO-11, 0.5–2.5 A; with

IIT 10–40 A

Pickup 0.5 A, time

0.1 sec at IT setting.

IT set 46IFL.

Provides ground

protection. Time

unit overrides false

residual during

starting.

Figure 9-22 Motor protection below 1500HP.


Figure 9-23 Motor protection 1500HP and above.

160 Chapter 9

Table 9-3 Typical Protection for Motors of �1500 hp (Fig. 9-23)

Device number Quantity Description Typical setting Remarks

49 1 DT-3, 50–190 8C,specify ohms of

RTD 120 vac.

Set for motor max.

safe operating

temperature.

Overload protection:

blocked ventilation

or high ambient.

51 1 CO-5, 1–12 A Current setting 1/2

locked rotor.

Locked-rotor

protection when

starting time is 20 to

70 sec.

CO-11, 4–12 A Time delay setting to

give operating

time> starting time.

Locked-rotor

protection when

starting time

�20 sec.

50 1 SC, 2 unit, 20–80 A Set 2 times locked

rotor.

Fault protection.

50G 1 IT, 0.15–0.3 A single

unit

0.15A. For use with 50/5

BYZ.

46 1 CM, 1–3 A For IFL � 3A: Set 2A. Unbalanced current

protection.

For IFL < 3A: Set 1A.

47/27 1 CVQ, 5 to 10% V2

sensitivity, 55- to

140-V range

Low voltage 75 to

80%, V2 ¼ 5%.

Undervoltage, phase

sequence, and

unbalanced voltage

protection. Note:

CP volt relay can be

used in place of

CVQ if all three-

phase motors on bus

are protected by CM

relays.

87f 1 IT, 0.15–0.3 A, 3 unit 0.15 A. Provides phase and

ground protection.

Use three 50:5 BYZ

transformers. 50G

still required for

cable protection if

BYZ at motor.

87 (alternative to 50

and 87f—use where

minimum 3f fault

current available is

less than 5 times

motor starting

current and 87fcannot be used)

3 CA, 10% None. Phase-fault protection.

51N/50N (alternative

to 50G where BYZ

cannot be applied)

1 CO-11, 0.5–2.5 A, with

IIT 10–40 A

Pickup 0.5 A, time

0.1 sec at IIT setting.

IIT set 4 times full

load.

Provides ground

protection.


10

Transformer and Reactor Protection

Revised By: J. J. McGOWAN

1 INTRODUCTION

Differential relays are the principal form of faultprotection for transformers rated at 10MVA andabove. These relays, however, cannot be as sensitive asthe differential relays used for generator protection.

Transformer differential relays are subject to severalfactors, not ordinarily present for generators or buses,that can cause misoperation:

Different voltage levels, including taps, that result indifferent primary currents in the connectingcircuits.

Possible mismatch of ratios among different currenttransformers. For units with ratio-changing taps,mismatch can also occur on the taps. Currenttransformer performance is different, particularlyat high currents.

308 phase-angle shift introduced by transformerwye-delta connections.

Magnetizing inrush currents, which the differentialrelay sees as internal faults.

Transformer protection is further complicated by avariety of equipment requiring special attention:multiple-winding transformer banks, zig-zag transfor-mers, phase-angle regulators (PAR), voltage regula-tors, transformers in unit systems, and three-phasetransformer banks with single-phase units.

All the above factors can be accommodated by thecombination of relay and current transformer design,along with proper application and connections. Mag-netizing inrush, the most significant variable intransformer protection, will be discussed first.

2 MAGNETIZING INRUSH

When a transformer is first energized, a transientmagnetizing or exciting inrush current may flow. Thisinrush current, which appears as an internal fault tothe differentially connected relays, may reach instan-taneous peaks of 8 to 30 times those for full load.

The factors controlling the duration and magnitudeof the magnetizing inrush are

Size and location of the transformer bankSize of the power systemResistance in the power system from the source to

the transformer bankType of iron used in the transformer core and its

saturation densityPrior history, or residual flux level, of the bankHow the bank is energized

2.1 Initial Inrush

When the excitation of a transformer bank is removed,the magnetizing current goes to 0. The flux, followingthe hysteresis loop, then falls to some residual value fR

(Fig. 10-1). If the transformer were reenergized at theinstant the voltage waveform corresponds to theresidual magnetic density within the core, there wouldbe a smooth continuation of the previous operationwith no magnetic transient (Fig. 10-1). In practice,however, the instant when switching takes place cannotbe controlled and a magnetizing transient is practicallyunavoidable.

163

In Figure 10-2, it is assumed that the circuit isreenergized at the instant the flux would normally be atits negative maximum value ð�fmaxÞ. At this point, theresidual flux would have a positive value. Sincemagnetic flux can neither be created nor destroyedinstantly, the flux wave, instead of starting at itsnormal value ð�fmaxÞ and rising along the dotted line,will start with the residual value (fR) and trace thecurve (ft).

Curve ft is a displaced sinusoid, regardless of themagnetic circuit’s saturation characteristics. Theoreti-cally, the value of fmax is þðjfRj þ 2jfmaxjÞ. Intransformers designed for some normal, economicalsaturation density fs, the crest of ft will producesupersaturation in the magnetic circuit. The result willbe a very large crest value in the magnetizing current(Fig. 10-2).

The residual flux fR is the flux remaining in the coreafter the voltage is removed from the transformerbank. The flux will decrease along the hysteresis loopto a value of fR, where i¼ 0. Because the flux in eachof the three phases is 1208 apart, one phase will have apositive fR and the other two a negative fR, or viceversa. As a result, the residual flux may either add to orsubtract from the total flux, increasing or decreasingthe inrush current.

A typical inrush current wave is shown in Figure10-3. For the first few cycles, the inrush current decaysrapidly. Then, however, the current subsides veryslowly, sometimes taking many seconds if the resis-tance is low.

The time constant of the circuit (L/R) is not, in fact,a constant: L varies as a result of transformersaturation. During the first few cycles, saturation ishigh and L is low. As the losses damp the circuit, thesaturation drops and L increases. According to a 1951

AIEE report, time constants for inrush vary from 10cycles for small units to as much as 1 min for largeunits.

The resistance from the source to the bankdetermines the damping of the current wave. Banksnear a generator will have a longer inrush because theresistance is very low. Likewise, large transformer unitstend to have a long inrush as they represent a large Lrelative to the system resistance. At remote substa-tions, the inrush will not be nearly so severe, since theresistance in the connecting line will quickly damp thecurrent.

In addition to the conditions that influence single-phase inrush, the wave shape of the inrush current intoa delta winding is influenced by the number of coresaffected and the vector sum of the currents from thebank windings. The net wave could, in fact, becomeoscillatory (Fig. 10-4). The shape of a polyphase orsingle-phase inrush to a delta winding is affected by the

Figure 10-1 Magnetizing current when transformers were

reenergized at that instant of the voltage wave corresponding

to the residual magnetic density within the core.

Figure 10-2 Magnetizing current when transformers were

reenergized at the instant when the flux would normally be at

its negative maximum value.

Figure 10-3 A typical magnetizing inrush current wave.

164 Chapter 10

nature of the line current itself, which is the vector sumof two currents from the bank windings. If we assumethat only one core has saturated, the nature of the linecurrent can result in either oscillatory waves ordistortion of the single-phase shape.

When there is more than one delta winding on atransformer bank, the inrush will be influenced by thecoupling between the different voltage windings.Depending on the core construction, three-phasetransformer units may be subject to interphasecoupling that could also affect the inrush current.

Similar wave shapes would be encountered whenenergizing the wye winding of a wye-delta bank or anautotransformer. Here, the single-phase shape wouldbe distorted as a result of the interphase couplingproduced by the delta winding (or tertiary).

Maximum inrush will not, of course, occur on everyenergization. The probability of energizing at the worstcondition is relatively low. Energizing at maximumvoltage will not produce an inrush with no residual. Ina three-phase bank, the inrush in each phase will varyappreciably.

The maximum inrush for a transformer bank can becalculated from the excitation curve if available, andTable 10-1 shows a typical calculation of an inrushcurrent (used phase A voltage as 08 reference).

From these calculated values it can be seen that:

The lower the value of the saturation density fluxfS, the higher the inrush peak value.

The maximum phase-current inrush occurs at the 08closing angle (i.e., 0 voltage).

The maximum line-current inrush occurs at +308closing angles.

Because of the delta connection of transformer wind-ing or current transformers, the maximum line-currentinrush value should be considered when applyingcurrent to the differential relay.

2.2 Recovery Inrush

An inrush can also occur after a fault external to thebank is cleared and the voltage returns to normal (Fig.10-5). Since the transformer is partially energized, therecovery inrush is always less than the initial inrush.

2.3 Sympathetic Inrush

When a bank is paralleled with a second energizedbank, the energized bank can experience a sympatheticinrush. The offset inrush current of the bank beingenergized will find a parallel path in the energizedbank. The dc component may saturate the transformeriron, creating an apparent inrush. The magnitude ofthis inrush depends on the value of the transformerimpedance relative to that of the rest of the system,which forms an additional parallel circuit. Again, thesympathetic inrush will always be less than the initialinrush.

Figure 10-4 Typical magnetizing inrush current wave that

can exist in one of the phases to a delta connection or in the

secondary of delta connected current transformers.

Table 10-1 Typical Inrush Current Calculation

Peak value of inrush current wave (p.u.)

fs Closing angle Ia Ib Ic Ia�Ib Ib�Ic Ic�Ia

1.40 08 5.60 �3.73 �3.73 8.33 �3.73 �8.33

1.40 308 5.10 �1.87 �5.10 5.96 5.10 �9.20

1.15 08 6.53 �4.67 �4.67 10.20 �4.67 �10.20

1.15 308 6.03 �2.80 �6.03 7.83 6.03 �11.06

Figure 10-5 Recovery inrush after an external fault is

cleared.

Transformer and Reactor Protection 165

As shown in Figure 10-6, the total current atbreaker C is the sum of the initial inrush of bank A andthe sympathetic inrush of bank B. Since this waveformlooks like an offset fault current, it could causemisoperation if a common set of harmonic restraintdifferential relays were used for both banks.

Unit-type generator and transformer combinationshave no initial inrush problem because the unit isbrought up to full voltage gradually. Recovery andsympathetic inrush may be a problem, but as indicatedabove, these conditions are less severe than initialinrush.

3 DIFFERENTIAL RELAYING FORTRANSFORMER PROTECTION

Since the differential relays see the inrush current as aninternal fault, some method of distinguishing betweenfault and inrush current is necessary. Such methodsinclude

A differential relay with reduced sensitivity to theinrush wave (such units have a higher pickup forthe offset wave, plus time delay to override thehigh initial peaks), such as types of CA and CA-26 transformer differential relays

A harmonic restraint or a supervisory unit used inconjunction with the differential relay, such astypes of HU, HU-1, HU-4, TPU, and RADSBtransformer differential relays

Desensitization of the differential relay during bankenergization

3.1 Differential Relays for TransformerProtection

3.1.1 CA (87) Transformer Differential Relay

The CA transformer differential and the CA generatordifferential are companion electromechanical relays.Though they have largely been replaced by solid-stateand microprocessor relays, their operating principle isstill of interest. Since there are thousands of theserelays in service, they are included here.

Figure 10-7 shows the basic design of the transfor-mer version of this relay. The generator relay has notaps and is more sensitive than the transformer version.The transformer relay is relatively insensitive to thehigh percentage of harmonics contained in magnetiz-ing inrush current. Because of this and its relativelyslow six-cycle operating time, it has been usedsuccessfully for many decades in less critical applica-tions where cost has a significant influence.

Operation occurs when the operating current (whichis the differential current) exceeds roughly 50% of theminimum restraint. Putting it in another frame ofreference, it operates when the operating currentexceeds approximately 20% of the summationrestraint. For external faults, note that the restraintampere-turns are additive; and for an internal faultthey are subtractive. For the idealized case of equalcontributions from sources on each side of thetransformer to an internal fault, the restraint cancels

Figure 10-6 Sympathetic inrush when a bank is paralleled

with a second energized bank.

Figure 10-7 Type CA transformer differential relay.

166 Chapter 10

completely, and the currents add together in theoperating coil. The transformer differential version ofthis relay allows currents with as much as a 2:1 ratio tobe matched. Figure 10-8 describes the case with 10Ainput from one set of current transformers and 5A tothe other set. Balance is accomplished in the operatingcoil by the autotransformer action. For this case, and

for all such ‘‘through’’ phenomenon that provide thissame ratio of currents, cancellation of operating torqueoccurs. Figure 10-9 shows a typical distribution ofcurrents for a ‘‘through’’ condition with the relay set tobalance currents with a 5 to 8 relationship.

3.1.2 CA-26 (87) Transformer Differential Relay

The CA-26 is a similar design to the CA-16 busdifferential relay, but with the ability to accommodateall of the necessary input currents for protecting athree-winding transformer. With no taps, the relayrequires auxiliary current transformers in two of theinput circuits to produce a match. One relay per phase,of course, is required for a three-phase transformer.Figure 10-10 describes the mechanical configurationused with this relay and the principle by whichrestraint is produced. It is sensitive and reasonablyfast, but its lack of harmonic restraint leads one to theselection of more modern relays.

3.1.3 HU and HU-1 Transformer DifferentialRelays

Since magnetizing inrush current has a high harmoniccontent, particularly the second harmonic, this secondharmonic can be used to restrain and thus desensitize arelay during energization. The method of harmonicrestraint is not without its problems. There must beenough restraint to avoid relay operation on inrush,without making the relay insensitive to internal faultsthat may also have some harmonic content.

The HU (two restraining winding) and HU-1 (threerestraining winding) variable-percentage differentialrelays have second-harmonic restraint supervisionthat adequately solves these problems. The connec-tions for these relays are shown schematically in

Figure 10-8 Distribution of currents in the type CA relay

set on the 5–10 taps.

Figure 10-9 Distribution of currents in the type CA relay

set on the 5–8 tap for example of Figure 10-14.

Figure 10-10 Type CA-26 transformer differential relay.


Figure 10-11. In the differential unit (DU), air-gaptransformers feed the restraint circuits, and a non-air-gap transformer energizes the operating coil circuit.Since the rectified restraint outputs are connected in

parallel, the relay restraint is proportional to themaximum restraining current in any restraint circuit.

The percentage characteristic varies from around20% on light faults, where current transformer

Figure 10-11 Schematic connections of the HU and HU-1 variable percentage differential relays with second harmonic

restraint supervision.

168 Chapter 10

performance is good, to approximately 60% on heavyfault, where current transformer saturation may occur.This variable-percentage characteristic is obtained viathe saturating transformer in the operating coil circuit.Taps provide a 3:1 difference in the ratio of the maincurrent transformer outputs. These taps are 2.9, 3.2,3.5, 3.8, 4.2, 4.6, 5.0, and 8.7.

The minimum pickup current is 30% of the tapvalue for the 30% sensitivity relay and 35% of tap valuefor the 35% sensitivity relay. The minimum pickup isthe current that will just close the differential unitcontacts, with the operating coil and one restraint coilenergized. The continuous rating of the relay is 10 to22A, depending on the relay tap used.

The harmonic restraint unit (HRU) has a second-harmonic blocking filter in the operating coil circuitand a second-harmonic pass filter in the restraintcoil circuit. Thus, the predominant second-harmoniccharacteristic of an inrush current produces amplerestraint with minimum operating energy. Thecircuit is designed to hold open its contacts whenthe second-harmonic component is higher than 15%of the fundamental. This degree of restraint in theHRU is adequate to prevent relay operation onpractically all inrushes, even if the differential unitshould operate.

For internal faults, ample operating energy is pro-duced by the fundamental frequency and harmonicsother than the second. The second harmonic is at aminimum during a fault. Since the HRU will operate atthe same pickup as the DU, the differential unit willoperate sensitively on internal faults, as shown in thetrip circuit of Figure 10-11. For external faults, thedifferential unit will restrain.

The relay operating time is one cycle at 20 times tapvalue. The instantaneous-trip unit (IIT) is included toensure high-speed operation on heavy internal faults,where current transformer saturation may delay HRUcontact closing. The IIT pickup is 10 times the relay tapvalue. This setting will override the inrush peaks andmaximum false differential current on external faults.

3.1.4 HU-4 Transformer Differential Relay

The HU-4 relay is used to protect multiple-windingtransformer banks or in the protective zone thatincludes the bus. The HU-4 relay is similar to theHU and HU-1 relays, but has four restraint windings.Also, the rectified outputs of the restraint transformersare connected in series, and the IIT unit is set at 15times the tap value. The application of this relay isdescribed in Chapter 11.

3.1.5 Modified HU Relays

In the HU relay, the 15% second-harmonic restraintvalue is based on a minimum second-harmonic content15% of fundamental, under an inrush condition offS ¼ 1:40 p:u: and a 08 closing angle. In moderntransformers, however, saturation density is moreoften 1.20 to 1.30 p.u. and can even be as low as1.0 p.u. At these lower saturation densities, theminimum second-harmonic content of the fundamen-tal is significantly lower. Also, service conditions aremore severe when the closing angle is +30�, ratherthan 08. (See Table 10-1.)

As a result, the second-harmonic content percentagemay be as low as 7% of the fundamental, and thepercentage of all harmonics may be as low as 7.5% ofthe fundamental. Under these conditions, the 15%second-harmonic restraint relays may not functionproperly during the energization of power trans-formers with low saturation density values.

The modified HU relay is designed to solve thisproblem. This relay is similar to the 15% second-harmonic HU relay, except that a 33-O, 3-W resistor isconnected across the HRU operating coil to calibratethe unit for a 7.5% second-harmonic restraint. Thecharacteristics of the modified HU are the same as forthe HU relay, and the modification does not affect thecharacteristics of the IIT unit. The differential unit ofthe modified relay, however, is about one cycle slowerthan the unmodified unit.

Ths modified HU relay has been used successfully ininrush current tests at a 138-kV generating station andin several installations where the 15% HU relaysexperience inrush problems.

3.1.6 Type RADSB Transformer DifferentialRelay

The type RADSB transformer differential relay is asolid-state three-phase package. Its basic versionprovides two restraining circuits for two-windingpower transformer protection. It can be expanded upto six restraints initially or in the future.

As shown in Figure 10-12, the relay utilizes thesecond harmonic for inrush current restraint. The relaywill restrain if the second harmonic content in any onephase is greater than 17% of its fundamental. Thisfeature is very unique in three-phase package design. Itwill solve the inrush current problem mentioned inSection 3.1.5, ‘‘Modified HU Relays.’’

The relay utilizes the fifth harmonic for the over-excitation restraint. The relay restrains if the fifth-harmonic content is greater than 38% of its funda-


Figu

re10-12

Block

diagram

ofRADSBtransform

erdifferentialrelay.

170 Chapter 10

mental. However, refer to Section 5.5, ‘‘OverexcitationProtection of a Generator-Transformer Unit,’’ in thischapter. It is necessary to apply the fifth harmonic withcaution, or a V/Hz relay, either type MVH orRATUB, should be considered for supervision.

The relay does not provide built-in taps; therefore, itrequires external auxiliary ct’s for current matching inall applications. Instantaneous high set trip function isalso built in the relay.

For applications that required more than tworestraining circuits, the RTQTB-061 unit(s) can beadded. Each RTQTB-061 unit contains six transfor-mers for two additional restraining circuits, as shownin Figure 10-12.

3.1.7 TPU 2000R Transformer Protection System

This unit takes full advantage of microprocessortechnology. It allows the selection of a wide varietyof characteristics such as second-harmonic restraint,all-harmonic restraint, and fifth-harmonic restraint.Inputs to it may be from wye- or delta-connectedcurrent transformers, irrespective of the transformerwinding connections. This allows monitoring of theindividual input phase currents rather than a combina-tion of them, as a delta-connected set of ct’s wouldprovide. The relay algorithm assures a proper match ofthe input currents from both (or all three) sides of thetransformer. It is compatible (different styles) with atwo- or three-winding transformer.

Any conceivable overcurrent unit curve shape canbe chosen as a setting for the overcurrent functions ofthis relay. Similarly, any of the communicationprotocols that are in popular use can be provided asan inherent part of this device. Operating curve slopeand minimum trip level are also obtained by setting.Inrush monitoring is possible, with blocking oftripping being selectable by second harmonic, fifthharmonic, or all harmonics (through the eleventh).Historically, each concept has been used successfully.

Though all of these methods of recognizing inrushas a distinctive phenomenon for which trip blocking ismandatory are useful, each has its own favorable andunfavorable nuances. Second-harmonic blocking isminimally less secure for those cases involving over-voltage. The higher the voltage upon energization, thelower the percentage of second harmonic. In general,the relay is not capable of tripping at an overvoltagelevel below which the typical transformer can supporton a prolonged basis. The relay expresses appropri-ately the need to trip. However, it would trip far toosoon.

Fifth harmonic blocking is favored for those over-voltage cases (long EHV line energization, hydroma-chine load rejection, etc.) where undesired high-speedtripping on elevated magnetizing current may occur. Itshould be recognized that while transformers cansupport short-time overvoltage, they are vulnerableto prolonged overvoltage heating. If differential trip-ping is to be blocked when inordinate fifth harmonic isobserved on overvoltage, the block must be released(or tripping imposed by other means) prior to theoccurrence of damage to the transformer.

All-harmonic blocking is moderately less depend-able because of the increased harmonics that arepresent in arcing faults.

Cross-blocking, the feature that blocks all differen-tial tripping when the harmonic restraint setting andthe operating current are exceeded in any one or morephases, increases security against an unwarrantedoperation during inrush. This feature is partiallyinherent where ct’s are traditionally connected in deltafor balancing the transformer phase shift and is usefulwith the harmonic-restraint mode of second or secondand fifth. For an internal ground fault, possibleelevated voltage on an unfaulted phase should beinvestigated to assure that excessive fifth harmonic inthat phase cannot block tripping of the faulted phasedifferential unit. Where this can occur, cross-blockingmust not be used.

3.2 General Guidelines for TransformerDifferential Relaying Application

The following guidelines are designed to assist inselecting and applying relays for transformer protec-tion. When two or more relays appear to be equallysuitable, engineering experience and economics willdetermine the final choice.

1. There is no clearcut answer to the question ofwhich relay or protective method to apply. As ageneral rule, however, the induction-disk differentialrelays (CA and CA-26) are used at substations remotefrom large generating sources where inrush is not aproblem and the kVA size of the bank is relativelysmall. The more complex and more expensive harmo-nic relays (HU, HU-1, HU-4, TPU and RADSB) areused at generating stations and for large transformerunits located close to generating sources, where asevere inrush is highly likely.

2. A current transformer tap that will giveapproximately 5A at maximum load is recommendedfor use with multiratio current transformers. This


arrangement provides good sensitivity without intro-ducing thermal problems in the current transformer,leads, or relay itself. Sensitivity can be improved byusing a tap that gives more than 5A; however, thecurrent transformer, leads, and relay capability mustbe checked carefully to guard against thermal over-load.

3. In general, for all except the TPU, the currenttransformers on the wye side of a wye-delta bank mustbe connected in delta, and the current transformers onthe delta side connected in wye. This arrangement (1)compensates for the 308 phase-angle shift introducedby the wye-delta bank and (2) blocks the zero sequencecurrent from the differential circuit on external groundfaults. As shown in Figure 10-13, zero sequence currentwill flow in the differential circuit for external groundfaults on the wye side of a grounded wye-delta bank; ifthe current transformers were connected in wye, therelays would misoperate. With the current transfor-mers connected in delta, the zero sequence currentcirculates inside the current transformers, preventingrelay misoperation.

4. Relays should be connected to receive ‘‘in’’ and‘‘out’’ currents that are in phase for a balanced loadcondition unless the relay itself is designed or set toaccommodate the difference. When there are morethan two windings, all combinations must be consid-ered, two at a time.

5. Relay taps or auxiliary current transformerratios should be as close as possible to the currentratios for a balanced maximum load condition. Whenthere are more than two windings, all combinationsmust be considered, two at a time, and based on thesame kVA capacity.

6. Ground only one point in the differentialscheme; never do multiple-point grounding.

7. After the current transformer ratios and relaytaps have been selected, the continuous rating of relaywindings should be checked for compatibility with thetransformer load. If the relay current exceeds itscontinuous rating, a higher-current transformer ratioor relay tap may be required. The relay’s requiredcontinuous rating may be determined from themaximum kVA capacity of the transformer bank. Ifthe transformer is allowed to exceed its maximum kVAcapacity for a short time, the expected 2-h maximumload should be used. The relay will reach finaltemperature within 2 h.

8. The percentage of current mismatch shouldalways be checked to ensure that the relay taps selectedhave an adequate safety margin. When necessary,current mismatch values can be reduced by changingcurrent transformer taps or adding auxiliary currenttransformers. Percentage mismatch M can be deter-mined from Eq. (10-1):

M ¼ILIH� TL

TH

S

��6100% ð10-1Þ

where

IL, IH¼ relay input currents, at the same kVA base,for low- and high-voltage sides, respectively

TL, TH¼ relay tap settings for low- and high-voltage sides, respectively

S¼ smaller of the two terms, (IL/IH) or (TL/TH)

When there are more than two windings, all combina-tions should be calculated, two at a time. When taps arechanged under load, the relays should be set on the basisof the middle or neutral tap position. The total mis-match, including the automatic tap change, should notexceed the recommended values shown in Table 10-2.

For example, for a transformer bank with a +10%on-load tap changer device, the calculated mismatchvalue should not be greater than +5% for a 30% HUrelay application. However, if the transformer bankdoes not have an on-load tap changer, then thecalculated mismatch value can be tolerated up to the‘‘limit of (MþLTC)’’ value.

Figure 10-13 Reason for delta-connected ct’s on wye

windings.

172 Chapter 10

9. To ensure correct operation of the relayingscheme, the current transformer performance shouldbe checked. A less accurate, but still acceptable,method is to use the ANSI relaying accuracy classifica-tion. For units with class C accuracy, performance willbe adequate if

NpVCL � ðIext � 100ÞRS

Iext> ZT ð10-2Þ

where

Np¼ proportion of total current transformer turnsin use, for example, if 1000/5 tap is used for a2000/5 MR current transformer, thenNp¼ 0.5

VCL¼ current transformer accuracy, class C vol-tage; for example, 200 for a class C200current transformer

Iext¼maximum external fault current in secondaryrms A (let Iext¼ 100 if maximum externalfault current is less than 100A)

Rs¼ current transformer secondary winding resis-tance in ohms

ZT¼ total current transformer secondary circuitburden impedance in ohms, determined byequation:

ZT ¼ 1:13ðm6RLÞ þ relay burdenþ ZA ð10-3Þ

where

RL¼ one-way lead resistance1.13¼multiplier used to accommodate temperature

rise of the conductors during faultsZA¼ burden impedance of any devices (other than

the relay) connected or reflected to thecurrent transformer secondary circuit

m¼multiplier, depending on the current trans-former connection and type of fault to beconsidered, as shown in Table 10-3

4 SAMPLE CHECKS FOR APPLYINGTRANSFORMER DIFFERENTIAL RELAYS

The following examples show the importance of thecurrent transformer connections, current ratios, relayratings, and current transformer performance inapplying the differential relay scheme for transformerprotection.

4.1 Checks for Two-Winding Banks

A worksheet for connecting differential relays arounda two-winding bank is shown in Figure 10-14. Thisdoes not apply to the TPU relay.

4.1.1 Phasing Check

It is very important to note that the transformer bank,as shown in Figure 10-14, is connected so that the highside lags the low side by 308, which is not an ANSI-standard-connected bank. (In a standard connection,the high side leads the low side by 308.) In thisexample, the nonstandard connection is used forillustration only. Practically, the actual connection ofthe bank should be confirmed with the informationfrom its nameplate before proceeding to the next stepof the phasing check.

Procedures for phasing check can be simplified asbelow (refer to Figures 10-14 and 10-15).

Step 1 Assume that Ia, Ib, and Ic on the wye sideflow through the bank to an external three-phasefault or maximum load.

Step 2 See Figure 10-14; use the statement on page12 to define the current in the windings: ‘‘thecurrent flowing out at the polarity-markedterminal on the secondary side is substantiallyin phase with the current flowing in at thepolarity-marked terminal on the primary side.’’

Three-phase transformers do not carry polar-ity marks as do their single-phase counterparts.However, from the nameplate data for thetransformer, it can clearly be seen which of thelow-voltage windings is drawn in parallel with

Table 10-2 Recommended Mismatch (M) Limitation

Relay

Sensitivity

(%)

Limit of

(MþLTC)

(%)

CA 50 35

HU, HU-1, HU-4, TPU 30 15

HU, HU-1, HU-4, TPU 35 20

CA-26, RADSB — 10

Table 10-3 Multiplier (m) for Eq. (10-3)

3fF fGF

Wye-connected ct m¼ 1 m¼ 2

Delta-connected ct m¼ 3 m¼ 2


one of the high-voltage windings. This is asymbolic identification that these two windingsare on the same core-leg. If these windings(actually lines) are vertical on the nameplatediagram, it may be assumed that the polaritymarkings could be placed at the upper end ofeach of these windings. Similar identification canbe made of the other windings. The terminals of athree-phase transformer are identified as H1, H2,H3, and, if there is a neutral, H0 (note this saysterminals, not windings). The low-voltage term-inals are called X1, X2, and X3 (and possiblyX0). If there is a third set of windings and theterminals are brought out of the case, they will beidentified as Y1, Y2, and Y3 (and again Y0 ifthere is a neutral and it is brought out of thecase).

The nameplate will show whether the voltagedrop from H1 to H3 is in-phase with X1 to X0(high leads low by 308), or H1 to H2 is in-phaseX1 to X0 (high lags low by 308) or some othercombination. It is assumed that the system‘‘A-phase’’ on the high-voltage side will beconnected to the H1 terminal and that the

‘‘A-phase’’ on the low-voltage side will beconnected to the X1 terminal, but surprisinglythis is seldom the case. This simply adds to thecomplexity of the relay engineer’s task, sortingout all of these factors. In fact, the transformerconnection dictates the relationship between thehigh- and low-voltage systems, and the phasorscan be named anything with which the user feelscomfortable.

Step 3 Trace these currents through the delta tothe delta-side phase wires, then through the wye-connected current transformers to the relays.

Step 4 Repeat the above relay currents to theother-side restraint windings.

Step 5 Trace the currents on the wye-windingphase wires; then determine the secondarycurrents (phase and direction) on each wye-sidecurrent transformer.

Step 6 Match up the information from steps 4 and5, enabling the wye-side current transformers tobe properly connected in delta for correctphasing under all conditions. The completedcheck for the example of Figure 10-14 is shownin Figure 10-15.

Figure 10-14 Worksheet for connecting differential relays around a two-winding transformer bank.

174 Chapter 10

4.1.2 Ratio Check

For the example in Figure 10-14, the ratio checkshould be executed as shown in Table 10-4. The stepsthat follow depend on the type relay being applied: CA(steps 7 to 10); CA-26 and RADSB (steps 11 to 13); orHU (steps 14 to 17):

4.1.3 For Type CA Relays (Steps 7 to 10)

For the application of CA relays, the tap settings,continuous coil ratings, and mismatch must also bechecked as described in steps 6 to 9. The taps in CArelays are 5-5, 5-5.5, 5-6.6, 5-7.3, 5-8, 5-9, and 5-10.Tap ratios are 1.00, 1.10, 1.32, 1.46, 1.60, 1.80, and2.00, respectively.

Step 7 Select relay taps with a ratio as close aspossible to the relay current ratio in step 5. In thiscase, tap 5-8, with a ratio of 1.60, is the closest.

Step 8 Connect relay terminal 9 to the 69-kV sideand terminal 7 to the 11.5-kV side. Always set thetime dial at position number 1.

Step 9 Check the continuous rating of the relaycoils. As shown in calculations, the continuouscurrents flowing in the restraint coils are less than10A, and any through currents flowing in theoperating coil are less than 5A. Therefore, therelay windings will not be subject to a thermalproblem.

Step 10 Calculate mismatch; use Eq. (10-1). Thismismatch is well within the 35% mismatch limitof Table 10-2.

4.1.4 For Type CA-26 and RADSB Relays (Steps11 to 13)

To calculate mismatch for CA-26 relays, perform steps11 to 13 as follows:

Figure 10-15 Complete phasing check for the example of Figure 10-14.


Step 11 Calculate mismatch; use Eq. (10-1).

M ¼6:524:18 � 1

S6100%

¼ 1:560� 1

16100% ¼ 56%

Step 12 Since the percent mismatch is higher thanthe recommended value (Table 10-2), an auxili-ary current transformer or a current-balancingautotransformer is required to decrease thecurrent to the relay. (Note: Use auxiliary currenttransformer to decrease the current to the relay.)The turns ratio of the balancing current trans-former is (4.18/6.52)6 100%, or 64.1%. Either atwo-winding auxiliary current transformer with aturns ratio of 3/2 (for example, ABB style no.7881A026G06), or a current-balancing auto-transformer (for example, ABB type A auxiliarycurrent transformers) would be satisfactory.The connections for these two transformers areshown in Figure 10-16. The continuous rating ofthe auxiliary current transformers should bechecked to guard against possible thermalproblems.

Step 13 After selecting the auxiliary current trans-former ratio, the mismatch should be rechecked.

If a 3/2 ratio auxiliary current transformer is used,refer to Figure 10-16 and apply Eq. (10-1):

M ¼4:344:18 � 1

16100% ¼ 3:80%

If a current-balancing autotransformer is used,refer to Figure 10-16 and apply Eq. (10-1):

M ¼4:174:18 � 1

16100% ¼ 0:24%

Both values are well within the 10% mismatchlimit of Table 10-2.

4.1.5 For Type HU Relays (Steps 14 to 17)

To apply HU relays, the tap settings, continuous coilratings, and mismatch must be checked, as described insteps 13 to 16. The taps in HU relays are 2.9, 3.2, 3.5,3.8, 4.2, 4.6, 5.0, and 8.7. The tap ratios are given inTable 10-5.

Step 14 Select relay taps that have a ratio as closeas possible to the relay current ratio in step 5. Inthis case, tap 5.0/3.2, with a ratio of 1.563, is theclosest.

Step 15 Use tap 3.2 on the 69-kV side and tap 5.0on the 11.5-kV side. In most HU applications,auxiliary current transformers are not requiredfor current balancing if the current ratio isbetween 1 and 3.

Step 16 Check the continuous thermal rating ofthe relay coils. Since the continuous rating of therelay is 12A for the 3.2 tap, there should be nothermal problem with the relay coils.

Table 10-4 Example of Ratio Check for Two-Winding Transformer

Step LV (wye) HV (delta)

1. For the example shown in Figure 10-14, assume that the maximum load

is 30,000 kVA. Then the rating of the bank IFL is

30;000ffiffi3

p611:5

¼ 1506A30;000ffiffi3

p669:0

¼ 251A

2. For increased sensitivity, select current transformer ratios as close to the

IFL value as possible. Practically, a calculated value of

(IFL/0.8) can be used as the reference for determining the current

transformer ratios for this example.

15060:8 ¼ 1882:5

2510:8 ¼ 313:7

Then use n ¼ 20005

n ¼ 3005

3. Calculate current transformer secondary currents IS¼ (IFL/n) ¼ 400 ¼ 60

¼ 1506400

¼ 25160

¼ 3.77A ¼ 4.18A

4. Calculate relay current IRL ¼ 3:77ð ffiffiffi3

p Þ IRH¼ 4.18A

¼ 6.52A

5. Calculate relay current ratioIRL

IRH¼ 6:52

4:18 ¼ 1:560

176 Chapter 10

Step 17 Calculate mismatch:

M ¼6:524:18 � 5:0

3:2

S6100%

¼ 1:560� 1:563

1:560¼ 100% ¼ 0:2%

This value is well within the 15% mismatch limitof Table 10-2. Also, the 30% sensitivity of theHU relay would be satisfactory for this applica-tion.

4.1.6 For Type TPU Relays

The TPU is applied similarly to the HU relays withadded benefits. Its wider tap range and finer tapincrements yield an even smaller mismatch than thatachieved with the HU relay. Also, the ct’s can beconnected in wye on both sides of the transformer,with compensation for transformer phase shift accom-plished internally for phase current metering andovercurrent protection on the wye side of thetransformer. For the example in Table 10-4, the tapcalculations and selections would be the same as thatfor the HU relay. The

ffiffiffi3

pfactor for IRL would still

apply because of the TPU internal compensation andnot because of delta-connected ct’s.

4.1.7 Current Transformer Performance Check

Assume that the three-phase external fault currents arehigher than the single-phase fault currents, with valuesof 15,000A on the 11.5-kV side and 2500A on the 69-kV side. In this case, the current transformer burdenlimit can be calculated as follows:

Low

voltage

High

voltage

Maximum external fault

current (primary

amperes), Ip

15,000 2500

Current transformer turns

ratio, n

400 60

Secondary amperesIpn

37.5 41.7

NP20003000

¼ 0:67300600

¼ 0:50

From Eq. (10-2) where (Iext� 100)RS¼ 0 since Iext isless than 100A secondary (Figure 10-14 shows the

Figure 10-16 Current balancing transformer connections

for Figure 10-15 when type CA-26 or RADSB relay is used.

Table 10-5 HU Relay Tap Ratios

2.9 3.2 3.5 3.8 4.2 4.6 5.0 8.7

2.9 1.000 1.103 1.207 1.310 1.448 1.586 1.724 3.000

3.2 1.000 1.094 1.188 1.313 1.438 1.563 2.719

3.5 1.000 1.086 1.200 1.314 1.429 2.486

3.8 1.000 1.105 1.211 1.316 2.289

4.2 1.000 1.095 1.190 2.071

4.6 1.000 1.087 1.890

5.0 1.000 1.740

8.7 1.000


C400 current transformers are used in this example),

0:676400

100¼ 2:67O

0:56400

100¼ 2:0O

Current transformer performance will be satisfactory ifthe total burden impedance values, as calculated fromEq. (10-3), are less than the above values.

4.2 Checks for Multiwinding Banks

The same types of phasing, ratio, continuous rating,and current transformer performance checks are usedfor multiwinding transformer as for two-windingtransformers. To determine the correct direction andphase of the restraint currents, one side of thetransformer is considered the primary and the otherwindings the secondaries. For ratio checks, any twowindings must be checked based on a same kVA value,as if the bank were a two-winding unit with no current

in the other winding. Any other pair is then checked inthe same manner. This process ensures that all ratiosare correct for any distribution of fault or load current.

A worksheet showing the connection of differentialrelays around a typical three-winding bank is given inFigure 10-17. The completed phasing checks are shownin Figure 10-18.

If the ratios are not correct for the relay, auxiliarycurrent-balancing autotransformers or current transfor-mers are required. In general, one or two sets are requiredfor the three-windingbank, dependingon theunbalancedcondition and relay type (CA-26, HU-1, or RADSB).

For the three-winding bank shown in Figure 10-18,the ratio check is performed per Table 10-6, steps 1 to 5.

4.2.1 For Type CA-26 and RADSB Relays (Steps6 to 8)

For CA-26 relay application, mismatch is checkedaccording to steps 6 to 8 below.

Figure 10-17 Worksheet for connecting differential relays around a three-winding transformer bank.

178 Chapter 10

Step 6 Calculate mismatch; use Eq. (10-1):

MHM ¼7:5776:012 � 1

16100% ¼ 26%

MML ¼6:0124:380 � 1

16100% ¼ 37:3%

MHL ¼7:5774:380 � 1

16100% ¼ 73%

Step 7 To decrease the currents to the relay, addcurrent-balancing autotransformers at the 66-and 26-kV sides. The turn ratios of thesetransformers are

4:380

7:5776100% ¼ 57:8%

Use 73 turns ratio:

4:380

6:0126100% ¼ 72:8%

Use 58 turns ratio:

Then the currents to the relay are

IRH ¼ 7:577658% ¼ 4:395A

IRM ¼ 6:012673% ¼ 4:389A

IRL ¼ 4:380A

Step 8 Recalculate the mismatch:

MHM ¼4:3954:389 � 1

16100% ¼ 0:1%

MML ¼43894:380 � 1

16100% ¼ 0:2%

MHL ¼4:3954:380 � 1

16100% ¼ 4:4%

Figure 10-18 Complete phasing check for the example of

Figure 10-17. The dotted lines show the connections for a

phasing check between the 66 and 11KV windings. Assum-

ing the 26KV circuit does not exist. The dashed lines show

the connections for a phasing check between the previous

connections made for the 66KV winding, assuming that the

11KV circuits do not exist. With this method, phasing is

correct for any distribution of currents through the three

windings.

Table 10-6 Example on Ratio Check for Three-Winding Transformer

Step 66 kV (Y) 26 kV (Y) 11 kV (D)

1. If 25,000 kVA flows through

the bank, the currents in

each winding are IFL

25;000ffiffi3

p666

¼ 21925;000ffiffi3

p626

¼ 55625;000ffiffi3

p611

¼ 1314

2. If we assume current

transformer turn ratios of

n¼ 250/5 ¼ 800/5 ¼ 1500/5

¼ 50 ¼ 160 ¼ 300

3. Then the current

transformer secondary

currents are IFL/n

¼ 219/50 ¼ 556/160 ¼ 1314/300

¼ 4.380 ¼ 3.475 ¼ 4.380

4. Relay currents are IRH ¼ ffiffiffi3

p64:380 ¼ 7:577 IRM ¼ ffiffiffi

3p

63:475 ¼ 6:012 IRL ¼ 4:380

5. Relay current ratios areIRH

IRH¼ 7:577

6:012 ¼ 1:260IRM

IRL¼ 6:012

4:380 ¼ 1:373IRH

IRL¼ 7:577

4:380 ¼ 1:730


4.2.2 For Type HU-1 Relays (Steps 9 to 12)

For HU-1 relay application, tap settings, mismatch,and current transformer performance are calculated asfollows:

Step 9 To select the relay taps, use Table 10-5 andstart from the highest current ratio, IRH/IRL¼ 1.730. The nearest tap ratio is 5/2.9¼ 1.724. Select TH¼ 5 and TL¼ 2.9. Next,select the tap ratio for the second higher-currentratio, IRM/IRL¼ 1.373, by using the lower tapfrom the first tap ratio (TL¼ 2.9) as a reference.This ratio would be 3.8/2.9¼ 1.310. In otherwords, TM¼ 3.8.

Step 10 Calculate the mismatch:

MHM ¼7:5776:012 � 5

3:8

1:266100% ¼ 4:4%

MML ¼6:0124:380 � 3:8

2:9

1:316100% ¼ 4:8%

MHL ¼7:5774:380 � 5

2:9

1:726100% ¼ 0:3%

Step 11 Check current transformer performance inthe same way as for the two-winding bank above.

Step 12 Even though the current ratios are within3.0 (step 5), auxiliary current transformers maybe required for current balancing. For example, ifIRH¼ 3.75, IRM¼ 8.109, and IRL¼ 6.222, thecurrent ratios are 2.162, 1.303, and 1.659. Eventhough these values are within 3, auxiliarycurrent transformers are necessary in this appli-cation.

4.2.3 For Type TPU Relays

Again, the TPU would be applied similarly to the HUrelay. In the three-winding TPU, all ct’s must beconnected in wye, resulting in a

ffiffiffi3

pfactor being

applied to IRH and IRM due to internal compensationas they were applied with the HU due to the ct’s beingconnected in delta.

4.3 Modern Microprocessor Relay

The modern microprocessor relay, such as the TPUrelay, is a protection system, rather than a simpletransformer differential overcurrent relay. Settings arerequired for phase and ground (time and instanta-neous) as well as negative sequence overcurrentfunctions. For the differential function, the tap rangesare wider and the steps are smaller than those

employed in electromechanical or solid-state relays.Differential comparisons are made between the I/tapvalues of all currents entering and leaving the relay.The taps are chosen to be proportional to the (actualor adjusted) input currents to the relay for a ‘‘through’’fault (or load). Contrary to the traditional method ofselecting the current transformer connections to delivercurrents to the relay with appropriate phase relation-ships, with this relay flexibility exists in those connec-tions. Factors are applied in setting the relay thatcompensate for the connections of the transformer andthe ct’s on the high and low (and third winding wherepresent) sides of the protected transformer. Theselection is based on the angle by which the winding-1, typically high voltage, input current leads thecomparable winding-2, typically low voltage (andwinding-3, typically tertiary voltage, where applicable)outgoing current.

The differential slope (IOP/IR) of the relay is linear(15 to 60%) with a minimum operating current. Bothslope and minimum trip level are selectable. Typically,a slope of 20 or 30% is selected, depending on whetheror not the transformer is equipped with a load-tapchanger. A minimum operating current of 20 to 40% oftap value may be chosen, with 30% being a traditionaland successful setting.

The relay system provides load current metering(including peak demand), menu-driven programmingof settings, three groups of setting tables, program-mable input and output contacts (with and withouttime delay), oscillographic data storage, fault records,and communications ports.

5 TYPICAL APPLICATION OF TRANSFORMERPROTECTION

5.1 Differential Scheme with HarmonicRestraint Relay Supervision

By setting this relay with second- and fifth-harmonicblocking, it will refrain from tripping on the disparitybetween incoming and outgoing currents that resultfrom inrush or exciting current due to overvoltage.Any combination of connections of a transformer canbe accommodated. Contrary to conventional practice,the current transformers should be connected in wyeon each side, irrespective of the phase shift between thecurrents on one side of the transformer and those onthe other. Figure 10-19 describes a typical externalconnection of the current transformers for a TPU2000R relay.

180 Chapter 10

The appropriate phase shift is taken into considera-tion internally in the relay. The currents that aredelivered to the relay are modified in the settingprocess to cause them to appear to have the same valuethey would have had if conventional ct connectionshad been used. This modifier is called the compensa-tion factor. Figure 10-20 shows the majority ofconnections that are in use, the appropriate phase-angle settings, and the compensating factors thatshould be used. Consider, for example, that a deltaconnection of the ct’s were in use in the conventionalsetup. The current delivered to the relay would bemultiplied, by the connection, to be

ffiffiffi3

ptimes the phase

current. Using wye currents only instead of theconventional delta currents, the currents are multi-plied, in the setting process, by

ffiffiffi3

p(the compensating

factor) to obtain equivalency. This internal compensa-tion is only applied to the phase differential algorithm.With the wye-wye connection, where the conventionalpractice would dictate the ct’s be connected in delta,the compensating factors of

ffiffiffi3

pand

ffiffiffi3

pare recom-

mended. This preserves the proportionality betweenthe input and output currents and is consistent with theother combinations. Similar settings to those appliedfor two-winding transformers are used for three-winding transformers.

5.2 Ground Source on Delta Side

As shown in Figure 10-21a, the differential relay willoperate falsely on external ground faults if thedifferential zone covers a grounding bank and aconventional wye-connected current transformer setis used. This misoperation can be eliminated byinserting a zero sequence current trap in the circuit(Fig. 10-21b). The zero sequence current trap consistsof wye-delta-connected auxiliary current transformers,which can have any ratio. These auxiliary currenttransformers provide a low-impedance path for thezero sequence current component and high-impedancepath for the positive and negative sequence current

Figure 10-19 Typical external connection for TPU relay.


Figure 10-21 Ground source on delta side.

Figure 10-20 TPU settings dictated by transformer connections.

182 Chapter 10

components of the fault current. The scheme can begrounded at the differential relay or trap, but only oneground point may be used.

5.3 Three-Phase Banks of Single-Phase Units

Figure 10-22 shows one phase of the transformerdifferential connection for transformer bushing currenttransformers used in a three-phase bank of single-phase units. In such cases, conventional currenttransformer connection cannot be used in the threecircuits to the delta, such as those shown in Figure 10-15. This differential relaying scheme will not detectinternal bushing flashovers if the power system isgrounded as illustrated in Figure 10-22a. This differ-ential relaying scheme will not detect internal bushing

flashovers if the power system is ungrounded, asillustrated in Figure 10-22b. The relay scheme alsowill not detect an internal bushing flashover andexternal ground fault on another phase.

Protection for these internal faults is obtained byplacing current transformers on both bushings of asingle-phase transformer winding that forms part ofthe three-phase delta connection (Fig. 10-22c). Allcurrent transformers can be wye-connected.

An unbalance of 2/1 results from connecting the twocurrent transformers in the delta winding in parallelwith one ct in the wye winding. When selecting currenttransformer ratios and/or relay taps, this unbalancemust be taken into account.

5.4 Differential Protection of a Generator-Transformer Unit

In the unit-type system shown in Figure 10-23, thetransformer differential relay is often connected toinclude the generator as well as the transformer. Thisarrangement provides additional and overlappingprotection for the rotating machine. Separate currenttransformers on the generator neutral are recom-mended to keep the burden low.

Because the station service transformer bank ismuch smaller than the generator unit, the transformerdifferential relay may not protect against secondary orinternal faults in the station service transformer unlesssuch faults occur near the high-voltage end of theprimary winding. The main units require a high currenttransformer ratio to limit secondary currents undercontinuous operation and high faults. Since the stationservice unit is small, it will have a high impedance andlight fault currents—frequently below the transformerdifferential relay sensitivity. In this case, a separatedifferential relay around the station service bank canbe operated with current transformer ratios appro-priate to the size of the bank. Overload relays, withouta separate differential, may be used to protect thisstation service bank.

If no appreciable current is fed back through thesmall station service unit for faults external to the largemain unit, faults on the low side of the station serviceunit may be well below the main differential relaysensitivity. If so, no current transformer connection tothe main differential relay circuit is required. Other-wise, a connection is required to prevent tripping onstation service bus faults.

Figure 10-22 Protection problems and solution for internal

faults on delta side of a three-phase bank consisting of single-

phase units.


5.5 Overexcitation Protection of a Generator-Transformer Unit

Overexcitation of a transformer may result in thermaldamage to cores due to excessively high flux in themagnetic circuits. Excess flux saturates the core steeland flows into the adjacent structure, causing higheddy current losses in the core and adjacent conductingmaterials.

Since flux is directly proportional to voltage andinversely proportional to frequency, the unit ofmeasure for excitation is defined as per unit voltagedivided by per unit frequency (volts/hertz). Over-excitation exists whenever the per unit volts/hertzexceeds the design limits of the equipment; forexample, a transformer designed for a voltage limitof 1.2 per unit at rated frequency will experienceoverexcitation whenever the per unit volts/hertzexceeds 1.2. Should the voltage exceed 120% at ratedfrequency, or the frequency go below 83.3% at ratedvoltage, overexcitation will exist. Severe overexcitationcan cause rapid damage and equipment failure.Figure 10-24 shows the curves of transformer over-excitation limitations for various manufacturers(curves are from Figure 11 of ANSI/IEEE Standard

C37.106—1987 ‘‘IEEE Guide for Abnormal-Fre-quency Protection for Power Generating Plants’’).

In a generator-transformer unit system, the trans-former may be subjected to an overvoltage or over-excitation condition on load rejection or as externalfaults are cleared by the high-side breaker. Duringperiods of high overexcitation, conventional transfor-mer differential relays and relaying schemes mayoperate. Some users, however, consider this ‘‘mis-operation’’ an advantage, since it protects againsttransformer damage from overvoltage.

Figure 10-25 shows a scheme for preventing unde-sired tripping of the differential relay and of protectinga transformer-generator combination against over-voltage. The differential relay must be equipped withsome form of restraint that will recognize excessivevolts/hertz. Either fifth-harmonic or ‘‘all-harmonic’’restraint is available for this function.

Figure 10-24 provides some help in establishingpermissible limits for transformers, but the individualmanufacturer should be contacted for assurances.

In Figure 10-25, device 87 is chosen to provideprotection to both the transformer and generator(and provide backup to the generator differentialrelay, not shown). As voltage increases, due to, for

Figure 10-23 Differential protection of the unit-type generator-transformer system with separate differential protection for the

station service unit.

184 Chapter 10

example, load rejection, the tendency of the differ-ential relay to operate is blocked by, say, thedetection of increased fifth harmonic. As the voltageincreases farther, to the point where damage mayoccur to the protected apparatus, device 24 (volts/hertz relay) takes over to provide tripping that is

time-delayed to coordinate with the capability of thetransformer and generator.

Care must be exercised in choosing both the level offifth harmonic at which blocking takes place as wellas the level of volts/hertz and time at which trippingtakes place. Information such as that provided inFigure 10-26 should be obtained from the individualmanufacturer to identify what may be expected for aparticular transformer or generator.

5.6 Sudden-Pressure Relay (SPR)

With the application of a gas-pressure relay, manytransformers can be protected by a simple differentialrelay set insensitively in the inrush current. Thesudden-pressure relay (SPR), which operates on arate of rise of gas in the transformer, can be applied toany transformer with a sealed air or gas chamberabove the oil level. The relay is fastened to the tank ormanhole cover, above the oil level. It will not operateon static pressure or pressure changes resulting fromnormal operation of the transformer.

The SPR relay is recommended for all units of5000 kVA or more. It is extremely sensitive to lightfaults as it will operate on pressure changes as low as0.33 lb/in.2/sec. In one case, this represented a fault of50A. The SPR relay is far more sensitive to lightinternal faults than the differential relay. The differ-ential relay, however, is still required for faults in thebushing and other areas outside the tank.

The SPR relay operating time varies from one-half cycle to 37 cycles, depending on the size of thefault.

In the past, large-magnitude through-fault condi-tions on power transformers have caused rate-of-change-of-pressure relays to occasionally operatefalsely. There has been reluctance on the part ofsome users to connect these rate-of-change-of-pres-sure relays to trip, and they have therefore usedthem for alarming only. Schemes have been devisedto restrict tripping of the rate-of-change-of-pressuredevice only to levels of current below which thetransformer differential relay cannot operate. Onemeans of doing this is with a high-speed current-blocking type RAICB relay (or a simple overcurrentrelay) to supervise the SPR trip circuit. By wiringthe SPR output to a TPU 2000R input, recording,supervision, and time delay of an SPR operationcan be accomplished.

Figure 10-24 Transformer overexcitation limitations for

various manufacturers transformer under no-load condi-

tions.


5.7 Overcurrent and Backup Protection

To allow transformer overloading when necessary, thepickup value of phase overcurrent relays must be setabove this overload current. An inverse-time charac-teristic relay usually provides the best coordination.Settings of 200 to 300% of the transformer’s self-cooledrating are common, although higher values are some-times used. Fast operation is not possible, since thetransformer relays must coordinate with all otherrelays they overreach.

Overcurrent relays cannot be used for primaryprotection without the risk of internal faults causingextensive damage to the transformer. Fast operationon heavy internal faults is obtained by using instanta-neous trip units in the overcurrent relays. These unitsmay be set at 125% of the maximum through fault,which is usually a low-side three-phase fault. Thesetting should be above the inrush current. Often,instantaneous trip units cannot be used because thefault currents are too small.

An overcurrent relay set to protect the mainwindings of an autotransformer or three-windingtransformer offers almost no protection to the tertiarywindings, which have a much smaller kVA. Also, these

tertiary windings may carry very heavy currents duringground faults. In such cases, tertiary overcurrentprotection must be provided.

A through fault external to a transformer results inan overload that can cause transformer failure if thefault is not cleared promptly. It is widely recognizedthat damage to transformers from through faults is theresult of thermal and mechanical effects. The thermaleffect has been well understood for years. Themechanical effect has recently gained increased recog-nition as a major concern of transformer failure. Thisresults from the cumulative nature of some of themechanical effects, particularly insulation compres-sion, insulation wear, and friction-induced displace-ment. The damage that occurs as a result of thesecumulative effects is a function of not only themagnitude and duration of through faults, but alsothe total number of such faults.

The transformer can be isolated from the faultbefore damage occurs by using fuses or overcurrentrelays. The former ANSI C37.91, ‘‘Guide for Protec-tive Relay Applications to Power Transformers,’’ wasbased on the former U.S. Standard C57 for PowerTransformer, in which mechanical effect was notconsidered. The latest published Standard C57.109–

Figure 10-25 Overexcitation protection for generator and transformer.

186 Chapter 10

1985, ‘‘IEEE Guide for Protective Relay Applicationsto Power Transformers,’’ considers both the thermaland mechanical effects.

For purposes of coordination of overcurrent pro-tective devices, ANSI/IEEE Standard C57.109–1985presents different curves for different-size transformersas listed in Table 10-7. For applications including thecategory I transformer, only the thermal effect fromthe through-fault current is considered in Figure 10-27;on the contrary, for applications including the categoryIV transformer, the thermal and mechanical effectsfrom the through-fault current should be considered as

Figure 10-26 Example of harmonic components on transformer overexcitation.

Table 10-7 Transformer Category (ANSI/IEEE Standard

C57.109-1985 Curves)

Minimum nameplate (kVA)Reference

Category

Single-

phase

Three-

phase

protective

curve

I 5–500 15–500 Fig. 10-27

II 501–1667 501–5000 Fig. 10-28

III 1668–10,000 5001–30,000 Fig. 10-29

IV above 10,000 above 30,000 Fig. 10-30


shown in Figure 10-30. For applications including thecategory II or III transformer, Figures 10-28 and 10-29, whether or not the mechanical effect from thethrough-fault current should be considered depends onthe frequency of the external fault.

For applications in which external faults occurinfrequently, for example, transformers with second-ary-side conductors enclosed in conduit or isolated in

some other fashion, the through-fault protection curveshould reflect primarily thermal damage considera-tions, since the cumulative mechanical damage effectof through faults will not be a problem.

For applications in which external faults occurfrequently, for example, transformers with secondary-side overhead lines, the through-fault protection curveshould reflect the fact that the transformer will be

Figure 10-27 Through-fault protection curve for category 1 transformers 5 to 500 kVA single-phase 15 to 500 kVA three-phase.

188 Chapter 10

subjected to the thermal and cumulative mechanicaldamage effects of through faults.

Figure 10-31 shows the infrequent-frequent faultincidence zones for the determination of curve selec-tion.

The following example describes the procedures forconstructing the dog-leg portion of the thermal/mechanical limit curves: a 230/25-kV, 30/50-MVAtransformer with an impedance of 10% on a 30-MVAbase and with secondary-side overhead lines.

Figure 10-28 Category II transformers 501 to 1667 kVA single-phase 501 to 5000 kVA three-phase.


Step 1 Select the category from the minimumname-plate rating of the principal winding. Forthis example, it is a category III transformer, andthe Figure 10-29 curve should be used for thecoordination.

Step 2 Plot the infrequent through-fault curve ofcategory III (Fig. 10-29), as shown in Figure 10-32, portion A.

Step 3 Determine the dog-leg portion of the curveas follows:

Figure 10-29 Category III transformers 1668 to 10,000 kVA single-phase 5001 to 30,000 kVA three-phase.

190 Chapter 10

1. Calculate the maximum per unit through-fault current:

I ¼ 1

0:10

� �¼ 106base current at 2 sec

This is point 1 in Figure 10-32.2. Calculate the constant K ¼ I2t at t ¼ 2 sec:

K ¼ 1

0:10

� �2

62 ¼ 200

Figure 10-30 Category IV transformers above 10,000 kVA single-phase above 30,000 kVA three-phase.


3. Calculate the time at 50% (note: use 50% forcategory III and IV, 70% for category II) ofthe maximum per unit through-fault current:

t ¼ K

I2¼ 200

½0:5ð10Þ�2 ¼ 8 sec

This is point 2 in Figure 10-32.4. Connect points 1 and 2 and draw a vertical

line from point 2 to the infrequent curve tocomplete the dog-leg portion of the curve asshown in Figure 10-32.

5.8 Distance Relaying for Backup Protection

Directional distance relaying can be used for transfor-mer backup protection when the setting or coordina-tion of the overcurrent relays is a problem. Thedirectional distance relays are connected to operate

when the fault current flows toward the protectedtransformer. They are set to reach into, but notbeyond, the transformer.

5.9 Overcurrent Relay with HRU Supplement

Three single-phase HRU units with instantaneoustrip elements can be used to supplement the time-delay overcurrent relays when inrush is a problemand no current transformers are available on thesecondary side of the protected bank. As illustratedin Figure 10-33, this arrangement provides high-speedtripping when the transformer is energized on faults.The scheme is not recommended, however, unlesstransformer loads are supervised by individual localbreakers, or load pickup does not occur duringtransformer energization.

Figure 10-31 Infrequent-frequent fault incidence zones for category II and category III transformers.

192 Chapter 10

6 TYPICAL PROTECTIVE SCHEMES FORINDUSTRIAL AND COMMERCIAL POWERTRANSFORMERS

The protection of industrial and commercial powertransformer banks is somewhat different from theprotective schemes used by utilities. The differences inprotective schemes are a function of several majorfactors, including system configuration, method ofgrounding, speed, coordination, operation, and cost.Some of the more commonly used industrial andcommercial protective schemes are shown in Figures10-34 to 10-39.

Figure 10-34 illustrates how a primary breaker canbe used for transformer protection. The basic protec-tion is provided by the 87T transformer differentialrelays. Either type TPU, CA, or HU relays can beused, depending on the severity of inrush andoperating speed requirements. Device 50/51, aninverse-time CO relay with IIT unit, provides trans-

Figure 10-32 Multiple of transformer full load (per unit).

Figure 10-33 Overcurrent relay with single-phase HRU

supplement for speed improvement.

Figure 10-34 Transformer protection with primary

breaker. The MSOC relay may be used for the overcurrent

functions.


Figure 10-35 Paralleled transformer protection with primary breaker.

Figure 10-36 Connections and operation at the CWC (87TG) ground differential relay where the connected system is

grounded.

194 Chapter 10

former primary winding backup protection for phasefaults; either device 50G (type ITH with a zerosequence current transformer) or 50N/51N can beused as transformer primary winding backup forground faults. Transformer overload, low-voltagebus, and feeder backup protection are provided bydevice 51 on the transformer secondary side. Since thelow-voltage side is medium-resistance-grounded, aground relay (51G) should be used to trip breaker52-1 for low-side ground faults and for resistor thermalprotection. Device 151G, which trips breaker 52–11,

provides feeder ground backup, whereas device 63,such as a type SPR relay, offers highly sensitiveprotection for light faults.

The current transformer ratings in this schemeshould be compatible with the transformer short-timeoverload capability: approximately 200% of transfor-mer self-cooled rating for wye-connected currenttransformers and 350% ð ffiffiffi

3p

6200%Þ for delta-con-nected current transformers. The neutral currenttransformer rating should be 50% of the maximumresistor current rating.

Figure 10-37 Connections and operation of the CWC (87TG) ground differential relay where the connected system is

ungrounded or the external ground source is not always available.


When a normally closed secondary bus tie breakeris used for paralleled transformer protection (Fig. 10-35), there are several differences, with the primarybreaker scheme shown in Figure 10-34. First, a typeCWC relay (87TG) provides selective and sensitiveprotection for ground faults within the secondarycircuit of the differential zone. The CWC relay, aninduction-disk relay, has two windings that operateon the product of the two currents. The operatingtorque is proportional to the product times the cosineof the angle between the two currents. Maximumtorque occurs when the currents are in phase; theconnections and operation of the scheme are shown inFigures 10-36 and 10-37.

For an external fault, the two currents in the relaycoils are essentially 1808 out of phase. In this case, theCWC relay has no operating torque. For an internalfault, the currents in the relay are essentially in phase,producing operating torque. The relay sensitivity is0.25VA. Make sure that the external ground source isalways available when applying Figure 10-36; other-wise, the CWC relay will not operate on internalfaults.

In these applications, the ratios of the currenttransformers do not have to be identical. Increased

sensitivity can be obtained by using a lower-rationeutral current transformer. It is desirable to keep thecurrents in the two relay coils within a 2:4 ratio so thatan auxiliary current transformer will be required if alarge ground resistor is used.

Figure 10-37 shows the connections when theexternal system is not grounded or the external groundsource is not always available. Iq must be positive forthe external fault. This can be done by using

n ¼ 1:2RCL

RCNor higher

for the auxiliary ct ratio as shown in Figure 10-37a.Current transformer ratios, as well as any effect ofsaturation of the line current transformer, must beconsidered.

In addition to devices 51G and 151G shown inFigure 10-35, a 251G relay is used to trip breaker 52Ton ground faults. The trip sequence of these threeground relays is as follows: (1) 251G trips 52T, (2)151G trips 52-11, (3) 51G trips 52-1. The 87TG trips52-1 and 52-11. Device 67 (type CR relays) providesreverse overcurrent protection.

If the transformer is too small to warrant the basicschemes described above, the scheme shown in Figures10-38 and 10-39 is recommended. Here, fuses providethe primary fault protection. Solid grounding willassure sufficient primary phase fault current to operatethe fuses for most secondary ground faults. Theopening of a single primary fuse will result in singlephasing of the transformer secondary system. Thismay be difficult to detect, particularly at light loads,and appropriate precautionary measures should betaken.

If the primary source is grounded and there is apower source on the secondary side, a ground fault onthe incoming line will be interrupted by the sourcebreaker; the transformer primary or secondarybreaker, however, will not be relayed open becauseof the delta primary transformer connection. Thefailure of these breakers to open can result in hazardsto personnel, possible damaging transient overvoltagesproduced by an arcing-type fault, and problems withautomatic reclosing of the source breaker. Severalschemes can be used to ensure opening of the sourceand transformer breakers, including pilot protection ofthe incoming line, transfer-trip, or potential grounddetection relaying schemes on the transformer pri-mary. Automatic reclosing is a special problem,requiring that the secondary breaker be opened beforethe primary source breaker is reclosed.

Figure 10-38 Transformer protection with primary fuses.

196 Chapter 10

When a normally open bus tie breaker is used, as inFigures 10-35 and 10-39, devices 67 and 67N are notrequired.

7 REMOTE TRIPPING OF TRANSFORMERBANK

Transformer banks are often applied as a part of theline section, with no high-side breakers. The protectionproblems associated with this combination aredescribed in Chapters 12, ‘‘Line and Circuit Protec-tion.’’

8 PROTECTION OF PHASE-ANGLEREGULATORS AND VOLTAGEREGULATORS

A phase-angle regulating transformer, which inserts orimpresses a regulated voltage on a line in quadraturewith its line-to-ground voltage, is used to controlpower flow in the system. A voltage-regulatingtransformer compensates for drops in IR by insertingor impressing a regulated voltage on a line in phasewith its line-to-ground voltage. These two transformersconsist basically of a series unit and an exciting unit,located on at least two separate cores and in separatetanks. Depending on the design and size of the bank,

Figure 10-39 Paralleled transformer protection with primary fuses.


the exciting unit may be wye- or delta-connected, andthe series unit may be constructed as one unit or splitinto two identical units. By mixing the control elementsin the exciting unit, as shown in Figure 10-40, a singlebank can sometimes provide both the phase-angleregulation and voltage control functions. The protec-tive schemes for the phase-angle regulator and voltageregulator are as varied as the ways in which the bank isconstructed. Figure 10-41 shows a typical scheme thatcould be used for the system depicted in Figure 10-40.

Device 87E (HU-1 relays, one per phase; TPUrelays, one per bank; or KAB relays, one per phase)provides overall regulator protection. As shown inFigure 10-42, the currents IS, IL, and Ie are appliedvectorially to the relay. Since all these currents flow inthe primary circuit of the series windings, saturation ofthe series windings from external-fault overvoltageswill not affect the 87E relays.

In application, the current transformer ratio andrelay tap’s selection for 87E would be similar to theconventional differential scheme for three-windingtransformer protection. The use of an equal currenttransformer ratio for the source, load, and primaryof the exciting unit is recommended. This allowsthe use of equal tap settings in the differentialrelay. For most applications, the current transfor-mer ratio is determined by the full-load currentthrough the series winding. However, for larger-angle shifts, the current transformer ratio isdetermined by the current of the exciting-unitprimary winding.

To set the 87E relays at their minimum tap, currenttransformer ratios should be identical. For HU-1 relayapplication, a setting of 2.9 for all restraint elementswill provide the best sensitivity. The series-unitprimary and exciting-unit primary are electrically

Figure 10-40 Typical 400MVA 115KV phase-angle regulator, +268 with voltage control.

198 Chapter 10

connected (instead of magnetically), and there is nophase shift. Consequently, the current transformers for87E relays can be either wye- or delta-connected. Wyeconnection will permit faulted-phase identification,whereas delta connection will produce more current

ð ffiffiffi3

p Þ to operate the relay and two out of the threerelays in the scheme will pick up on an internal faultproviding redundant backup.

Device 87S (HU relays, one per phase or TPU)provides differential protection for the series unit. Asshown in Figure 10-44, the current transformers forcurrent Ie’ should be located at the neutral end of thewindings to provide some backup protection for theexciting unit.

The series winding has very low impedance and isdesigned for rated voltage equivalent to the quadraturevoltage at maximum phase-angle shift. For theexample shown in Figures 10-43 and 10-44, it isapproximately [2 sin (268/2)] or 45% of the line-to-neutral voltage. As a result, the transformer windingsare subject to over-voltage or overexcitation conditionson external fault, which may produce saturation of theseries windings and cause false operation of the 87Srelays.

Figure 10-41 Typical scheme for protecting the phase-angle regulator of Figure 10-40.

Figure 10-42 Overall differential protection for the phase-

angle regulator of Figure 10-40.


Figu

re10-43

Phase-angle

regulator(87S).

200 Chapter 10

Whether or not an external fault will cause suchovervoltage on the winding and false relay operationdepends on several factors, including the characteristicof the series winding (saturation curve, impedance, andtap position), location and type of fault, and power-system condition. A typical analysis shows that faultson one side of the bank produce false operation of therelay, whereas faults on the other side of the bank donot. Other cases indicate no overvoltage problems atall. If overvoltage is a problem, therefore, the 87Srelays should be supervised by volts/hertz or excessivefifth harmonic.

Note in Figure 10-43 that for this phase-shifting-regulator protective scheme that the currents aresummed in the relay at the right. As shown in thisdrawing, the current in R2 (restraint 2) is equal to thesum of the currents in restraints R1 and R3. Differencesin the tap settings are required only if the RCL does notequal RCE (nt/ny).

For illustration purposes, a typical example ofcurrent transformer ratio and relay tap selection forthe 87S device is given below.

It should be noted that the ampere turns in theseries winding, primary and secondary, are alwaysbalanced for any ‘‘through’’ condition.

An example illustrates the 87S relay taps selectionfor a phase shifter as shown in Figure 10-43. It has thefollowing information:

345 kV, plus/minus 608 189/252/315MVASeries winding primary 26 280 turnsSeries winding secondary 232 turnsExciting winding 334 turns

Use the maximum load-current value to determinethe source-/load-side ct ratios:

315; 000

1:736345¼ 528A

That is, use the 1200:5 ratio for the series-unit primary-side ct’s:

Series-unit primary current ¼ 528ASeries-unit secondary

current¼ 528 (26 280/232)¼ 1274A

Exciting-unit secondary current¼ 1.736 1274

¼ 2204A

Therefore, the ct ratio at this location should not belower than 2204/0.8¼ 2755 to 5, i.e., use the 3000/5 ct.87S relay taps selection would be as follows:

Series-unit primary-side ct ¼ 1200/5 deltaExciting-unit secondary-side ct ¼ 3000/5 wyeCurrent from series-unit primary

to relay¼ 1.73(528)(5/1200)¼ 3.804

Current from exciting-unitsecondary to relay¼ 2204(5/3000)

¼ 3.673

Current ratio¼ 26 3.804/3.673 ¼ 2.071

Note the following:

1. A factor of 2 is included in the current’s ratiocalculation for this particular example. This isdifferent from the approach in a conventionaldifferential scheme.

2. The following illustration shows a simpler wayfor finding the current’s ratio:

ne ¼ 232 Rce ¼ 600

ns ¼ 280 RcL ¼ 240

Current ratio ¼ ne6Rce

ns6RcL¼ 2326600

2806240¼ 2:071

Selected relay taps’ ratio ¼ 8:7=4:2 ¼ 2:070

Calculated mismatch ¼<0:01%

Use 8.7 taps for the series-unit ct’s and 4.2 tapfor the exciting-unit ct.In the event that ct ratios are to be

determined, the following expression may behelpful. Make M% approach 0; then the ct

Figure 10-44 Differential protection for series winding of

the phase-angle regulator of Figure 10-40.


ratios can be determined:

M% ¼ne6Rce

ns6RcL� TL

Te

S! 0:0

where

M%¼ percent mismatchTL¼ relay line/restraint tap settingTe¼ relay exiting-unit/restraint tap settingne¼ secondary series-windingnumberof turnsns¼ half-primary series-winding number of

turnsRcL¼ line/source ct ratioRce¼ secondary of exciting-unit ct ratioS¼ smallest of the two ratios

Devices 51N1 (short time) in the neutral circuit of theexciting-unit secondary provides sensitive ground faultprotection for single-phase-to-ground faults on thesecondary side of the exciting unit. The zero sequencecurrent distribution for a ground fault in this area isshown in Figure 10-45.

Device 51N2 provides backup for devices 51N1 and87E during single-phase-to-ground faults on theexciting-unit primary. The current flow in neutraldepends on the autotransformer action of the faultedwinding (Fig. 10-46).

Refer to Figure 10-47. Since there is no delta-connected winding in the exciting unit for this

particular phase shift to provide a path for circulatingzero sequence current, it will provide no zero sequencecurrent for external ground fault. Therefore, the 51N1and 51N2 devices do not have a coordination problemon external faults. However; coordination should beconsidered if there is a zero sequence current path forexternal ground faults in this area.

The sudden-pressure relays are recommended forthese units, especially when there is high potential foran arcing fault in the tap-changing equipment.

9 ZIG-ZAG TRANSFORMER PROTECTION

Since the connected system will be ungrounded in someapplications, a zig-zag grounding transformer can beprotected against ground faults by the scheme shownin Figure 10-48. The overcurrent relays for the delta-connected current transformers provide phase-faultprotection (Fig. 10-48). The time-overcurrent relay(device 51N) in the neutral provides backup groundprotection. The ground relay must be set to coordinatewith ground relays in the connected system. Rate-of-pressure-rise relays, such as the sudden-pressure relays,are recommended for light internal faults.

The grounding banks are seldom switched bythemselves. When they are switched, however, they aresubjected to magnetizing inrush—just as for other types

Figure 10-45 Sensitive ground protection zero sequence currents for a secondary ground fault in the exciting unit of the phase-

angle regulator of Figure 10-40.

202 Chapter 10

of transformers. The harmonic restraint relay (single-phase-typeHRU), as shown inFigure 10-48, can be usedto prevent inadvertent tripping during energization.

Some power transformer banks consist of zig-zag-connected windings for phase correction or systemgrounding. As shown in Figures 10-49 and 10-50, thephase-angle shift between the primary and secondarysides of the banks depends on their connections. Inthese examples, windings on one side are delta-connected. As shown in Figure 10-51, however, awye connection could also be used, introducing aphase-angle shift of either plus or minus 308. Thegrounded wye is not a zero sequence current source forground faults on the wye side, even if both windingsare grounded. It is, however, a good zero sequencecurrent source for ground faults at the zig-zag side.Both the phase-angle shift and zero sequence currentsource should be considered when applying thedifferential scheme for these transformers.

10 PROTECTION OF SHUNT REACTORS

10.1 Shunt Reactor Applications

Both EHV transmission lines and long HV transmis-sion lines and cables require shunt reactance tocompensate for their large line-charging capacitance.This capacitance produces VAR generation that thesystem generally cannot absorb. This VAR genera-

tion increases as the square of the voltage and is afunction of line length and the conductor configura-tion. In many cases, it is necessary to absorb theseVAR and provide voltage control at both terminalsduring normal operation. High overvoltage onsudden loss of load must be limited as well. Systemswitching and operation may require a differentamount of VAR absorption and even, at times,some VAR generation.

Shunt reactance for VAR control is obtained by

Fixed shunt reactorsSwitched shunt reactors or capacitorsSynchronous condensersStatic VAR compensators

Fixed shunt reactors are generally used for EHV andlong HV lines and for HV cables. Switched shuntreactors or capacitors and synchronous condensersare applied in the underlying system and near loadcenters.

Shunt reactors vary greatly in size, type, construc-tion, and application. Their capacities range from 3 to125MVA, at voltage levels from 4.6 to 765 kV. Theycan be single- or three-phase, oil- or dry-type, witheither air or gapped-iron cores. The connections maybe directly (1) to the transmission circuit, (2) to thetertiary winding of a transformer bank that is part ofthe line, or (3) to the low-voltage bus associated withthe line transformer bank. (This third application isnot common.)

Figure 10-46 Ground backup protection and zero sequence currents for a primary ground fault in the exciting unit of the

phase-angle regulator of Figure 10-40.


Line reactors, which are connected directly orthrough a disconnect switch, are a part of thetransmission circuit. Circuit breakers are seldomused. The neutrals of the reactors are solidly groundedor grounded through a neutral reactor. Reactor faultsrequire that all line terminals be open.

When connected to the tertiary of a transformerbank, circuit breakers are generally used, either in thesupply or on the neutral. Opening the neutral breakerdoes not isolate a reactor fault. Tertiary applicationsare operated either ungrounded or grounded throughimpedance.

Line operation without a reactor can result in veryhigh overvoltage when load is lost, such as when oneend is opened. This factor encourages the use of direct-connected reactors to avoid the accidental loss ofservice should load be lost.

Line-connected reactors are generally includedwithin the line protection zone and are often wellprotected by the line relays adjacent to the units.Separate reactor relays are recommended, however,since the remote terminal may have difficulty detectinga reactor fault. These relays can be applied withcurrent transformers sized to the reactor MVA andshould include some way of transfer-tripping theremote line terminals—especially on long lines orwhen the remote terminal is a relatively weak source.With separate reactor relays, the line relays provideadditional backup.

Tertiary-connected reactors can be included in thetransformer bank differential zone. Separate reactor-protection relays are recommended. When practical,the transformer protection zone overlap should beused as backup. Line-side reactor breakers allow the

Figure 10-47 The exciting unit provides no zero sequence current path for external ground fault. Devices 51N1 and 51N2 do

not have coordination problem on external faults.

204 Chapter 10

protection to be separated, so that the transformerbank need not be tripped for reactor faults. In suchcases, the possibility of high voltage during operationwithout the reactors should be examined.

The protective techniques commonly used forreactor primary and backup protection are

Rate-of-rise-of-pressure (applicable to oil units witha sealed gas chamber above the oil level)

Overcurrent (three-phase and/or ground)Differential (three-phase or ground only)

Other protective relaying techniques, such as distance,negative sequence, and current balance, have been usedto a limited extent.

10.2 Rate-of-Rise-of-Pressure Protection

Rate-of-rise-of-pressure protection provides the mostsensitive protection available for light internal faults.

Tripping is recommended, although such protection issometimes used for alarm purposes only. An alarmoperation should be monitored carefully since there arecases where a fault left no tangible evidence after thefirst pressure relay operation but later developed into asevere fault. Even on the severe fault, the pressure relaywas distrusted because of the initial assumed-falseoperation.

Rate-of-rise-of-pressure protection can be used asseparate primary protection only if line or transformerdifferential protection is available for faults outside thereactor tank and for backup protection. Rate-of-rise-of-pressure protection is, of course, not applicable todry-type units.

10.3 Overcurrent Protection

Overcurrent phase and ground protection for reactorsare shown in Figure 10-52. To avoid operation on

Figure 10-48 Protection of a zig-zag grounding transformer and the zero sequence currents for an external ground fault.


transients, the phase-type CO time-overcurrent units(51) are set at 1.5 times the rated shunt reactorcurrent: the IIT instantaneous units (50) are set at fivetimes the rated current. The ground relay unit (51N)can be set at 0.5 to 1.0A and the relay (50N) at fivetimes the 51N setting. Both ground units should be setabove the zero sequence current (3I0) contribution ofthe reactor for faults outside the reactor protectionzone. This setting will avoid operation on line-deenergized oscillations. If the reactor is connectedto an ungrounded system, 50N and 51N should beomitted. This scheme requires only one set of currenttransformers.

10.4 Differential Protection

Separate-phase differential relays (87), as shown inFigure 10-53, are applicable for either three- or single-

phase reactor units. With single-phase units, theseparate differential relays aid in identifying the fault.The relays detect bothwinding and bushing faults. Sincethe relays will see magnetizing inrush as a ‘‘through’’condition, generator-type relays can be used. Either theSA-1 or generator CA-type relays may be applied; bothprovide sensitive internal fault protection (0.14A for theSA-1 and 0.18 A for the CA). The ground relay (50N/51N) provides backup protection.

A single CA-16 or HU-4 relay can be used for aground differential. The four restraints are connectedto the three lines and one neutral current transformer,as shown in Figure 10-54. The minimum pickup of theCA-16 is 0.15 A and of the HU-4 is 0.87 A.

The scheme shown in Figure 10-55 provides anexcellent combination of phase instantaneous and timeovercurrent with ground differential. For single-phasereactors, phase faults that do not involve groundcannot occur—at least within the tank. Therefore, the

Figure 10-49 Interconnected delta zig-zag transformers with voltages in phase on the two sides.

206 Chapter 10

three 50/51 relays represent backup protection, whichcould be omitted.

The type SA-1 or CA generator differential relays orbus differential type KAB relays can also be used forthe ground differential (87N). Additional security canbe obtained by using the type CWC relay, particularlyif the current transformer performance is inferior. Thisconnection is shown in Figure 10-36. Also, the ratiosdo not have to be identical. Sensitivity can be increasedby using a lower-ratio neutral current transformer, asdescribed above.

When the shunt reactor is grounded and connectedto an ungrounded system, a CWC relay can be used,with the connections shown in Figure 10-37.

10.5 Reactors on Delta System

On delta systems, shunt reactors are usually con-nected to the tertiary of a power transformerassociated with the line. Since most faults will involveground, the units or associated system are groundedthrough high resistance for detection purposes. Neu-tral resistance grounding is shown in Figure 10-56,and voltage transformer grounding in Figure 10-57(see Chap. 7). To limit both transient overvoltage andground fault current, the resistor is sized so that I0Requals or exceeds I0C. Since the system capacitance toground is very large, the impedance of the associatedsystem is essentially negligible and is not shown in the

Figure 10-50 Interconnected delta zig-zag transformer with voltages 608 apart on the two sides.


zero sequence diagrams. Whereas the primary currentfor a ground fault is quite small, the secondarycurrent will be large. If 59N is used for alarm insteadof tripping, the secondary current may exceed thecontinuous thermal rating of the voltage transfor-mers.

The CV-8 relay for 59N provides sensitive protec-tion: Its pickup is 8% of its continuous rating. Foralarm applications, the 3E0 voltage should not exceedthe 69- or 199-V rating unless a series resistor is used tolimit the voltage across the relay to its rating. (SeeChap. 8.)

Phase protection for three-phase reactors can beobtained by overcurrent or differential relay schemes.Overcurrent protection is the same as for Figure 10-52,without 50N/51N; differential protection is as shownin Figure 10-54 or 10-53, without 50N/51N. Thearrangements offer little protection for single-phasereactors unless a second ground fault should develop inanother unit.

Although including the reactor within the transfor-mer differential circuit provides some phase-faultprotection, it offers no ground-fault protection withhigh impedance grounding. Even the phase-fault

Figure 10-51 Interconnected wye zig-zag transformer with wye phase voltages lagging 308 from zig-zag phase voltage.

208 Chapter 10

protection is limited, since the current transformers ofthe transformer differential are sized for transformercapacity and not the smaller-reactor MVA.

Low impedance or solid grounding of the reactorsmay be used. In this case, either the 50N/51N neutralovercurrent relay (Figs. 10-56 and 10-57) or 87Nground differential of Figure 10-37 should be applied.

10.6 Turn-to-Turn Faults

Light turn-to-turn faults are extremely difficult todetect. Although the rate-of-rise-of-pressure relay

offers the greatest sensitivity, its application is limited.The reactors must be oil-type, and the fault must causea sufficient pressure change to operate the unit. Whiletransformer action in a turn-to-turn fault can producea large current within the shorted turn, there is very

Figure 10-52 Phase and ground instantaneous and time

overcurrent protection for shunt reactors.

Figure 10-53 Separate phase differential protection with

ground time overcurrent backup protection for shunt

reactors.

Figure 10-54 Phase instantaneous and time overcurrent

with ground differential protection for shunt reactors.

Figure 10-55 Combined phase and ground differential

protection for shunt reactors.


Figure 10-56 Neutral resistance grounding for ground fault detection.

Figure 10-57 Resistance grounding through voltage transformer for ground detection.

210 Chapter 10

little current change at the terminals of the unit. Theeffect is equivalent to an autotransformer with ashorted secondary. The impedance change that willoccur in one phase can be represented by symmetricalcomponents as a shunt unbalance. As shown inFigure 10-58, impedance ZA of phase A is not equalto the other two phases, shown with a total reactorimpedance of ZB. For this condition, the sequencenetworks are connected as shown in Figure 10-58.

Because of the transformer action, the change inimpedance of the total phase circuit for a shorted turnis difficult to calculate. As a rough estimate, assume achange of 3% in the phase with the shorted turn. Alsoassume that the fault has not yet involved ground orother phases. Given these assumptions and if weneglect phase angles, distributed winding capacitance,and transformer action, negative and zero sequencecurrents will be less than 1%. Removing the groundfrom the units does not change the positive andnegative sequence currents significantly, although itdoes eliminate the zero sequence. The magnitudes ofthe currents are largely a function of the total reactorimpedance; the source impedance is relatively lowcompared to the reactor impedance.

The small unbalances and sequence currents asso-ciated with turn-to-turn faults generally are no largerthan the normal or tolerable unbalances. Conse-quently, there seems to be no reliable ‘‘handle’’ todistinguish between the intolerable and tolerable

conditions. Although special schemes or relays havebeen reported, they will require very careful ‘‘custo-mized’’ applications.

As the turn-to-turn fault spreads to more turns, thecurrent will increase. A negative sequence relay may beset at 0.2A negative sequence. The relay should beapplied with a timer to avoid operation on systemtransients and external faults and should be disabledwhen the breaker is opened. This latter safeguardavoids possible operation on low-frequency lineoscillations after the line is deenergized. With verylittle resistance in the line, such oscillations can last anappreciable time.

Figure 10-58 Shunt reactor with a shortened turn in phase

‘‘a’’ so that ZA<ZB and the sequence connections for an

unbalanced impedance in phase ‘‘a.’’


11

Station-Bus Protection

Revised by: SOLVEIG WARD

1 INTRODUCTION

A bus is a critical element of a power system, as it is thepoint of convergence of many circuits, transmission,generation, or loads. The effect of a single bus fault isequivalent to many simultaneous faults and usually,due to the concentration of supply circuits, involveshigh-current magnitudes. High-speed bus protection isoften required to limit the damaging effects onequipment and system stability or to maintain serviceto as much load as possible. The bus protectiondescribed refers to protection at the bus location,independent of equipment at remote locations.

Differential protection is the most sensitive andreliable method for protecting a station bus. Thephasor summation of all the measured current enteringand leaving the bus must be 0 unless there is a faultwithin the protective zone. For a fault not in theprotective zone, the instantaneous direction of at leastone current is opposite to the others, and the sum ofthe currents in is identical to the sum out. A fault onthe bus provides a path for current flow that is notincluded in these summations. This is called thedifferential current. Detection of a difference exceedingthe predictable errors in the comparison is oneimportant basis for bus relaying.

In dealing with high-voltage power systems, therelay is dependent on the current transformers in theindividual circuits to provide information to it regard-ing the high-voltage currents. Figure 11-1 showstypical examples of the location of current transfor-mers that are used for this purpose. The arrowheadsindicate the reference direction of the currents.

1.1 Current Transformer Saturation Problemand Its Solutions on Bus Protection

Bus differential relaying is complicated by the fact thatfor an external fault on one circuit, all of the othercircuits connected to the bus contribute to that fault.The current through the circuit breaker for the faultedcircuit will be substantially higher than that for any ofthe other circuits. With this very high current flowingthrough the current transformer and its circuit breaker,there is a very high likelihood that some degree ofsaturation will occur. A saturated current transformerwill not deliver its appropriate current to the bus relay.With the lower currents in the other circuits for thisexternal fault, the degree of saturation is expected to beconsiderably lower. This may lead to a large differ-ential current that will tend to cause the relay to sensean internal fault rather than the actual external faultthat exists. The relay must accommodate this errorcurrent without misoperation.

A widely used equivalent diagram for a currenttransformer appears in Figure 11-3b. It consists of aperfect transformation from the high current side tothe low current side (e.g., 600:5). All of the significantimperfections are lumped into Rp, Rs, and Xm. The Rsrepresents the internal secondary resistance of the ct(current transformer), and the X represents a currentpath that accommodates the exciting requirements.The ct is assumed to have a uniformly distributedwinding and, therefore, to manifest no significantleakage reactance.

When the ct is subjected to excessive flux, the ct issaid to ‘‘saturate,’’ meaning that the core of the ct has

213

Figu

re11-1

Commonbusarrangem

ents

withrelayinputsources.

214 Chapter 11

been forced to carry more flux than it can handle. Theflux then spills into the area surrounding the core,causing the magnetizing reactance to have a muchlower value than normal. It can be seen that anycurrent that flows in X subtracts from the perfectlytransformed current, producing a deficiency in thecurrent that is delivered to the devices connected to thect. The black blocks are the polarity markers. A singlepolarity marker has no significance. With two, it isacknowledged that, at the instant of time at whichcurrent is flowing into the polarity marker on the highcurrent side of the ct, current is flowing out of thepolarity marker on the low current side. Of course, thecurrent reverses every half cycle, but both the high andlow reverse together.

Direct current saturation is much more serious thanac saturation because a relatively small amount of dcfrom an asymmetrical fault wave will saturate thecurrent transformer core and appreciably reduce thesecondary output. The L/R ratio of the power-systemimpedance, which determines the decay of the dccomponent of fault current, should strongly influencethe selection of the bus protective relaying. Typically,the dc time constants for the different circuit elementscan vary from 0.01 sec for lines to 0.3 sec or more forgenerating plants. The nearer a bus location is to astrong source of generation, the greater the L/R ratioand the slower the decay of the resulting dc componentof fault current.

Of the several available methods for solving theunequal performance of current transformers, four arein common use:

1. Eliminating the problem by eliminating iron inthe current transformer [a linear coupler (LC)system]

2. Using a multirestraint, variable-percentage dif-ferential relay which is specifically designed tobe insensitive to dc saturation (CA-16 relaysystem)

3. Using a high impedance differential relay with aseries resonant circuit to limit sensitivity to ctsaturation (KAB relay system)

4. Using a Differential Comparator relay withmoderately high impedance to limit sensitivityto ct saturation (RED-521)

1.2 Information Required for the Preparation ofa Bus Protective Scheme

Some bus protection schemes rely on the operation ofa remote breaker. It is simple and economic, but slow

(zone-2 trip) and may interrupt unnecessarily a tappedload. When local bus protection is applied, thefollowing information is required for the schemeselection, relay selection, and setting calculations:

1. Information about the bus configuration isrequired. The common bus arrangements areas shown in Figure 11-1, such as single bus,double bus, main-and-transfer bus, ring bus,breaker and a half, bus tie-breaker, double-bus-single-breaker, etc.

2. Maximum and minimum bus fault currents(single-phase-to-ground fault and three-phasefault)

3. Current transformer information, including

Current transformer locationCurrent transformer ratiosCurrent transformer accuracy classCurrent transformer saturation curves

4. Operating speed requirement

1.3 Normal Practices on Bus Protection

The normal practices on bus protection are

1. There is one set of bus relays per bus section.2. Use a dedicated ct for bus differential protec-

tion. If possible, the connection of meters,auxiliary ct’s, and other relays in differential-type bus schemes should be avoided since thesedevices introduce an additional burden into themain circuit.

3. Lead resistance, as well as ct winding resistance,contributes to ct saturation. Therefore, thelength of secondary lead runs should be heldto a minimum.

4. Usually, the full-ct secondary winding tapshould be used. This has two advantages. Itminimizes the burden effect of the cable and,second, leads by minimizing the secondarycurrent and makes use of the full-voltagecapability of the ct.

5. Normally, there is no bus relay required for thetransfer bus on a main-and-transfer busarrangement. The transfer bus is normallydeenergized and will be included in the mainbus section when it is energized.

6. No bus relay is required for a ring bus becausethe bus section between each pair of circuitbreakers is protected as a part of the connectedcircuit.

Station-Bus Protection 215

7. Special arrangements should be considered ifthere is any other apparatus, such as stationservice transformers, capacitor banks, ground-ing transformers, or surge arresters, inside thebus differential zone.

8. There is no simple scheme available for adouble-bus-single-breaker arrangement (Fig.11-1e), because its current transformers arenormally located on the line side. Theseapplications greatly benefit from numericalschemes, such as the RED-521. (Refer to Sec.9 of this chapter for more information.)

2 BUS DIFFERENTIAL RELAYING WITHOVERCURRENT RELAYS

2.1 Overcurrent Differential Protection

This differential scheme requires that a time-over-current relay be paralleled with all of the currenttransformers for a particular phase, as shown inFigure 11-2. It is permissible to use auxiliary ct’s tomatch ratios, but it is preferred that all of the ct’s havethe same ratio on the tap chosen and that the use ofauxiliary ct’s be avoided.

In this scheme, the overcurrent relay must be set tooverride the maximum error current that results froman external fault (phase or ground). It may also benecessary to have sufficient time delay to refrain fromtripping during the time that one or more of thecurrent transformers is severely saturated by the dccomponent of the primary current. To assure this,using a simple overcurrent relay, the current transfor-mers must be chosen to have no more than 20 timesrated current flowing in their primary for the worst-case external fault, and each have a burden no morethan the rated value (relaying-accuracy-class voltage/100). The operating time of the relay must not be less

than three primary time constants, and its setting mustbe greater than the exciting current of the currenttransformer under worst-case conditions. This mayrequire a setting of 10 or more amperes and a timesetting of, say, 18 cycles. These values may beacceptable for smaller substation buses, but moresophistication and faster relaying speed are generallymandatory for more extensive and higher-voltagebuses.

In these applications a ‘‘short time’’ or ‘‘extremelyinverse’’ characteristic overcurrent relay is used in theinterests of getting faster tripping speeds at highcurrent. Operating times of 8 to 18 cycles are expected.Although the relay cost is low, the engineering costmay be high because of the usual need for considerablestudy for the application to assure correct operation.

2.2 Improved Overcurrent DifferentialProtection

The sensitivity of the overcurrent differential scheme(Fig. 11-2) can be improved by externally connecting aseries resistor with each overcurrent relay, as shown inFigure 11-3. These resistors are called stabilizingresistors. If we assume that an external fault causesthe ct on the faulted feeder to be saturated completely,the ct excitation reactance will approach 0. As shownin Figure 11-3, the error current Id that flows through

Figure 11-2 The overcurrent differential bus protection.

Figure 11-3 The improved overcurrent differential bus

protection.

216 Chapter 11

the overcurrent unit would be

Id ¼ IF2RL þRS

2RL þRS þRd

� �ð11-1Þ

where Rd is the resistance in the differential path.In order to reduce the error current Id in the

differential path for improving the sensitivity of thescheme, the most effective way is to increase the valueof Rd. The limitations of this additional resistance aredetermined by (1) the overvoltage to the ct circuit and(2) the minimum available internal fault current. Itshould be limited to

Rd ¼ VCL

46IminpickupO ð11-2Þ

Note: The multiplier 4 includes a safety factor of 2.

3 MULTIRESTRAINT DIFFERENTIAL SYSTEM

Multirestraint differential schemes use conventionalcurrent transformers, which may saturate on heavyexternal faults. For this reason, the secondary currentoutput may not represent the primary. In a differentialscheme, the current transformers and relay function asa team. When the current transformers do not performadequately, the relay can within limits make up for thedeficiency.

The multirestraint differential scheme uses the CA-16 variable-percentage differential relay, which con-sists of three induction restraint units and oneinduction operating unit per phase. Two of the unitsare placed opposite each other and operate on acommon disc. In turn, the two discs are connected to acommon shaft with the moving contacts. All four ofthe units are unidirectional; that is, current flow ineither direction through the windings generates con-tact-opening torque for the restraint units or contact-closing torque for the operating unit. Each restraintunit (called R, S, and T) also has two windings toprovide restraint proportional to the sum or difference,depending on the direction of the current flow. If thecurrents in the two paired windings are equal andopposite, the restraint is cancelled. Thus, the pairedrestraint windings have a polarity with respect to eachother. With this method six restraint windings areavailable per phase.

In addition to providing multiple restraint, thevariable-percentage characteristic helps in overcomingcurrent transformer errors. At light fault currents,current transformer performance is good, and the

percentage is small for maximum sensitivity. For heavyexternal faults, current transformer performance islikely to be poor, and the percentage is large. Thevariable-percentage characteristic is obtained by ener-gizing the operating unit through a built-in saturatingautotransformer.

The saturating autotransformer also presents a highimpedance to the false differential current, which tendsto limit the current through the operating coil and toforce more equal saturation of the current transfor-mers. On internal faults, in which a desirable highdifferential current exists, saturation reduces theimpedance. A further advantage of the saturatingautotransformer is that it provides a very effectiveshunt for the dc component, appreciably reducing thedc sensitivity of the operating units. At the minimumpickup current of 0.15+ 5% A, the restraining coilsare ineffective.

When using the CA-16 relay, the current transfor-mers should not saturate when carrying the maximumexternal symmetrical fault current; that is, the excitingcurrent should not exceed one secondary ampere rms.This requirement is met if the burden impedance doesnot exceed

½NPVCL � ðIEXT � 100Þ�RS

1:33 IEXTð11-3Þ

where

NP¼ proportion of total current transformerturns in use

VCL¼ current transformer accuracy-class voltageIEXT¼maximum external symmetrical fault current

in secondary (amperes rms) (use IEXT¼ 100if IEXT< 100)

RS¼ current transformer secondary winding resis-tance of the turns in use (in ohms); forexample, if the 400:5 tap of a 600:5 wye-connected class C200 current transformer isused, then NP¼ 400/600¼ 0.67 andVCL¼ 200

If IEXT¼ 120A and RS¼ 0.5O, then the burden ofthe ct’s secondary circuit, excluding current transfor-mer secondary winding resistance, should not exceed

0:676200� ð120� 100Þ0:51:336120

¼ 0:78O

Settings for the CA-16 relay need not be calculated.Field experience indicates that one CA-16 relay perphase is satisfactory for the vast majority of applica-tions.


External connections are as shown in Figures 11-4through 11-6. Figure 11-4 may be used if only threecircuits are involved. The term circuit refers to a sourceor feeder group.

When several circuits exist and the bus can bereduced to four circuits, then the scheme of Figure 11-5may be used. For example, assume a bus consists oftwo sources and six feeders, and that the feeders arelumped into two groups. The bus now reduces to fourcircuits.

In paralleling current transformers, each feedergroup must have less than 14 A load current (restraintcoil continuous rating).

If the bus reduces to more than four circuits, thenthe scheme of Figure 11-6 should be used. In applyingthe scheme of Figure 11-6, each primary circuit mustbe identified as either a source or feeder. As definedhere, a feeder contributes only a small portion of thetotal fault current for a bus fault. All other circuits aresources. Next, a number of feeders are lumped into afeeder group by paralleling feeder current transfor-mers. Each feeder group must have less than 14 A loadcurrent and not contribute more than 10% of the totalphase- or ground-fault current for a bus fault. Thenconnect the ‘‘source’’ and ‘‘feeder groups’’ alternatelyas shown in Figure 11-6.

Note that in Figures 11-4 through 11-6, electro-magnets R, S, and T are referred to. Each of theseelements has two windings. The polarity markings areextremely significant as related to one another on thesame electromagnet, but have no significance withrespect to one another on different electromagnets. Ifthe current into a polarity marker is equal to the

Figure 11-6 Connection of one CA-16 relay per phase to

protect a bus with six equivalent circuits. (Connections for

one phase only are shown.)


protect a bus with four equivalent circuits. (Connections for



protect a bus with three equivalent circuits. (Connections for


218 Chapter 11

current out of the polarity marker on the sameelectromagnet, there will be no restraining torqueproduced by that electromagnet. The sum of all of therestraint torques is compared to that produced by theoperating coil. Current into the operating coil circuitproduces a much stronger effect than the same currentthrough a single restraint winding. For an externalfault, there is no current through the operating coil ifthe current transformers perform perfectly. There willbe substantial restraint for this same condition, eventhough in some restraint electromagnets some (or eventotal) cancellation may take place.

Consider a fault on the bus of Figure 11-5 in whichall of the high-voltage circuits contribute the samevalue of current. All of the restraint cancels because ineach of the electromagnets the current into the polaritymarker equals the current out of its paired coil. All ofthe internal fault current (in secondary terms, ofcourse) flows into the operating coil circuit and fasttripping occurs. Practical cases with widely differingfault contributions produce similar effects even thoughconsiderable restraint torque may be present.

Consider, now, an external fault on the upper circuitoff of the bus with the equal fault current contributionsthat were assumed in the previous case. Torquecancellation occurs in electromagnet T, as before.Substantial restraint torque is produced by R and S.The operating coil current cannot exceed the errorcurrent in the faulted circuit (which may well beextreme due to the effect of saturation).

This is a very sensitive bus relaying scheme, and it isvery secure against operation for external faults eventhough severe ct saturation may occur for one or morect’s. It is reasonably fast. Another advantage is that itcan accept auxiliary ct’s in the circuit, which allowsdifferent ratios of the main ct’s. Two shortcomings areits comparative inflexibility as other circuits are addedto the bus and the need to bring all circuits back fromthe switchyard to the relay location.

4 HIGH IMPEDANCE DIFFERENTIAL SYSTEM

Although the high impedance differential scheme alsouses conventional current transformers, it avoids theproblem of unequal current transformer performanceby loading them with a high impedance relay (Fig.11-7).

This arrangement tends to force the false differentialcurrents through the current transformers rather thanthe relay operating coil. Actually, the high impedancedifferential concept comes from the above ‘‘improved

overcurrent differential’’ approach. It uses a highimpedance voltage element instead of ‘‘a low impe-dance overcurrent element plus an external resistor.’’

The high impedance differential KAB relay consistsof an instantaneous overvoltage cylinder unit (V), avoltage-limiting suppressor (varistor), an adjustabletuned circuit, and an instantaneous current unit (IT).

On external faults, the voltage across the relayterminals will be low, essentially 0, unless the currenttransformers are unequally saturated. On internalfaults, the voltage across the relay terminals will behigh and will operate the overvoltage unit. Since theimpedance of the overvoltage unit is 2600O, this highvoltage may approach the open-circuit voltage of thecurrent transformer secondaries. The varistor limitsthis voltage to a safe level.

Since offset fault current or residual magnetismexists in the current transformer core, there is anappreciable dc component in the secondary current.The dc voltage that appears across the relay will befiltered out by the tuned circuit, preventing relaypickup.

The IT current unit provides faster operation onsevere internal faults and also backup to the voltageunit. The range of adjustment is 3 to 48 A.

The KAB relay has successfully performed opera-tions up to external fault currents of 200 A secondaryand down to an internal fault current of 0.27Asecondary. Its typical operating speed is 25 msec.

The overvoltage unit is set by calculating themaximum possible voltage for an external fault asfollows:

VR ¼ KðRS þRLÞ IFN

ð11-4Þ

Where

VR¼ pickup setting of the V unit in volts rmsRS¼ dc resistance of current transformer secondary

winding, including internal leads to bushingterminals

RL¼ resistance of lead from junction points to themost distant current transformer (one-waylead for phase faults, two-way lead forphase-to-ground faults)

IF¼maximum external primary fault current, inamperes rms, contributed by the bus

N¼ current transformer turns ratioK¼margin factor

The maximum voltage occurs for the external faultwhen the faulted circuit current transformer is


completely saturated, and there is no saturation in thesource current transformers. The maximum voltage isequal to the resistance drop produced by the secondarycurrent through the leads and secondary winding ofthe saturated current transformer. In practice, thefaulted current transformer will never completelysaturate, and the source current transformers willtend to saturate. As a result, the actual maximumvoltage is less than the theoretical value. The marginfactor K, which modifies this voltage, varies directlywith the current transformer saturation factor SF:

1

SF¼ ðRS þRLÞIF

NVkð11-5Þ

where

Vk¼ knee voltage value of the poorest currenttransformer connected to the relay. For type

KAB relay application, the knee voltage isdefined as the intersection of the extension ofthe two straight-line portions of the saturationcurve. The ordinate and abscissa must use thesame scales.

The margin factor curve, shown in Figure 11-8, isbased on tests of the KAB relay in the high-powerlaboratory. A safety factor of 2 has been included inconstructing this curve.

The maximum number of circuits that can beconnected to the relay, or the minimum internal faultcurrent required to operate the relay, can be estimatedfrom the following equation:

Imin ¼ ðXIe þ IR þ IVÞN ð11-6Þ

where

Figure 11-7 External connection of type KAB bus differential relay.

220 Chapter 11

Imin¼minimum primary fault current in amperesrms

Ie¼ secondary excitation current of the currenttransformer at a voltage equal to the settingvalue of the V unit in amperes

IR¼ current in the V unit at setting voltage VR inamperes, that is, IR¼VR/2600

IV¼ current in varistor circuit at a voltage equal tothe setting value of the V unit in amperes(generally negligible)

N¼ current transformer turns ratioX¼ number of circuits connected to the bus

In general, the following factors should be consideredwhen applying a high impedance bus differential relay.

4.1 Factors that Relate to the Relay Setting

The V-unit setting of the KAB relay is based on thecalculated result of Eq. (11–4), which is determined bythe values of K, RS, RL, and IF. In order to keep thissetting value within the available relay range of 75 to400V, it is necessary to keep the values of (RSþRL)and any additional burden in the ct secondary as lowas possible. This includes the consideration of thefollowing:

Use fully distributed winding current transformers,such as bushing ct’s or current transformers withtoroidally wound cores, such as those used inmetal-clad switchgear. These ct’s provide anegligible leakage reactance and therefore donot contribute to the internal impedance in theequivalent circuit of the ct. Only the RS resistanceis needed in series with RL in Eq. (11–4).

The use of auxiliary ct’s is discouraged, though,with proper consideration of their resistance inseries with the lead resistance (raising theeffective RL), they may be used at the sacrificeof some sensitivity of fault recognition. The samecomment applies to the introduction of otherdevices in the current transformer circuits.

The junction point for all of the ct’s in the busdifferential system should be in such a location asto equalize as much as possible the distance fromeach ct to this point. This will minimize RL, thevalue used in the setting calculation and thusallow better sensitivity to be achieved. Departurefrom this requirement is permissible in metal-cladswitchgear because of the comparatively shortdistances usually involved.

The lead resistance from the junction point to therelay terminals is not critical.

Note that with this system total saturation of thecurrent transformer on a circuit feeding an externalfault is allowed and the relay remains secure.

4.2 Factors that Relate to the High-VoltageProblem

All ct’s in the bus differential circuit should beoperated on their full-tap position. Refer toFigure 11-9; a high voltage will be induced on theunused portion of the ct circuit due to auto-transformer action.

All current transformers should have the same ratio.If taps must be used, the windings between the

Figure 11-9 High voltage induced by autotransformer

action.

Figure 11-8 Empirical margin factor for setting the V-unit

of the KAB relay.


taps must be completely distributed, and anyhigh voltage at the full-tap terminal caused byautotransformer action should be checked toavoid insulation breakdown. In general, auxiliaryct’s should not be used to match ratios.

4.3 Setting Example for the KAB Bus Protection

Assume a six-circuit bus for which the maximumexternal three-phase fault current is 60,000A rms,symmetrical; the maximum external phase-to-groundfault current is 45,000A, and the minimum internalfault current is 10,000A. The current transformerratios are 2000:5, ANSI class C400, Vk is 375V. Thesecondary winding resistance RS is 0.93, and one-waylead resistance to junction point RL is 1.07O.

4.3.1 Settings for the V Voltage Unit

For the three-phase fault condition [using Eq. (11-5)],

1

SF¼ ð0:93þ 1:07Þ60,000

4006375¼ 0:8

From Figure 11-8, 1.2>K� 0.82 (use the lower valueof 0.82 for sensitivity); therefore, using Eq. (11-4), weget

VR � 0:82ð0:93þ 1:07Þ 60,000400

¼ 246V

For the phase-to-ground fault condition,

1

SF¼ ð0:93þ 261:07Þ645,000

4006375¼ 0:92

And from Figure 11-8, 1.1>K� 0.77; therefore, usingEq. (11-4) yields

VR � 0:77ð0:93þ 261:07Þ 45,000400

¼ 266V

The minimum setting of the V unit in the KAB relay,therefore, should be 266V, the larger value for eitherthe three-phase or phase-to-ground conditions, ascalculated.

4.3.2 Setting for the IT Current Unit

The IT setting is determined from Figure 11-10. Thehigher value is used as the ordinate as determined fromthe three-phase and phase-to-ground fault. Thus, for

the example, the ordinate value is

Three-phase fault ¼ ð0:93þ 1:07Þ60,000400

¼ 300

Phase-to-ground fault

¼ ð0:93þ 261:07Þ45,000400

¼ 345

From these numbers, it is obvious from Figure 11-10that the IT unit is incapable of operating for anexternal fault. The lowest available setting of 3A willusually be adequate because of the high conductionlevel of present-day varistors. The principal trippingfunction is accomplished at high speed by the voltageunit, and only in extreme circumstances will the IT unitoperate for an internal fault.

5 DIFFERENTIAL COMPARATOR RELAYS

These relays use the fundamental principle described inFigure 11-11. The RADSS is a solid-state version, theREB-103 is similar to this, but the logic is accomplishedwith a microprocessor, while the RED-521 is entirely anumerical relay. All are very high-speed relays (9- to 16-msec tripping) and are very secure against misoperationfor external faults; all reliably and sensitively detectinternal faults and are quite flexible in accommodatingadditional circuits. They may also be used for generatorstator protection and for shunt reactor protectionthough their prime application area is for bus protection.

The RADSS and the REB-103 relays use externalauxiliary current transformers which allow substan-

Figure 11-10 Setting of KAB instantaneous unit.

222 Chapter 11

tially different main circuit current transformers to beaccommodated and also reduce current to a suitablelevel for the relay. The RED-521, being a micropro-cessor relay, is able to accept widely varying inputsfrom the main current transformers and to provide,internally, the appropriate scaling factors. The RED-521 is therefore very suitable for double-bus-single-breaker arrangements as no external ct switching takesplace. The ct is connected to the appropriate protectionzone numerically inside the relay.

Taking advantage of Kirchoff’s law, the schemecompares the sum of all of the currents entering thebus with the sum of all of the currents leaving the bus.These are instantaneous currents (as opposed to rms oraverage currents.)

In the circuit of Figure 11-11, the currents aredelivered to the relay through the diodes. The sum ofthe currents through the lower group of diodes isrepresentative of the instantaneous sum of the incom-ing currents to the bus, and the current flowing to theupper group of diodes is representative of theinstantaneous sum of the currents leaving the bus.These two sum currents are always in perfect balanceprovided the current transformers perform their jobfaithfully and there is no fault on the bus (or to state itmore correctly, provided there are no current paths offof the bus that are unaccounted for).

If an internal fault (phase or ground) were to occur,the currents in and out would no longer match. Theywould differ by the amount of the fault current. Thisdifference current appears as I DIFF in the relay.

To accommodate the inherent errors in the currenttransformers for an external fault, particularly in the ct

associated with the circuit on which the external faultoccurred, restraint is developed across the resistor Rs.

Any condition that produces I DIFF current will,through the transformer and the full-wave bridge,generate a voltage Vd3. For the through fault case, therestraint voltage Vs will exceed the operating voltageVd3, and the relay will refrain from operating. For theinternal fault case, I DIFF will be large, Vd3 will exceedVs, current will be passed through the diode and thereed relay DR, and tripping will occur. SR is a ‘‘start’’relay whose contact supervises tripping to add to theoverall security of the relay. It is obvious that this relayis extremely fast because the decision to trip is basedon instantaneous currents.

The RED-521 numerical relay uses this principle,but is not encumbered by need for the auxiliarymatching current transformers, the diodes, or anyother of the components inherently required in thecomparison process. The individual samples of cur-rents are collected and summed appropriately todevelop numerically the I IN and I OUT values andthe corresponding restraint quantity. This is comparedwith the difference of these individual sums, I DIFF,and a determination of the need to trip is established.

6 PROTECTING A BUS THAT INCLUDES ATRANSFORMER BANK

Ideally, when the bus includes a power transformerbank, separate protection should be provided for thebus and transformer, even though both protectionschemes must trip all breakers around the two units.

Figure 11-11 Differential comparator relay.


Such a system offers maximum continuity of service,since faults are easier to locate and isolate. Also, usinga bus differential relay for bus protection andtransformer differential relay for transformer protec-tion provides maximum sensitivity and security withminimum application engineering.

However, economics and location of current trans-formers often dictate that both units be protected inone differential zone. For these applications, either themultirestraint HU-4 or CA-26 relays should be used.

TheHU-4 relay is similar to theHUandHU-1 relays,except that it has four restraint windings. Also, therectified outputs of the restraint transformers areconnected in series, providing a higher restraint forcewhen a through fault occurs on the bus. Since the dcsaturation of current transformers will allow current topass into theHRUtransformers andpossibly pickup theIIT, the IIT unit of the HU-4 relay is set at 15 times therms tap value toprevent false tripping for external faults.

Similar to the CA-16, the CA-26 relay has astronger contact spring and higher pickup of1.25+ 5% A to help override inrush. Its variablerestraint curve is steeper than the CA-16, and itsoperating time is approximately three cycles.

Of the two types, the HU-4 relay is preferred, as it isimmune from operation on transformer magnetizinginrush. The HU-4 should always be applied for largetransformer banks or those associated with HV andEHV buses. A typical application, shown in Figure 11-12, protects a three-winding transformer bus with fourcircuits. Figure 11-13 illustrates another typical appli-cation used in EHV systems.

The CA-26 relay is applicable to relatively smalltransformers remote from generating stations, HV,and EHV buses. Here, inrush will usually be light andnot cause the CA-26 to operate. If, however, completesecurity against inrush is required, the HU-4 must beapplied.

With CA-26 relays, the four-circuit bus connectionsof Figure 11-5 are not recommended for bus protec-tion, since the relay may have too much restraint for abus fault.

The bus CA-16 relay should not be used for thetransformer differential, since it is too sensitive tooverride magnetizing inrush.

7 PROTECTING A DOUBLE-BUS SINGLE-BREAKER WITH BUS TIE ARRANGEMENT

The double-bus single-breaker with bus tie (Figure 11-1e) provides economic and operating flexibility com-parable to the double-bus double-breaker arrangement(Fig. 11-1c). However, the ct’s are normally on theline-side location, which results in increased differen-tial relaying problems. Two different approaches havebeen used in the bus protection of such arrangements:the fully switched scheme (Fig. 11-14) and theparalleling switch scheme (Fig. 11-15). They are bothcomplicated (inserting switch contacts in the ctcircuits) and/or imperfect in protection. These schemeseither require switching ct’s and/or disabling the busprotection before any switching operation. This is aperiod when the probability of a bus fault occurring ishigh and it is most desirable that the bus protection bein service. A third scheme as shown in Figure 11-16 canbe considered. It is similar to the paralleling switchedscheme except a check-zone relay is added as shown.

Figure 11-12 Typical application of HU-4 relay for

protecting a large transformer bank associated with HV

and EHV buses. (Auxiliary current transformers for ratio

matching are not shown.)

Figure 11-13 Protection of a typical transformer section

where the transformer tertiary is brought out for load or

connected to an external source.

224 Chapter 11

Figure 11-14 Fully switched scheme.

Figure 11-15 Paralleled switched scheme.


Two bus differential zones are provided, one for eachbus, with each one overlapping the bus breaker. Eachprimary circuit is normally switched to a specific bus,and relay input circuits and breaker control circuits arewired accordingly. The additional check-zone devicesupervises the trip circuits. If it becomes necessary toclear one of the buses, all the primary circuits may beswitched to the opposite bus and it is needless todisable the bus protection before any switchingoperation. However, this scheme still has two draw-backs when any one or all of the primary circuits isswitched to the opposite bus: (1) It will lose itsselectivity, and (2) it will reduce its sensitivity sincethe two relays are paralleled.

A numerical scheme, such as RED-521, overcomesthese drawbacks as there is no external ct switchinginvolved. The ct’s are connected to the appropriatezone by numerical switching in the relay.

8 OTHER BUS PROTECTIVE SCHEMES

Other methods for protecting buses are in limited use:(1) partial differential schemes, (2) directional compar-ison relaying, and (3) the fault-bus method. Except forthe latter, these schemes are most often applied aseconomic compromises for the protection of buses.

8.1 Partial Differential Relaying

This type of protection is also referred to as ‘‘busoverload’’ or ‘‘selective backup’’ protection. It is avariation of the differential principle in which currents

in one or more of the circuits are not included in thephasor summation of the current to the relay.

In this scheme, only the source circuits aredifferentially connected, as shown in Figure 11-17b,using a high-set overcurrent relay with time delay. Thect’s protecting the feeders or circuits are not in thedifferential connection.

Essentially, this arrangement combines time-delaybus protection with feeder backup protection. Thesensitivity and speed of this scheme are not as good aswith complete differential protection. This methodmay be used as a backup to a complete differentialscheme, as primary protection for a station with loadsprotected by fuses, or to provide local breaker failureprotection for load breakers.

In modern microprocessor systems, provision hasbeen included to allow communications between thefeeder breaker relaying and the source breaker relaying.The feeder breakers are each equipped with a nontrip-

Figure 11-16 Paralleled switch with check zone scheme.

Figure 11-17 Partial differential protection.

226 Chapter 11

ping low-set instantaneous overcurrent function that isset somewhat above their maximum load. The sourcebreakers have an instantaneous overcurrent unit withslight time delay that is set above the maximum totalload current for the bus, and they are equipped toreceive a status input from the feeder breakers. For afault on one of the feeder circuits, the low-setinstantaneous overcurrent unit operates and applies ablock signal to the source relay. The instantaneous unitof the source breaker operates, but is unable to tripbecause of the block signal. The time-delayed andcoordinated tripping of the source breaker is notaffected so its backup function stays intact.

For a bus fault, the block signal is absent, andtripping of the source breaker occurs at high speed.

Some partial differential circuits use distance-typerelays in the scheme. The use of a distance relay for thisscheme produces both faster and more sensitiveoperation than the overcurrent scheme.

8.2 Directional Comparison Relaying

Occasionally, it is desirable to add bus protection to anolder substation where additional ct’s and control cableare too costly to install. In this instance, the existing ctcircuits used for line relaying can also be used for thedirectional comparison bus relaying protection.

As shown inFigure 11-18, the directional comparisonrelaying uses individual directional overcurrent relayson all sources and instantaneous overcurrent relays onall feeders. The directional relays close contacts whenfault power flows into the bus section. Back contacts onthe overcurrent relays open when the fault is external onthe feeder.All contacts are connected in series, andwhenthe fault occurs on the bus, the trip circuit is energizedthrough a timer. A time delay of at least four cycles willallow all the relays to decide correctly the directionof thefault and to permit contact coordination.

In this scheme, the ct’s in each circuit do not requirethe same ratio and can be used for other forms ofrelaying and metering.

The disadvantage of this scheme is the large numberof contacts and complex connections required. There isalso the remote possibility of the directional elementsnot operating on a solid three-phase bus fault as aresult of 0 voltage.

8.3 Fault Bus (Ground-Fault Protection Only)

This method requires that all the bus supportingstructure and associated equipment be interconnected

and have only one connection to ground. An over-current relay is connected in this ground path as shownin Figure 11-19. Any ground fault to the supportingstructure will cause fault current to flow through therelay circuit, tripping the bus through the multiple-contact auxiliary tripping relay. A fault detector,energized from the neutral of the grounded transfor-mer or generator, prevents accidental tripping. Thisscheme requires special construction measures and isexpensive.

Figure 11-18 Directional comparison bus protection.

Figure 11-19 Fault bus.


12

Line and Circuit Protection

Revised by: ELMO PRICE

1 INTRODUCTION

1.1 Classification of Electric Power Lines

Alternating current lines are commonly classified byfunction, which is related to voltage level. Althoughthere are no utility-wide standards, typical classifica-tions are as follows:

1. Distribution (2.4 to 34.5 kV) Circuits trans-mitting power to the final users.

2. Subtransmission (13.8 to 138 kV) Circuitstransmitting power to distribution substationsand to bulk loads.

3. Transmission (69 to 765 kV) Circuits trans-mitting power between major substations orinterconnecting systems, and to wholesaleoutlets. Transmission lines are further dividedinto

High voltage (HV): 69 to 230 kVExtra-high voltage (EHV): 345 to 765 kVUltra-high voltage (UHV): greater than 765 kV

1.2 Techniques for Line Protection

Most faults experienced in a power system occur onthe lines connecting generating sources with usagepoints. Just as these circuits vary widely in theircharacteristics, configurations, length, and relativeimportance, so do their protection and techniques.

There are several protective techniques commonlyused for line protection:

1. Instantaneous overcurrent2. Time overcurrent3. Directional instantaneous and/or time over-

current4. Step time overcurrent5. Inverse time distance6. Zone distance7. Pilot relaying

1.3 Selecting a Protective System

Several fundamental factors influence the final choiceof the protection applied to a power line:

1. Type of circuit Cable, overhead, single line,parallel lines, multiterminals, etc.

2. Line function and importance Effect on servicecontinuity, realistic and practical time require-ments to isolate the fault from the rest of thesystem

3. Coordination and matching requirementsCompatibility with equipment on the associ-ated lines and systems

4. Influence on power system stability

To these four considerations must be added economicfactors and the relay engineer’s preferences based onhis or her technical knowledge and experience. Becauseof these many considerations, it is not possible toestablish firm rules for line protection. This chapter,however, focuses on basic application rules andcoordination procedures to aid the engineer in theselection of proper protective systems for both phase

229

and ground faults with the techniques as listedin Section 1.2, except for pilot relaying. Also, thischapter covers the basic protective concept of series-compensated transmission lines using distance relaytechniques.

Pilot relaying is covered in a companion book, PilotProtective Relaying (Marcel Dekker).

1.4 Relays for Phase- and Ground-FaultProtection

Relay systems for the phase-fault protection of powerlines are outlined in Table 12-1, those for ground faultsin Table 12-2.

1.5 Multiterminal and Tapped Lines and WeakFeed

The protection of multiterminal and tapped lines andweak feed will also be discussed in this chapter.Multiterminal and tapped lines, although usuallyeconomical in their breaker requirements, need com-plex relaying for adequate protection and operation. In

fact, these lines are the most difficult to protect,particularly with weak feed (limited-fault current) andwhen high-speed reclosing is desired at one or moreterminals. Weak-feed protection may also be requiredfor two terminal lines. Although the weak-feedterminal can maintain the fault arc, the current maynot be sufficient to operate conventional protectiverelays adequately.

The following definitions will be used throughoutthis chapter:

Multiterminal lines. Transmission lines with morethan two terminals, each connected to a majorpower source. The source will provide positivesequence fault current and, usually, zerosequence as well. A transformer bank may beincluded as part of the transmission line at one ormore of the terminals.

Tapped lines. Transmission lines that are tapped(usually through a transformer bank) primarilyto supply loads. Behind the tapped line there maybe a positive sequence source, either localgeneration or an interconnecting tie with anotherpart of the power system. There may also be azero sequece source.

Table 12-1 Relay Protection Systems for Phase Faults

Basic relay type

Type of protection

Device

no. Electromechanical Static or numericala

Time overcurrent 51 CO MCO, MMCO, 51

IMPRS, MICRO-51

Instantaneous and time

overcurrent

50/51 CO with IIT MCO, MMCO, 51

IMPRS, MICRO-51

Directional time overcurrent 67 CR 32þMMCO, 32þ 51

32þMICRO-51

Directional instantaneous

overcurrent

67 KRV 32þ 50Db

Step time overcurrent 51 CO-4 51þ 50D

Directional instantaneous and

directional time overcurrent

67 IRV 32þMMCO, 32þMICRO-51

Inverse time distance system 21/51 Two KD-10, plus two-element CO

Zone distance system 21 Two KD-10, plus KD-11, plus

two TD-5 (or one TD-52)

Complete zone phase distance

system

REL-300 (MDAR), REL-301/REL-302,

REL-512

Complete distribution package DPU 1500R, DPU 2000R, MSOC

(nondirectional)

a Type numbers refer to ABB circuit-shield types. Certain functions require two relays, with the output of the controlling relay wired to the torque-

control input of the second relay.b Select 50D with a 0.01 to 0.03 adjustable range.

230 Chapter 12

Weak-feed terminal. A terminal whose source doesnot supply enough current for faults on the lineto operate the line protective relays at thatterminal. This situation can occur for eitherphase (positive sequence), ground (zerosequence), or both. The terminals may only be‘‘weak’’ during some operating periods, but‘‘strong’’ or have only load at other times. Atapped terminal is frequently a weak feed sourceif the tapped load has limited local generation,synchronous motors, and does not have manylarge induction motors.

2 OVERCURRENT PHASE- AND GROUND-FAULT PROTECTION

2.1 Fault Detection

Most of the faults on power lines can be detected byapplying overcurrent relays, since the fault currents arenormally higher than the load current.

Radial circuits can be protected by nondirectionalovercurrent relays. Figure 12-1 shows several sectionsof a typical radial circuit. Because the circuit is radial,each section requires only one circuit breaker at thesource end. To clear a fault at (1) and other faults tothe right, then, only the breaker at R needs to be

tripped. To clear faults at (2) and (3) and in the areabetween them, the breaker at H must be tripped.Likewise, to clear faults at (4) and (5) and betweenthem, the breaker at G must be tripped.

However, none of the relays at the breaker locationscan distinguish whether the remote fault is on theprotected line, the remote bus, or an adjacent line. Therelays at H, for example, cannot distinguish betweenfaults at (1) and (2), since the current magnitudemeasured at H will be the same in either case. Openingbreaker H for fault (1) is not desirable, since it wouldinterrupt the load at R unnecessarily. Two techniquesare available to solve this problem: time delay or pilotrelaying. The latter requires a communication channelbetween the two stations and is covered in thecompanion book, Pilot Protective Relaying.

Table 12-2 Relay Protection for Ground Faults

Basic relay type

Type of protection Device no. Electromechanical Static or numericala

Time overcurrent 51N CO MSOC

MICRO-51

Instantaneous and time overcurrent 50N/51N CO with IIT MSOC

MICRO-51

Product overcurrent 67N CWC or CWP

Directional time overcurrent 67N CRC, CRP, CRD, or CRQ MSOCþ 32D or 32Q, MICRO-51

þ 32D or 32Q, 51þ 32D or 32Q

Directional instantaneous

overcurrent

67N KRC, KRP, KRD, or KRQ 50Dþ 32D or 32Qb

Directional instantaneous and time

delay

67N/50N IRC, IRP, IRD, or IRQ MSOCþ 32D, 51þ 32D, or 32Q

Complete zone ground distance

system

REL-300 (MDAR), REL-301/REL-302,

REL-512

Complete distribution package DPU 1500R, DPU 2000R

a Type numbers refer to ABB circuit-shield types. Certain functions require two relays, with the output of the controlling relay wired to the

torque-control input of the second relay.b Select 50D with a 0.01 to 0.03 adjustable range.

Figure 12-1 A typical radial feeder.

Line and Circuit Protection 231

2.2 Time Overcurrent Protection

2.2.1 Time-Delay Relaying

Time relaying delays the operation of the relay for aremote fault, allowing relays and breakers closer to thefault to clear it, if possible. In the example shown inFigure 12-1, relays at H will delay for faults at (1) or(2). If the fault is at (1), this delay will allow the Rrelays and breaker to operate before H. Thus, althoughH would not open for a fault at (1) (unless the R relaysor associated breaker failed), it would operate for afault at (2). This technique, called coordination orselectivity, is designed to combine minimum operatingtime for the close-in faults with a long enough delay forremote faults. In Figure 12-1, e.g., the relays andbreaker at R must coordinate or select with those tothe right (not shown), H must coordinate with R, andG with H.

2.2.2 Coordination

Relays are coordinated in pairs. If, in Figure 12-1,breaker H relay-tripping characteristics have alreadybeen coordinated with whatever protective devicesexist at R and beyond, the breaker at G must then becoordinated with those at H.

For the three critical fault points, (5), (3), and (2),the following data are required:

1. Fault at (5). Maximum and minimum faultcurrents

2. Fault at (3). Maximum fault current, whichdetermines the required coordination betweenbreakers G and H

3. Fault at (2). Minimum fault current, whichdetermines when the G relays must operate toprovide backup protection for faults on line HRnot cleared by the breaker at H

Relays within a system can be coordinated usinggraphs or tables, although graphs are generally moreuseful for radial systems. Semilog (log abscissa forcurrent and linear ordinate for time) or log-log papercan be used. Log-log is preferred when a number ofdifferent types of devices, including fuses, are beingcoordinated on one graph. The current scale can be inprimary amperes or per unit. Any difference in currenttransformer ratios must be taken into considerationwhen determining actual relay currents at differentlocations.

The coordination procedure is conducted as follows(Fig. 12-2). First, assume that the desired relay type(tap range and time characteristic) and current

transformer ratio have been determined. (The selectionof these variables will be discussed later in thischapter.) Then perform the following:

1. Determine the critical fault locations and faultcurrent values.

2. Plot these variables on the time-current graph,drawing vertical lines at the various values.

3. Determine the setting for the most downstreamrelay for the maximum and minimum faultcurrents. Set the relay as sensitive and fast aspossible if there is no other device downstreamthat has to be coordinated with. For example,consider the relay R in Figure 12-1. If there issome other device, such as a power fuse, to theright of this relay, then relay R should be

Figure 12-2 Coordination setting procedure for relays at

breaker ‘‘G’’ of Figure 12-1.

232 Chapter 12

coordinated with the power fuse first. If thereare no other devices to coordinate with down-stream, set the overcurrent relay equal to orgreater than 2.0 maximum load.

4. Plot the operating time of relay R on the time-current graph, shown as XR and YR points inFigure 12-2, respectively.

5. Add a one-step coordinating time interval(CTI); (see Sec. 2.2.3) to points XR and YR.This step gives two set points for the character-istic curve of the relay at H.

6. Select a tap for relay H to operate for fault (1)minimum and, for a phase relay, not to operateon maximum load. The fault (1) minimumshould operate the relay on at least twicepickup, although compromises may be neces-sary (see Sec. 2.2.4). For phase relays, thesetting must always be above the maximumload.

7. Select a time lever such that the relay H time-current curve passes through or above one orboth of the set points XR and YR and providesthe minimum operating time for maximum andminimum fault.

8. Repeat the above steps for each time section‘‘up-stream.’’ For example, add a one-step CTIto XH and YH in Figure 12-2 for relay G,respectively; then select a tap and time lever forrelay G, etc.

2.2.3 Coordinating Time Interval

The coordinating time interval is the minimum timebetween the operating characteristics of two seriesdevices. Factors influencing the CTI are as follows:

1. Breaker fault interruption time2. Relay-impulse-time overtravel of the induction

disk or solid-state relay after the fault currenthas been interrupted

3. Safety margin to compensate for possibledeviations in calculated fault currents, relaytap selection, relay operating time, and currenttransformer ratio errors

For coordinating at above approximately three timesminimum trip current (at least two times the settingvalue), the CTI should be in the range of 0.2 to 0.5 sec.Larger CTIs should be used on the steep part of thecurve to compensate for errors below a multiplier of 3.A CTI of 0.3 sec is commonly used. Lower valuesshould be used only after careful consideration of 1through 3 above.

2.2.4 Selecting an Overcurrent Relay Tap

As indicated above, phase overcurrent relays must notoperate on the maximum load current that can occuron the line. Situations in which temporary overloadsmay occur, such as the cold loads discussed in the nextsection, must be factored into the value used for settingthe overcurrent relays. Thus, it is important for therelay engineer to cooperate with the operating engi-neers in determining the maximum possible load foreach circuit. This maximum-value STM (short-timemaximum) load can differ from the rating of the lineand is the value that should be used for setting therelays.

The tap (minimum pickup value of the phaseovercurrent relays) should be at least 2 (a safety factorfor security) times normal maximum load and neverless than 1.5 times. If we assume that the STM isgreater than the normal maximum load, the tap can beselected as the next available tap greater than 1.25STM.

Dependability should be checked once the relay tapis selected. In Figure 12-8, the minimum fault currentI2min through the relay for a fault on the remote bus H,divided by 2 (a dependability factor), should be greaterthan the selected tap value if the relay is set forprotecting the line only; or the minimum fault currentI3min through the relay for a fault on the remote bus R,divided by 2, should be greater than the selected tapvalue if the relay is set for protecting the line andremote backup of the line (HR) beyond.

Current transformers are normally selected toprovide secondary currents between 4 and 5A duringrated maximum load. As a result, the phase relaypickup will usually be above 5A.

The above limitation does not apply to groundrelays, since load current does not produce current intheir operating windings unless it is unbalanced. Toavoid operation on possible imbalances in a normallybalanced circuit, a good rule of thumb is to set theground relays for not less than 10% of the maximumload current. Four-wire distribution circuits will, ingeneral, require a much higher setting than this.

2.2.5 Selecting an Overcurrent Relay TimeCurve

Five different curve shapes have been established bythe vast number of electromechanical relays that are inservice on power systems and that dictate coordinationrequirements. Solid-state and numerical relays imple-ment these shapes while often allowing others. These


widely used time-current characteristics are describedas:

1. Definite time, CO-62. Moderately inverse, CO-73. Inverse, CO-84. Very inverse, CO-95. Extremely inverse, CO-11

These time-current characteristics are compared inFigure 12-3. The time lever settings are selected so thatall relays operate in 0.2 sec at 20 times the tap setting.

The microprocessor-based overcurrent relay, typeMCO, is a single-phase one, and type MMCO is athree-phase and ground package. All the above time-current curves are built in these relays, and can beselected by settings. The equations for the MCO orMMCO time-current curves are

TðsecÞ ¼ T0 þ K

ðM� CÞp� �

D

24; 000for M�� 1:5

¼ R

ðM� 1Þ� �

D

24; 000for M < 1:5

where

T¼ trip time in secondsD¼ time dial setting from 1 to 63M¼ operating current in terms of multiple of tap

settingT0¼ definite time termK¼ scale factor for the basic inverse timeP¼ an exponent determining inverseness

T0, K, C, P, and R are constants and are shown asbelow:

Curve

no. T0 K C P R

CO-2 111.99 735.00 0.675 1 501

CO-5 8196.67 13,768.94 1.130 1 22,705

CO-6 784.52 671.01 1.190 1 1475

CO-7 524.84 3120.56 0.800 1 2491

CO-8 477.84 4122.08 1.270 1 9200

CO-9 310.01 2756.06 1.350 1 9342

CO-11 110.00 17,640.00 0.500 2 8875

Operating time as shown in Table 12-3 illustratesthat the MCO or MMCO provides a fairly good time-current characteristic for coordinating with the con-ventional type of CO relay. Similar curves and other

variations are available in the Micro-51, DPU 2000R,DPU 1500, and Microshield relays.

The choice of a relay time-current characteristic is afunction of the sources, lines, and loads. Since thesefactors vary throughout a system, a characteristic thatis ideal for one line and one operating conditionrequires compromises for other conditions and associ-ated lines.

If possible, time curves with the same or approxi-mately the same characteristics should be used.Identical or similar curves applied at different placesin the system tend to ‘‘track’’ together as operatingconditions change. If different time characteristiccurves must be used, all possible operating conditionsmust be checked carefully to ensure that the CTI ismaintained for selective tripping. (Using similarcharacteristics, in other words, minimizes coordinationstudies.)

Fixed- and inverse-time characteristics for a systemare compared in Figure 12-4. The last feeder supplyingone load center can be protected with an instantaneousovercurrent device set into the load. Since nocoordination at the load is involved, no time delay isrequired, as shown with bus R in both time-distancecharts.

In the upper chart of Figure 12-4, the fixed-timecharacteristics approximate the definite minimum time(CO-6). The relay at H is coordinated with R, and Gis coordinated with H, as shown. The advantage ofthis arrangement is that the operating times arerelatively constant and independent of changes infault levels from maximum to minimum generation.On the other hand, the operating times for heavy

Figure 12-3 Type CO curve shape comparison.

234 Chapter 12

faults near the source are very long. For this reason,this arrangement is not practical when there are morethan one or two radial feeders from the distributionsubstation.

The lower chart of Figure 12-4 shows inverse-timerelay characteristics. For faults near the relay, parti-cularly for the maximum conditions, operating timesare very short. Unfortunately, as system conditionschange from maximum to minimum, operating timesvary considerably. Even though this arrangement can

produce very long operating times for minimum faultsnear the remote bus, it is commonly used.

Line length is also an important factor. For a shortline, one whose impedance is low compared to thesource impedance, the fault currents for the close-inand far-end faults are essentially the same; that is, theinverse-time characteristic gives a relatively fixedoperating time over the line. In such cases, the definiteminimum time characteristic is preferred, since theoperating time will not vary as much for differentgeneration levels as with inverse relays.

In general, the following apply:

1. The flatter curves (CO-6 and CO-7) are moresuitable when:

(a) There are no coordination requirementswith other types of protection devicesfarther out in the system.

(b) The variation in current for faults at thenear and far ends of the protected circuit istoo small to take advantage of the inversecharacteristic.

(c) Instantaneous trip units give good cover-age (see Sec. 3.1).

2. The ‘‘inverse-time’’ relay (CO-8) provides fasterclearing time than the ‘‘more inverse-time’’relays for low-current faults. This would beadvantageous on long lines where the availablefault current is much less at the end of the linethan at the local end. It does not provide muchmargin for cold load pickup.

3. The steeper (more inverse) curves (CO-9 andCO-11) are more suitable when

(a) Fault currents are significantly differentfor the close-in and remote faults (for

Table 12-3 Relay Operation Times

Conventional CO MCO or MMCO

Type of relay

Time dial

set at

Time of operation

at 46 pickup current

(Fig. 12-3) (sec)

Time dial

set at

Time of operation

at 46 pickup current

(sec)

CO-6 0.6 0.25 6 0.25

CO-7 1.0 0.40 7 0.44

CO-8 1.2 0.60 7 0.58

CO-9 2.1 0.70 11 0.62

CO-11 5.0 2.00 31 2.00

Figure 12-4 Comparison of fixed time vs. inverse time

overcurrent relays on radial feeder circuits.


example, when the line impedance is largecompared to the source impedance).

(b) There is an appreciable current inrush onservice restoration (cold load).

(c) Coordination with other types of deviceswith very inverse characteristics, such asfuses and reclosers, is required.

General Comments on Curve Shape Selection

There is no known scientific means of determining theideal curve shape for a specific application, except tomake preliminary setting calculations. However, somegeneral comments can be made:

1. Use a CO-6 definite minimum time relay whencoordination is not a problem.

2. Use a CO-6 relay for short line application.3. Use a CO-11, extreme inverse-time relay when

fuses are involved.4. The more inverse shape (CO-8, CO-9, and CO-

11) is more suitable in loop systems.5. Use a comparable shape within a system

segment for easier coordination.

One of the advantages of using a microprocessor-basedovercurrent relay is that a different time curve can beselected in the same device without changing the unit.

2.2.6 Effect of Extended Load Outage/Cold-LoadInrush

A particularly critical phenomenon for distributioncircuits serving residential and commercial loads is thehigh transient current inrush that may occur when afeeder is energized after a prolonged outage. For these‘‘cold-load’’ conditions, the diversity of intermittentloads is lost: Consumers tend to leave more than thenormal load connected, and thermostatically con-trolled equipment will start as soon as the voltage isrestored. The overall effect is a very high initialcurrent, or cold-load inrush.

In general, the pickup of time-overcurrent relayscannot be set above this transient without severelycompromising protection. Setting the relay below thetransient will cause it to begin to operate on the coldload; however, the current will decrease below thepickup value before the relay has time to operate.

A current-time curve for the cold-load inrush on atypical feeder is shown in Figure 12-5. Since suchcurves vary considerably with different feeders, eachutility must develop its own system history andprobability data. If we assume that the time-over-current CO relays are set at twice the normal

maximum load, they will receive operating currentfor the first 2.3 sec that the feeder is energized (Fig. 12-5). The average current for this period is 3.4 p.u., anequivalent of 1.7 times pickup. To prevent tripping thebreaker, then, the relay operating time at 1.7 times thetap value should be slightly more than 2.3 sec (about2.5 sec). For the CO-9 very inverse relay, this conditionrequires a time dial setting of 1.25; for the CO-11extremely inverse relay, a time dial setting of 0.75 isrequired. Time dial settings of 1.5 for the CO-9 relayand 1.0 for the CO-11 relay are suggested, unlessoperating experience indicates otherwise.

Although the extremely inverse relays may providefaster fault operations than the less inverse type ofrelays, they still override the cold-load inrush.

Modern microprocessor relays contain logic thatallows sensing of the energization of a feeder circuit. Ashort time delay in instantaneous tripping can beintroduced to prevent operation on cold-load pickupwhile still permitting sensitive and relatively fast faultrecognition.

Often, it is necessary to sectionalize the feeder andto pick up the load in increments in order to reenergizethe cold load without undesired tripping.

2.2.7 Fuse and Relay Coordination

Because fuses have a time-current characteristic that ismuch more inverse than most induction-disc time-overcurrent characteristics, coordinating these relaysand fuses can be difficult (Fig. 12-6). A fuse curve andtwo sets of relay curves, one for the extremely inverserelay and one for the very inverse relay, are plotted ona linear time vs. logarithmic current scale. The right-hand set of relay curves provides a good margin ofprotection at high levels of fault current, but isunsatisfactorily slow for medium values of faultcurrent, particularly with extremely inverse character-

Figure 12-5 Typical example of feeder load current follow-

ing extended outage (cold load inrush).

236 Chapter 12

istics. If the right curve were moved to the left, it wouldcoordinate better at higher current values. As shown inFigure 12-6, the curve would then cross the fuse curveat lower values of fault current. Usually, either the veryinverse or extremely inverse characteristic can be set tocoordinate with fuses. By adjusting the tap and timelever settings, areas of crossing such as those shown inFigure 12-6 may impose impractical or impossibleoperating conditions on the circuit.

When plotting fuse curves, the following three timecharacteristics must be considered:

1. Maximum time that the fuse will carry currentwithout suffering damage

2. Melting time for the fuse links3. Total clearing time for the fuse to clear the

circuit

The first two characteristics are used for coordinationwith protective devices beyond the fuse. Normally,curves based on the melting-time characteristic areprovided with a ‘‘safety band’’; in this case,maximum time curves are not required. The totalclearing-time characteristic is used for coordinationwith other protective devices, including relays, aheadof the fuse.

2.3 Instantaneous Overcurrent Protection

Adding instantaneous trip units to time-overcurrentrelays provides high-speed relay operation for close-infaults and may also permit faster settings on the relaysin the adjacent section.

Instantaneous trips may be used on a circuit if themaximum close-in fault current is on the order of 1.1to 1.3 or more times the maximum fault current on busH (see Fig. 12-8).

I1max > ð1:1 to 1:3Þ I2max ð12-1ÞThe factor of (1.1 to 1.3) in Eq. (12-1) is for preventingthe instantaneous unit from overreaching. In otherwords, the instantaneous unit must operate for as manyof the line faults as possible, but to avoid miscoordina-tion, it must not operate for the far-end fault. Thegreater the ratio of close-in to far-end faults, the more ofthe line the instantaneous unit will protect. In terms ofsystem constants and setting, the reach or coverage ofthree-phase faults on a line can be determined as follows:

n ¼ SIRð1�KiÞ þ 1

Kið12-2Þ

Figure 12-6 Current (logarithmic scale) comparison of fuse

and relay curves.

Figure 12-7 Connections for overcurrent ground relay.

Figure 12-8 Criteria for a directional unit requirement at

relay breaker A.


where

n¼ per unit of line section length protected bythe instantaneous unit

Ki¼ instantaneous unit pickup current; IITmaximum far-end fault current; IF

SIR¼ source impedance ratio

¼ source impedance; ZS

protected line impedance; ZL

Refer to Appendix A in this chapter for moreinformation about Eq. (12-2).

Recommended values of Ki, are 1.3 for the solenoidor plunger units with transient overreach (IIT, SC, ITunits), 1.2 for static or mumerical units (MCO, LIunits), and 1.1 for the cylinder units with negligibletransient overreach (KC-2, KC-4, KO, KR, and IRtypes). The value of 1.25 can be used as a generalfactor.

The minimum value that can justify the use of aninstantaneous unit for line protection is a matter ofchoice. Since the relative cost of adding the instanta-neous units is quite low, they are recommended evenwhen the line coverage is low for maximum faultsand 0 for minimum faults. The arrangement providesfast protection for the most severe, heavy, close-infaults.

Cold-load inrush may be above the instantaneousunit setting desired for maximum fault protection. Toavoid operation when a setting above this inrush isnot practical, the instantaneous trip circuit can bemanually opened at restoration and left open untilthe instantaneous trip unit resets. For manualoperation, a slip contact on the control switch thatis open while the switch is held in the ‘‘close’’position (or its equivalent) prevents operation untilthe inrush subsides to drop out the instantaneousunit. If set above cold-load inrush, the instantaneoustrip unit setting should be at least three times theovercurrent tap setting, or around six times thenormal maximum load.

2.4 Overcurrent Ground-Fault Protection

Ground overcurrent relays are for faults involving zerosequence quantities, primarily single-phase-to-groundfaults and sometimes two-phase-to-ground faults.With a few significant differences, the general applica-tion rules for phase relays also can be applied toground relays.

The directional or nondirectional overcurrent typesare used widely at most voltage levels. In addition to

their lower cost and complete independence of load,the power system provides more rapid attenuation ofcurrent with distance and the relatively higher inde-pendence of system changes. This makes their applica-tion and setting easier than for phase relays.

Ground relays usually can be set and coordinatedindependently of phase relays, even though the faultedphase current does flow through the one or more phaserelays for a single-phase-to-ground fault. The primaryreason for this independence is that ground relays areset at one-fifth to one-tenth of the sensitivity of phaserelays. The more sensitive settings obtainable withground overcurrent relays may mean (for electro-mechanical relays) a higher burden on the currenttransformers, and their performance should bechecked as described in Chapter 5.

A circuit may be protected with a single, nondirec-tional overcurrent ground relay, as shown in Figure 12-8. Positive and negative sequence currents are balancedout at the current transformer neutral, so only 3I0currents pass through the ground relay (50N/51N).Since, under normal balanced conditions, 3I0, is at orapproaches 0, a very low pickup current is used,typically 0.5 to 1.0A. Although ground-fault currentson distribution circuits are generally higher at thesubstation than phase-fault currents, they decrease at amuch greater rate with the distance from the sub-station because X0 is considerably larger than X1 forthe feeder circuits. With the exception of fault currentvalues, the application and coordination of thenondirectional overcurrent relays are the same as forphase relays, as given above.

As with phase relays, instantaneous ground trip canbe used to improve relaying, particularly for close-infaults. Instantaneous ground-trip units are moreapplicable in general, with the higher attenuation ofthe fault currents with distance. Unless care isexercised in the choice of settings, high transientoverloads and unequal current transformer perfor-mance can give rise to ‘‘false residual currents’’ and,hence, misoperation.

The choice of a relay time characteristic for lineprotection is usually limited to the inverse or veryinverse type. The very inverse type is the morecommonly used. However, when coordination withfuses and/or series trip reclosers is required, anextremely inverse characteristic would probably bepreferable.

The foregoing descriptions can be summarized asfollows:

1. Factors that are favorable to ground-faultprotection:

238 Chapter 12

Normally load current does not affect the ground-relay operation. This means that the ground relaycan be set more sensitively than the phase relay.

Ground relays are not affected by out-of-stepconditions.

Ground relays always have available the unfaultedphase voltages for polarizing. They may notrequire a memory circuit.

The higher zero sequence line impedance Z0L, ascompared with the positive line impedance Z1L,may allow one to use a high-set ground over-current unit and make coordination easier thanfor phase faults.

The zero sequence isolated system may makecoordination easier.

2. Factors that are not favorable to ground-faultprotection:

Most of the time, ground faults involve higher faultresistance; this may introduce an over- or under-reach problem to the relay.

Zero sequence mutual effect may cause a ground-relay directionality problem.

The zero sequence current distribution factor is notequal to the positive sequence current distribu-tion factor, except for a single end feed condition;this makes the ground relay more complicatedthan the phase relay in design.

The ground relay faces more problems than thephase relay on reverse fault clearing (e.g., contactbounce, unequal pole clearing, etc.).

3 DIRECTIONAL OVERCURRENT PHASE- ANDGROUND-FAULT PROTECTION

3.1 Criteria for Phase Directional OvercurrentRelay Applications

When there is a source at more than one of the lineterminals, fault and load current can flow in eitherdirection. Relays protecting the line are thereforesubject to fault power and reactive flowing in bothdirections. If nondirectional relays were used, theywould have to be coordinated with not only relays atthe remote end of the line, but also the relays behindthem. Since directional relays operate only when faultcurrent flows in the specified tripping direction, theyavoid both this complex coordination and the possi-bility of compromising line protection.

Figure 12-8 shows a line, with a source at each end,that could be a section of a loop. The following

procedure determines the criteria for a directional unitby comparing the currents flowing through the relayfor faults at either bus.

3.1.1 For Phase Directional Time-OvercurrentRelays

A directional time-overcurrent relay should be appliedat G if, as in Figure 12-8, the maximum reverse-faultcurrent (I4max) for a fault on bus G or the maximumreverse-load current IRLoad through the relay exceeds0.25 (this value includes the safety factor 2 anddependability factor 2) times the minimum faultcurrent I2min through the relay for a fault on theremote bus H for the protection of the line only; orthe minimum I3min through the relay for a fault on theremote bus R if the relay is set for the protection ofthe line on the remote bus R and remote backup of theline (HR) beyond. In other words, a directional relayshould be used when

I4max or IRLoad

I2min or I3min>¼ 0:25 ð12-3Þ

3.1.2 For Phase Directional Instantaneous-TripOvercurrent Relays

A directional instantaneous-trip overcurrent unitshould be used if the maximum reverse-fault currentI4max is greater than the maximum I2max (Fig. 12-8).Also, as mentioned in Section 2.3 before, for instanta-neous-trip overcurrent application, the criteria asshown in Eq. (12-1) should be met.

3.2 Criteria for Ground DirectionalOvercurrent Relay Applications

With a few significant differences, the general rules ofapplication for directional overcurrent phase relaysalso apply to ground relays. Normal balanced loadcurrent is not a consideration. However, the groundovercurrent unit still should be set above anymaximum expected unbalanced load current.

3.3 Directional Ground-Relay Polarization

To determine the direction to a fault, a directionalrelay requires a reference against which line currentcan be compared. This reference is known as thepolarizing quantity and, in this context, reference andpolarizing are synonymous terms. With zero sequence


line current, either a zero sequence current or voltageor both must be used. In power systems with mutualinduction problems, the trend is toward the use ofnegative sequence quantities for the ground directionalunit.

3.3.1 Voltage (Potential) Polarization

The zero sequence voltage at or near any bus in aninductive power system can be used for polarization.The voltage measured on the bus or just on any linenear the station will have the same direction for anyfault location. However, the current through the linebreaker will change direction according to the faultlocation.

The polarizing zero sequence voltage is obtainedfrom the broken-delta secondary of grounded-wyevoltage transformers (Fig. 12-9). Phase voltages arealso required for the phase relays, instrumentation, etc.In such cases, either a double-secondary voltagetransformer or device, or a set of auxiliary wye-grounded, broken-delta auxiliary transformers, can beused. The voltage across the broken delta, VXY, alwaysequals 3V0, or VAG plus VBG plus VCG.

3.3.2 Current Polarization

The current polarization reference depends on theavailability and connection of the power transformerat the relay location. Various bank connections arediagrammed in Figure 12-10. The current in theneutral of a wye-grounded delta power transformercan be used for polarizing. For almost any groundfault on the wye-side system, this current flows up theneutral when current is flowing to the fault. (Anexception with high mutual induction will be discussedlater.) Thus, a current transformer in the neutralmeasures 3I0 for IP current polarization, the polarizingcurrent shown in Figure 12-10a for a wye-delta bank,and that shown in Figure 12-10d for a zig-zag bank.Wye-wye banks, either grounded or ungrounded,cannot be used for polarizing (Fig. 12-10b and12-10c). The grounded wye of three-windingwye-wye-delta banks can be used (Fig. 12-10e, f,and g).

The separate neutral currents of the three-windingwye-delta-wye transformers cannot be used for polar-izing, but current transformers in each groundedneutral must be parallel with inverse ratios, as shownin Figures 12-10f and 12-10g. If we assume that boththe high- and low-voltage sides connect to a groundsource, ground faults on the low-voltage side (Fig. 12-10g) result in current flowing up the low-voltageneutral and down the high-voltage neutral. Conversely,for faults on the high-voltage side (Fig. 12-10f), currentflows up the high-voltage neutral and down the low-voltage neutral. Hence, the reversal in either neutraldoes not provide a reference. By paralleling the twoneutral current transformers, however, IP always fallsin the same direction for faults on either side since, ona per unit basis, the current flowing down the neutral isalways less than the current flowing up the otherneutral. The actual current distribution will vary asdetermined by the zero sequence network.

The tertiary or delta winding can also be used as apolarizing source. If there are no external circuits fromthe delta, one current transformer connected in any legof the delta will provide I0. A current transformer isrequired in each of the three windings if the delta isconnected to external circuits, so that positive and/ornegative sequence currents can exist during load orfaults. These current transformers must be connectedin parallel to cancel out positive and negative sequenceand provide 3I0 only.

Autotransformers should not be used for currentpolarizing without careful analysis because they arefrequently unreliable as a reference. AutotransformersFigure 12-9 Zero sequence polarizing voltage source.

240 Chapter 12

that are ungrounded or without a delta tertiary cannotbe used. Such units can bypass zero sequence inroughly the same way as the wye-wye-grounded banks.A grounded autotransformer with a tertiary is shownin Figure 12-11.

For Ground Faults on the High-Voltage Side

(Figure 12-11a, if we assume the high-side source isopened for simplicity). The zero sequence currentflowing through the high-voltage winding of theautotransformer is thus I0H. The current flowing inthe low-voltage circuit will be P0I0H(VH/VL). Becauseof the physical connections, the current in the neutral

of the bank is

INH ¼ 3 I0H � P0VH

VLI0H

� �

¼ 3I0H 1� P0VH

VL

� �ð12-4Þ

From this, it is evident that a positive value of ð1�P0VH=VLÞ will assure that the neutral is a reliablepolarizing source.

From the zero sequence network, we obtain

P0 ¼ ZT

Z0LS þ ZL þ ZT

� �ð12-5Þ

Figure 12-10 Zero sequence polarization from power transformer banks.


Substituting yields

INH ¼ 3I0H6 1� ZT

Z0SLþ ZL þ ZT6

VH

VL

� �ð12-6Þ

The tertiary current (inside the delta) is

ITH ¼ I0Hð1� P0Þ VHffiffiffi3

pVT

at VT ð12-7Þ

and since

ð1� P0Þ ¼ Z0SL þ ZL

Z0SL þ ZL þ ZT

� �ð12-8Þ

ITH ¼ I0HZ0SL þ ZL

Z0SL þ ZL þ ZT

� �

6VHffiffiffi3

pVT

at VT ð12-9Þ

For Ground Faults on the Low Voltage Side

(Figure 12-11b, if we assume the low-side source isopened for simplicity). I0L is the zero sequence currentin amperes from the autotransformer. The current

from the high voltage system will be R0I0L(VL/VH) atVH. Again, the bank neutral current is

INH ¼ 3 I0L �R0VL

VHI0L

� �

¼ 3I0L 1�R0VL

VH

� �ð12-10Þ

From the zero sequence network, we get

R0 ¼ ZT

Z0SH þ ZH þ ZT

� �ð12-11Þ

Substituting yields

INL ¼ 3I0L6 1� ZT

Z0SH þ ZH þ ZT6

VL

VH

� �ð12-12Þ

The tertiary current (inside the delta) is

ITL ¼ I0Lð1�R0Þ VLffiffiffi3

pVT

at VT ð12-13Þ

Figure 12-11 Polarizing from grounded autotransformer banks.

242 Chapter 12

and since

ð1�R0Þ ¼ Z0SH þ ZH

Z0SH þ ZH þ ZT

� �ð12-14Þ

ITL ¼ I0LZ0SH þ ZH

Z0SH þ ZH þ ZT

� �

6VLffiffiffi3

pVT

at VT ð12-15Þ

If the neutral or tertiary current is used forpolarizing, Eq. (12-6) and (12-12) or (12-9) and (12-15) must give a positive operating value for all possiblevariations of Z0SH and Z0SL. However, in sometransformer designs, especially those for autotransfor-mers, the ZH or ZL may be negative, the direction ofthe INH or INL may be reversed, and this current is nota reliable source for polarization, when the combinedimpedance of (Z0SHþZH) or (Z0SLþZL) is a negativevalue.

3.3.3 Dual Polarization

The approach is called dual polarization if thedirectional ground overcurrent relay uses both zerosequence voltage 3V0 and zero sequence current Ip forpolarization. It provides more flexibility in applica-tions.

3.3.4 Negative Sequence Polarization

The negative sequence directional ground unit isoperated by the quantities V2 and I2. One typicaldesign operates when I2 leads V2 by 988. The output ofthe negative sequence directional unit D2 can be usedfor either supervising or the torque control of an I0 orI2 unit.

Negative sequence relays can be tested easily andquickly using load current flow. The directional unit ischecked by simply interchanging B and C currents andvoltages to provide negative sequence from thebalanced load quantities.

Negative sequence directional sensing has becomeincreasingly necessary because of a mutual inductionproblem. Refer to Section 3.5 in this chapter for moredetails on applying negative directional sensing relays.

3.4 Mutual Induction and Ground-RelayDirectional Sensing

Transmission lines on the same tower or parallel alongthe same right of way are mutually coupled. For apositive and negative sequence, mutual impedances are

less than 10% (usually they do not exceed 3 to 7%) ofself-impedances and can be considered negligible. Forzero sequence, however, mutual impedance can be 50to 70% of the zero sequence self-impedance Z0L and is,therefore, significant. Mutual impedance affects themagnitude of ground-fault currents and can result inincorrect directional sensing.

Figure 12-12 shows two parallel three-phase lineswith zero sequence isolation, except for mutualcoupling. The two lines can be completely isolated,or more commonly, tied together to common genera-tion sources. The lines can be at the same or differentvoltage levels.

A ground fault at or near G on line GH involves nodirectional sensing problems at either station G or H,as shown by the I0 current directional arrows. The 3I0current flowing from H to G induces a zero sequencevoltage in the parallel line RS, causing current to flowfrom R to S as shown. The polarizing and operatingquantities are properly oriented to operate the zerosequence quantity-polarized directional ground relaysat both R and S, so that this current appears as aninternal line RS fault to the relays at terminals R andS.

The zero sequence current I0M in line RS induced byuniform mutual coupling (Z0M) for length GH will be

I0M ¼ �K0I0ðnZ0MÞ þ ð1� k0ÞI0ð1� nÞZ0M

Z0SR þ Z0L þ Z0SS

¼ ½1� ðnþK0Þ�I0Z0M

Z0SR þ Z0L þ Z0SS

¼ DVSR

Z0SR þ Z0L þ Z0SSð12-16Þ

where

n¼ per unit fraction of Z00L from bus G to the

faultDVSR¼ induced voltage drop from S to R

As the fault moves from G to H, the induced currentI0M in line RS will decrease, reverse, and then increase

Figure 12-12 Mutual coupled transmission lines with zero

sequence isolation.


in the opposite direction. For example, when (nþK0)equals 1, I0M will be 0. If (nþK0) is greater than 1,then I0M reverses.

For some conditions and fault locations, it may bedifficult to set the overcurrent units to distinguishbetween faults on line GH and those on line RS.Mutual inductance will also cause incorrect directionalsensing on line GH for faults on line RS. A similaranalysis applies.

Correct directional sensing can be obtained forFigure 12-12 if the neutrals of the current transformerscan be paralleled at G and R, and at H and S as well. Apractical example of this situation would be parallellines terminating in transformer banks at each end withgrounded wye on the line side and delta on the bus. Ona per unit basis, the current flowing up the neutral onthe faulted line is always greater than the current,resulting from induction, flowing down the neutral.Hence, paralleling the current transformer provides anet current up the neutral for any fault.

Mutually coupled lines, with only one common bus(Fig. 12-13a), provide no directional sensing problemsin most of the conditions, as long as all breakers areclosed. The problems are as follows:

1. See Figure 12-13a. For example, zero sequencecurrent flow arrows are for a ground fault near or atbus G. The voltage drop between the zero potentialbus along line SG starting from the transformer at S is

I0SZ0SS þ I0HZ0M þ I0SZ0L � I0GZ0SG ¼ 0

Solve for I0S:

I0S ¼ I0GZ0SG � I0HZ0M

Z0SS þ Z0Lð12-17Þ

I0S will reverse and flow down the transformer bankneutral at station S if the mutually induced voltages aregreater than the drop across the bank at G. I0S thenwill reverse when

I0HZ0M > I0GZ0SG ð12-18Þ2. When one breaker is opened, zero sequence

isolation and incorrect directional sensing can occur.In Figure 12-13b, for example, a fault near G would becleared by breaker 1 relays. When breaker 1 opens, lineGS current flow reverses, and incorrect directionalsensing occurs at S.

3.5 Application of Negative SequenceDirectional Units for Ground Relays

3. A negative sequence directional ground unit can beconsidered in application for any one of the followingsystem configurations, as shown in Figure 12-14:

1. In Figure 12-14a, there are no high-side VT andno I0S source available for the relay polariza-tion.

2. Figure 12-14b cannot get 3V0 from high-sideVT; it is an open delta connection and/or thereis no I0S source available.

3. Figure 12-14c, an autotransformer neutralcurrent, is not a reliable polarizing quantity.

4. In Figure 12-14d, there is mutual inductionbetween parallel lines.

5. In Figure 12-14e, there is a mutually coupledtransmission line with one common terminatingbus.

3.6 Selection of Directional Overcurrent Phaseand Ground Relays

All of the desired functions are available in solid-stateversions, usually with curve shape selection inherent inthe relay. Multifunction microprocessor relays are soflexible they allow not only curve shape selection, butalso the choice of the type of polarization (includingdual). Electromechanical relays require a foreknowl-edge of the requirements of the power system beforethe equipment can be chosen.

Figure 12-13 Mutual coupled transmission lines with one

common terminating bus.

244 Chapter 12

3.6.1 Available Phase and Ground OvercurrentRelays

Nondirectional Phase and Ground OvercurrentRelays

Type Hi-Lo CO (single-phase, electromechanical) andtype MMCO (three or four elements, microprocessor-based) relays are suitable for this application.

Directional Phase and Ground Overcurrent Relays

The available electromechanical and solid-state direc-tional phase and ground overcurrent relays are asshown in Table 12-4.

3.6.2 Type of Relay Selection

The procedures for selecting a type of relay for eitherphase or ground protection are as below (refer to Fig.12-8):

Step 1 A directional relay should be used if

4ðI4max or IRLoadÞ �� ðI2min or I3minÞStep 2 An instantaneous relay can be applied and

set to underreach. If

I1max > 2:5 I2max

the unit can be set IT> (1.1 to 1.3) I2max and step3 should be checked.

Step 3 The instantaneous unit should be direc-tional if

I4max > I2max

3.6.3 Evaluation of Ground-Relay PolarizingMethods

Both zero sequence and negative sequence polarizingmethods require some evaluation to ensure thatsufficient operating quantities are available. Faultstudies provide the 3I0, 3V0, V2, and I2 values thatexist for various faults.

It should be noted that I0 is equal to I2 only for asingle-phase-to-ground fault on a radial circuit.

Typical profiles for V0 and V2 are shown inFigure 12-15. For remote faults, where nZ0L can bequite large compared to ZT, 3V0 for the relay at bus Gcan be quite low. Since negative sequence flowsthrough the transformer to the source, V2 can belarger than 3V0. The factor of 3 for both I0 and V0

helps, but it is still not uncommon for V2 to be largerthan 3V0 near a strong ground source.

Figure 12-14 Typical applications of negative sequence

directional ground relay.


The variation in negative sequence energy isconsiderably less than that for zero sequence. Also,the negative sequence energy will often be higher thanthe zero sequence energy for remote phase-to-groundfaults. The higher zero sequence impedance of lines, inconjunction with the remote ground source, highlyattenuates the zero sequence distribution to a remotefault.

Negative sequence directional ground relays arewidely used because they are not subject to polariza-tion reversals or mutual induction. Neither are theysubject to switching isolation, which produces mutualreversal; they require neither a transformer bankgrounded neutral nor zero sequence voltage transfor-mers.

Zero sequence generally provides higher relayoperating quantities for short- and medium-lengthlines, although values can be quite low for remotefaults on long lines. Current polarization is preferred at

Table 12-4 Summary of Available Independent Single-Function Directional Phase and Ground Relays (67/67N)

Fault-sensing unit operating quantities

Relay type

Directional unit

polarizing or

operating quantities

Directional controlled

time-overcurrent unit

Nondirectional

controlled

instantaneous unit

Directional controlled

instantaneous unit

CRa L-L voltage Yes Option

CRC I0 Yes Option

CRP V0 Yes Option

CRD I0 and/or V0 Yes Option

CRQ I2 and V2 Yes Option

IRVa L-L voltage Yes Yes

IRC I0 Yes Yes

IRP V0 Yes Yes

IRD I0 and/or V0 Yes Yes

IRD I2 and V2 Yes Yes

KRVa L-L voltage Yes

KRC I0 Yes

KRP V0 Yes

KRD I0 and/or V0 Yes

KRQ I2 and V2 Yes

CWC I0 Yes

CWP V0 Yes

CWP-1 V0 Yes

32b L-L voltage

32Db I0 or V0

32Qb I2 and V2

a Phase units.b Solid-state high-speed directional units that can be used to control overcurrent relay units such as MMCO, MICRO-51, 51, or 50D.

Figure 12-15 Typical voltage profiles for negative and zero

sequence voltages for ground faults.

246 Chapter 12

stations where there are ground sources, since V0 canbe very small. When there are several power trans-former banks in a station, the current transformers inall the grounded neutrals should be paralleled. Thismethod will avoid the loss of polarization if one of thebanks is removed from service. Zero sequence polar-ization from autotransformers should be used onlyafter a careful analysis.

Dual-polarized ground relays offer flexibility. Onerelay type can be used in various applications, ascurrent polarized only, potential polarized only, orboth, depending on system conditions.

4 DISTANCE PHASE AND GROUNDPROTECTION

Fault levels are usually high for high-voltage transmis-sion circuits and, if faults are not cleared rapidly, theycan cause system instability as well as extensivedamage and hazards to personnel. For these reasons,phase distance relays are generally used in place of thedirectional overcurrent relays, except at the lower-voltage levels. Even at the lower-voltage levels, thetrend is toward distance relays. For the higher-voltagelines, one or two pilot systems are used in conjunctionwith or as a supplement to phase distance relays.

The advantages of the application of a distancerelay in comparison to that of an overcurrent relay are

1. Greater instantaneous trip coverage2. Greater sensitivity (overcurrent relays have to

be set above twice load current)3. Easier setting calculation and coordination4. Fixed zone of protection, relatively independent

of system changes, requiring less setting main-tenance

5. Higher independence of load

4.1 Fundamentals of Distance Relaying

A distance relay responds to input quantities as afunction of the electrical circuit distance between therelay location and point of faults. There are manytypes of distance relays, including impedance, reac-tance, offset distance, and mho.

4.1.1 Operation of Distance Relays

Basically, a distance relay compares the current andvoltage of the power system to determine whether afault exists within or outside its operating zone. The

pioneer beam-type distance relay can be used toillustrate the operating principle. Consider a horizontalbeam pivoted in the center, with a voltage coil on oneend and current coil on the other. In Figure 12-16, therelay coils are connected to a power line throughinstrument transformers. Suppose a solid fault occurson the line at a distance of nZ1L O from the relay. Sincethe voltage at the fault is 0, the voltage VR on the relaywill be the IRZL drop from the relay to the fault. Thisvoltage provides a magnetic force or ‘‘pull’’ on one endof the beam. If, for this fault, the current or operatingforce IR on the other end of the beam is adjusted toequal the voltage or restraint force VR, the beam willbe balanced. That is,

VR

IR¼ IRnZ1L

IR¼ nZ1L ð12-19Þ

Should a fault occur between the relay and ZL

distance, say at (ZL�DZL) O from the relay, thenthe restraint force IR(ZL�DZL) would be less than theoperating force at the same current magnitude. As aresult, the beam would tilt down at the current end,closing the contacts. If the fault were beyond the ZL

distance, say at (ZLþDZL) O from the relay, then therestraint force VR would be greater than the operatingforce IR. The beam would then tilt down at the voltageend, and the contacts would not close. The expression

Figure 12-16 Beam type distance relay.


‘‘balance point,’’ which describes this threshold locusof operation for distance relays, is still commonly used,even though modern distance relays operate on quitedifferent principles.

In general, the operating torque T of a cylinder-typedistance relay is

T ¼ K1I2R �K2V

2R ð12-20Þ

At the threshold or balance point, T is 0. Then,

K1

K2¼ V2

R

I2Rð12-21Þ

and the reach, or ohms to the balance point, is

VR

IR¼ ZL

¼ffiffiffiffiffiffiK1

pffiffiffiffiffiffiK2

p ð12-22Þ

If delta voltage and delta currents are used, a set ofthree relays will provide operation by one or morerelays for all types of phase faults within the set-balance point impedance (ZL).

4.1.2 Application of Distance Relays

The major advantage of distance relays is apparentfrom Eq. (12-19) or Eq. (12-22). The relay’s zone ofoperation is a function of only the protected lineimpedance, which is a fixed constant, and is relativelyindependent of the current and voltage magnitudes.Thus, the distance relay has a fixed reach, as opposedto overcurrent units, for which reach varies as sourceconditions change.

Any line section in a power system can berepresented as shown in Figure 12-17. In this figure,ZL is the impedance of the line to be protected frombus G to bus H. ZS is the equivalent source impedanceup to bus G, and ZU the equivalent source impedanceup to bus H. ZE could be a parallel line equal to ZL,but, more generally, ZE represents the equivalence of

the interconnecting system between buses G and H,except for line ZL.

Figure 12-18 shows a simplified representation ofthe protected line for which distance relays are to beapplied at the bus G line terminal. The system can beplotted on an R-X diagram as follows. With G as theorigin, the phasor impedance ZL of the line is drawnto scale in the first quadrant. Either per unit or ohmscan be used, although secondary or relay ohmsare generally preferred. Modern distance relays arenormally connected to wye-connected current trans-formers,

Zsec ¼ Zrelay

¼ ZpriRC

RVð12-23Þ

where RC and RV are the ratios of current transformerand voltage transformer, respectively. ZS, the sourceimpedance, can be plotted from G into the thirdquadrant; at H, the source impedance ZU can beextended, both impedances at their respective magni-tudes and angles. ZS in this figure is assumed to be verylarge or infinite relative to the others. In applicationsinvolving several line sections, ZU would be the remoteline section beyond bus H; ZS would be the line sectionbehind the G line relay or to the left of bus G (if weassume there were no other lines or sources at eitherbus G or H).

4.1.3 General Characteristics of Distance Relays

A number of distance-relay characteristics plotted onthe R-X diagram are shown in Figure 12-19. The

Figure 12-17 General representation of a line section

between two buses.

Figure 12-18 (a) Representation of a line section; (b) the

R-X diagram.

248 Chapter 12

operating zones are inside the circles for the typeslabeled a, b, and c. That is, whenever the phasor ratioof V/I falls inside the circle, the distance unit operates.The beam-type distance relay, described earlier, wouldhave the nondirectional impedance characteristicshown in Figure 12-19a. When used for fault protec-tion, a separate directional unit is added to limit thetripping to line faults. Since the beam-type relay is nolonger manufactured, this characteristic is obtained byother techniques.

By modifying either the restraint and/or operatingquantities, the circle can be shifted as shown in Figures12-19b and 12-19c. The characteristics given in Figures12-19d and 12-19e can be obtained in the same generalway. There are a number of methods for obtainingthese characteristics, the details of which are beyondthe scope of this section. Appendix B of this chapterprovides more information for the distance-relaycharacteristics.

Load can be represented on these R-X diagrams asan impedance phasor, generally lying near the R axis(depending on the power factor of the load currenton the line). The phasor lies to the right (firstquadrant of the R-X diagram) when flowing into the

protected line from the bus and to the left (thirdquadrant of the R-X diagram) when flowing out ofthe line to the bus. Load is between 0 and 5Asecondary at or near rated voltage; faults generallyproduce much higher current levels and lowervoltages, so that the load phasor usually falls outsidethe distance operating circles. Since these conditionsdo not hold for the reactance type shown inFigure 12-19d, this unit cannot be used alone. Also,because the reactance unit is not directional, itwould, without supervision, operate for faults behindthe relay. The reactance unit then needs very carefulsupervision and is not a particularly desirablecharacteristic for most applications. A blinder char-acteristic (Fig. 12-19e) unit can be used for loadrestriction on impedance (Fig. 12-19a), modifiedimpedance or offset (Fig. 12-19b), and mho (Fig.12-19c) distance units.

The blinder characteristic, Figure 12-19e, is essen-tially two reactance-type units shifted to the lineimpedance angle. The right unit operates for a widearea to the left, and the left unit operates for a widearea to the right of the X axis. Together, the twoprovide operation in the band shown.

Figure 12-19 Distance relay characteristics.


With single-phase-type relays connected as phase-to-ground elements, those with one distance elementconnected to each phase, the R-X characteristic ofFigure 12-19 applies to only the unit on the faultedphase. For three-phase faults, load, and three-phasepower swings, all three single-phase distance relayshave a reach of nZ1L along the line (if arc resistance isneglected). All three will operate for any three-phasefault between the relay and nZ1L set point.

Three-phase-type relays respond to all phase faults,regardless of the specific phases involved. For example,the KD-10 relays have two operating units. One unitresponds to any three-phase fault between the relayand set reach of nZ1L; the other responds to any phase-to-phase fault (AB, BC, and CA) between the relay andset reach of nZ1L. One or both of the two units willrespond to all two-phase-to-ground (ABG, BCG, andCAG) faults from the relay to the set reach of nZ1L.

4.1.4 Characteristics of Mho-Type DistanceRelays

The operation of the mho-type distance relays on theR-X diagram is given in Figure 12-20. The three-phaseunit is a mho circle, where the locus at any point on theR-X diagram is

Zreach ¼ ZC � ZCffy2

ð12-24Þ

This defines a circle whose center is offset from G at

ZC/2 and radius is ZCffy=2, where ffy ¼ 0 to 3608.When y is 08, Zreach is 0 at the origin. When y is 1808,Zreach equals ZC, the forward reach or set point.

For the phase-to-phase unit, the reach at any pointis

Zreach ¼ ðZC � ZSÞ þ ðZC þ ZSÞffy2

ð12-25Þ

In Eqs. (12-24) and (12-25), the first term inparentheses is the center, and the second the radiusof the circle.

For the phase-to-phase unit, the circle does not passthrough the origin and varies with source impedance,except at the set point.

If Zs equals 0, then the characteristic is a mho circlethrough the origin, as Eq. (12-25) reduces to Eq. (12-24). When y equals 1808, a reverse reach is established.This function has no practical significance, since thecurrent reversal that occurs when the fault moves fromthe line side of the current transformers to the bus sidealways produces restraint (see Fig. 12-21 and 12-23)faults (1). As a result, the relay is directional, and theoperating area exists only in the first quadrant, withinthe area of the sector of the curve from ZC to the R axis.

When y equals 0, the terms containing ZS cancel, sothat Zreach equals ZC. Thus, the phase-to-phasecharacteristic is fixed at the set point and variable atall others. This variation poses no disadvantage, sinceneither load, power swings, nor any type of balancedconditions can produce operating torque. In otherwords, the operating and setting of the phase-to-phaseunit are completely independent of load and swings.

4.2 Phase-Distance Relays

4.2.1 The KDAR Phase-Distance Relays

The KDAR phase-distance relays, types KD-10 andKD-11, are three-phase, single-zone packages indesign. These relays provide mho-type characteristics,as shown in Figures 12-19c and 12-20. They consist ofa three-phase unit for three-phase fault detection and aphase-to-phase unit for two-phase fault detection. Theoperating units are a electromechanical cylinderelement. They use line-drop compensator theory thatproduces two phasor voltages, which are in phasewhen a fault occurs at the balance or reach set point.This condition produces no output. Faults inside thebalance point shift the voltages in a direction toprovide operation; faults beyond the balance pointproduce a shift in the opposite direction to providerestraint (Figs. 12-21 and 12-22).Figure 12-20 The K-DAR relay on the R-X diagram.

250 Chapter 12

Figure 12-21 illustrates the operation of the three-phase unit of the KD-10 relays. With three-phasepotential VA, VB, and VC applied with line current(IA� 3I0),

VX ¼ 1:5VA � 1:5ðIA � 3I0ÞZC

VY ¼ VB

VZ ¼ VC ð12-26Þ

The I0 component in Eq. (12-26) provides double-phase-to-ground fault coverage for systems with a verylow Z0s relative to Z2s. For these faults, the threesequence networks are connected in parallel. If, in thelimit, Z0s goes to 0, the negative sequence network isshorted out, leaving only the positive sequence net-work, as for a three-phase fault. For very low Z0s

systems, therefore, double-phase-to-ground faults‘‘look like’’ three-phase faults, and the three-phaseunit responds. The 3I0 helps cover these faults but doesnot enter into the phasors of Figure 12-21, which arefor three-phase faults.

The compensator is set so that ZC equals thepositive sequence line impedance from the relay to thebalance point (ZL). For the three-phase fault at the

balance (fault 3), VX terminates on the line between, Yand Z. The result is a zero-area triangle and nooperating torque on the cylinder unit connected to X,Y, and Z (see Chap. 3, ‘‘Basic Relay Units,’’ for adiscussion of the cylinder unit). Faults beyond thebalance point (4) and behind the relay (1) provide anXYZ triangle that produces opening torque. Faults (2)inside the balance point produce an XYZ triangle andoperating torque proportional to the area of thetriangle.

Close-in, solid three-phase faults produce an ABCtriangle with a very small area. That is, Y and Z wouldcollapse to the origin with very little or no operatingarea. To avoid this, memory action is provided bydelaying the collapse of Y and Z when B and Cvoltages approach or equal 0. The memory circuit, asshown in Figure 12-21, consists of a reactor and acapacitor in phase ‘‘c’’ of the voltage circuit and tunesthe oscillating voltage drop to the power-systemfrequency.

This delay is long enough to allow the instantaneous(zone 1) three-phase unit to operate and trip a breakerfor 0-V three-phase faults. For backup (zone 3) andother fault applications where continuous torque is

Figure 12-21 Phasor diagrams of KD-10 phase distance relays for faults at various locations (three-phase unit).


required, a different unit with added current-onlytorque is applied. This arrangement makes the unitsnondirectional for the heavy, close-in three-phasefaults. This is the KD-11 relay.

The operation of the phase-to-phase unit of theKD-10 phase-distance relay is shown in Figure 12-22.With three-phase applied voltages and current IA, IB,and IC from wye-connected current transformers,

VXY ¼ VAB � ðIA � IBÞZC

VZY ¼ VCB � ðIC � IBÞZC ð12-27Þ

The compensator is set so that ZC equals thepositive sequence line impedance from the relay to thebalance point (nZ1L). With a phase-to-phase fault atthe balance point, the XYZ triangle has 0 area for anAB, BC, or CA fault. Figure 12-22 shows the phasorsfor various BC faults. At the balance point, a BC fault(3) results in VZY being equal to 0. For BC faultsbeyond (4) or behind (1), a restraint triangle XYZ isproduced. An operating triangle XZY is produced forinternal BC faults (2).

The KD-10 phase-to-phase unit will also operate formost double-phase-to-ground faults. Together, the twounits (three-phase and phase-to-phase) provide com-plete coverage for all types of double-phase-to-groundfaults from the relay to the balance point setting. Withthe positive, negative, and zero sequence networks inparallel for these faults, the fault tends to ‘‘look like’’ aphase-to-phase fault when Z0s is large compared to Z2s.If Z0s goes to infinity, the double-phase-to-groundnetwork becomes equivalent to the phase-to-phasefault network.

Memory action is not required for this unit, andhigh torque exists for the solid, 0-V, phase-to-phasefault at the relay. As seen in Figure 12-22, when VBC

equals 0, VZY and VXY are large, providing a largeoperating XYZ triangle.

4.2.2. The Microprocessor-Based Phase-Distance Relays:

The microprocessor-based relay, type REL-300(MDAR) or REL-301/REL-302, is a numerical trans-

Figure 12-22 Phasor diagrams of KD-10 phase distance relays for faults at various locations (phase-phase unit).

252 Chapter 12

mission line protection systemwith three nonpilot zonesand one optional pilot zone of distance protection.

These relays use a single processor approach indesign. All measurements and logic in REL-300 andREL-301/REL-302 use microprocessor technology. Allthe distance units provide a mho-type characteristic, asshown in Figure 12-23. The zone 3 units may be chosento have a forward or reverse application.

Equations (12-28) and (12-29) show the phasors ofthe operating and reference (restraint) quantities forthree-phase and ff faults, respectively. The unit willproduce a trip output when the operating quantityleads the reference quantity.

Operating Reference

For three-

phase fault

VXG� IXZC VQ (12-28)

For phase-to-

phase faults

VAB� IABZC VCB� ICBZC (12-29)

where

VXG¼VAG, VBG, or VCG

IX¼ IA, IB, or ICZC¼ zone reach settings in terms of positive

line impedanceVQ¼ quadrature phase voltages, i.e., VCB,

VAC, and VBA for fA, fB, and fC,respectively

VAB, VCB¼ line-to-line voltagesIAB, ICB¼ delta currents (e.g., IA� IB)

4.2.3 The Microprocessor-Based Phase-DistanceRelay: Type REL-100

The REL-100 uses a multiprocessor design with threeprocessors for the basic distance measuring functionand up to three additional signal processors perform-ing the optional functions. The impedance unitsmeasure the apparent impedances of the fault loops.The resulting impedance is compared against reactanceand resistance limits determined by the relay setting.Either zone of the relay can be selected as reverse-looking. The quadrilateral characteristics with indivi-dual settings of reactive and resistive reach are asshown in Figure 12-24.

4.2.4 Microprocessor-Based REL-512

This microprocessor-based relay utilizes the full powerof digital techniques to identify whether or not a faultis within the reach of the distance unit. The samefundamental concept is used, in which two developedquantities as described by Eqs. (12-28) and (12-29) arecompared. Contrasted with previous methods thatdeveloped the equivalent phasor quantities for com-parison, REL-512 uses individual samples of voltageand current to determine fault location.

In the various measurements (phase fault, blinder,directional, etc.) quantities S1 and S2 are defined.In Eq. (12-28), for ‘‘a-phase,’’ for example,S1¼VAG� IAZC, and S2¼VCB. Operation occurswhen S1 phasor leads S2 phasor, with maximumsensitivity occurring when S1 leads S2 by 908.

Figure 12-25 shows the manner in which the REL-512 system determines fault location by using twoadjacent samples of two waveforms. The two quan-tities are similar to those previously described such as

Figure 12-23 REL-300 (MDAR) characteristics. Figure 12-24 REL-100 characteristics.


in Figure 12-22. In this example,

S1 ¼ VXY ¼ VAB � IABZC

S2 ¼ VZY ¼ VCB � ICBZC

Inputs of VAG, VBG, VCG, IA, IB, and IC allow this

digital algorithm to be implemented. Operation isproduced when S1 leads S2.

In Figure 12-25, y represents the angle by which S1leads S2. This quantity is negative when S2 leads S1. Krepresents the instant at which a sample is taken of S1and S2. K-1 is the point at which the previous samplesof S1 and S2 were taken. A and B are the peaks of thetwo waveforms.

S1 and S2 are shown as two arbitrary sine functions.S1k is the instantaneous value of the sample of S1taken at time k. S1 k� 1 is the value of S1 taken at theprevious sample. Similar values are taken of S2.

The purpose of this example is to show the ability toidentify which of two waveforms leads the other basedupon two adjacent samples of each of the waveforms.In the actual implementation, samples of all of thevoltage and current inputs are combined appropriately(along with acknowledgment of the various settings) toproduce the proper samples for comparison. This, inturn, allows the location of a fault to be identified aswithin or outside of the operate zone of the relay.

4.3 Ground-Distance Relays

4.3.1 Fundamentals of Ground-DistanceRelaying

The general representation of a line section, given inFigure 12-17, is expanded in Figure 12-26 for a line-to-ground fault at F. The equations of the relay currentand voltage at bus G are developed for phase ‘‘a.’’ Forthe phase a relay, consider the following combinationsfor a ground-distance relay.

1. Using zero sequence quantities 3V0 and 3I0, weget

Zrelay ¼ 3V0G

3I0G

¼ �I0P0Z0S

K0

ð12-30Þ

Unfortunately, this method is not useful. It measuresthe source impedance and current distribution factors,both of which are variable, rather than the protectedline impedance.

2. Using phase voltage and phase current. Thismethod depends on voltage compensation. By usingonly VAG and IAG, we obtain

Zrelay ¼ VAG

IAG

¼ ðK1I1 þK2I2ÞnZ1L þK0I0nZ0L

K1I1 þK2I2 þK0I0ð12-31Þ

Figure 12-25 Criterion for Lead-Lag Relationship.

254 Chapter 12

Again, the method is unsatisfactory.3. Using VAG modified by subtracting the positive

and negative sequence drop of (K1I1þK2I2)nZ1L, andI0, we get

Zrelay ¼ VAG � ðK1I1 þK2I2ÞnZ1L

K0I0

¼ nZ0L ð12-32ÞThis method is satisfactory and has been used.

4. With phase-to-ground voltage and modifiedphase current. This is the current compensationmethod. Assuming that nZ0L equals pnZ1L, we obtain

Zrelay ¼ VAG

IAG

¼ ðK1I1 þK2I2ÞnZ1L þK0I0pnZ1L

K1I1 þK2I2 þK0I0

¼ nZ1LK1I1 þK2I2 þ pK0I0

K1I1 þK2I2 þK0I0ð12-33Þ

By modifying the relay current IR to ðK1I1 þK2I2 þ

pK0I0Þ instead of IAG, we get

Zrelay ¼ nZ1L

IR ¼ K1I1 þK2I2 þ pK0I0

¼ K1I1 þK2I2 þK0I0 þ ðp� 1ÞK0I0

¼ IAG þ ðp� 1ÞK0I0 ð12-34Þ

where

p ¼ Z0L

Z1Land ðp� 1Þ ¼ Z0L � Z1L

Z1L

Substituting yields

IR ¼ IAG þ Z0L � Z1L

Z1L

� �K0I0 ð12-35Þ

This method is the one used in the KDXG, MDARand REL-512 ground-distance units. The completeformula, including arc resistance and the mutual effect

Figure 12-26 A line-to-ground fault on line section ‘‘GH’’.


of the parallel line, can be written as

Zrelay ¼ VAG

IR¼ nZ1L þRG

3I0

IRð12-36Þ

where

IR ¼ IAG þ Z0L � Z1L

Z1L

� �K0I0

þ Z0M

Z1L

� �I0E ð12-37Þ

In these formula,

RG¼ arc plus tower footing resistance and includesground wires, when used

Z0M¼mutual impedance to a parallel line where I0Ecurrent flows

I0¼ total zero sequence current, of which K0 I0 isthe portion through the relay

4.3.2 The KDXG Reactance Ground-DistanceRelay

The KDXG is an electromechanical cylinder unit,single-phase multizone relay. It permits three reactancezones of protection.

The operating principle of the reactance unit of theKDXG relay is similar to the description in Section4.3.1, Step 3, except the compensator setting jX is thereactance to the balance point, which equals thereactance part of nZIL in Eq. (12-38).

Figure 12-27 shows the phasor diagrams for thereactance unit of the KDXG relay for faults at variouslocations. The unit will produce contact-closing torquewhen the operating quantity Vop ¼ VL � jXIR lags thereference quantity Vref ¼ IR.

Voltage switching in the three single-phase reac-tance relays changes the reach through all three zones,in order. Switching is initiated by the zones 2 and 3timers.

Figure 12-27 Phasor diagrams for KDXG relay for faults at various locations.

256 Chapter 12

The reactance units are nondirectional and must besupervised by two units: (1) an external connecteddirectional (KRT or KDTG), and (2) an internal ratiodiscriminator, RD unit. The RD unit determines thefaulted phase and also provides the phase-to-phasefault blocking feature for the relay.

A terminal consists of three KDXGs (reactance andratio discriminator units) and one KRT or KDTG(directional unit and two zone timers). The basic tripand control circuits are illustrated in Figure 12-28.

4.4 Effect of Line Length

4.4.1 Zone Application of Distance Relaying

Historically, three zones of protection have been usedto protect a line section and provide backup for theremote section (Fig. 12-29). Each of the three zonesuses instantaneous operating distance relays. Zone 1 isset for 80 to 90% of the line impedance. Zone 2 isadjusted for 100% of the line, plus approximately 50%of the shortest adjacent line off the remote bus. Zone 3is set for 100% of both lines, plus approximately 25%of the adjacent line off the remote bus. These classicalsettings define the protective zones only if there are noinfeed effects. In practice, there is almost always aninfeed effect at the buses, which reduces the reach asdescribed later.

Since zone 1 (Z1) tripping is instantaneous, the zonemust not reach the remote bus, hence the 80 to 90%settings. The 10 to 20% margin provides a safetyfactor, for security, to accommodate differences orinaccuracies in relays, current, potential transformers,and line impedances. The 10 to 20% end zone isprotected by the zone 2 (Z2) relay, which operatesthrough a timer T2, set with one step of CTI(coordination time interval), as for overcurrent relays.Two zones at each terminal are required to protect allof the line section, with 60 to 80% of the line havingsimultaneous instantaneous protection. This protec-tion is independent of system changes and loading.

The backup zone 3 also operates through a timerT3, set as shown to coordinate with the zone 2 unit ofthe remote bus. Coordinating distance relays, withtheir fixed reach and time, is much easier thancoordinating overcurrent relays.

For directional comparison blocking pilot relaying,zone 3 is used to start the carrier. It must consequentlybe set with reverse reach, in the opposite direction tothe protected line section. T3 must be coordinated withrelays operating on the lines behind, rather than ahead.

An application of distance relays to parallel lines isshown in Figure 12-30. Here, the T2 settings on eachparallel line should be the same. Thus, the remote endsof both lines will trip if a fault in any end zone is notcleared by the nearby breaker and relay operation.

Figure 12-28 Basic trip and control circuit for the KDXG relay system.


Suppose that for fault F1 the H-end breaker does notclear as it should on zone 1. At G, both the top- andbottom-line zone 2 relays ‘‘see’’ this fault and operatein time T2. A similar operation would result if the faultoccurred at F2.

If both T2 settings were not the same, the unfaultedline may be cleared first. If T2 on the bottom line weregreater than T2 on the top line, for example, the top-line zone 2 relay at G would operate to clear the F1fault first. But if the fault were at F2, the top-line zone

2 relay at G would operate, even though the bottom-line zone 2 relay should clear this fault.

When several remote lines have different lengths,the zone 2 and 3 settings involve compromises (Fig. 12-31). Since line HV is short compared to lines HS andHR, setting zone 2 at G for 50% of line HV provides amaximum of 5.5% coverage for line HR and 8.4% forline HS. This coverage is further reduced by the infeedeffect. Additional coverage could be obtained byincreasing the G zone 2 setting and the corresponding

Figure 12-29 Step time zones of distance relay protection.

Figure 12-30 Distance relays on parallel lines.

258 Chapter 12

T2 setting to coordinate with the T2 times on lines HV,VW, and WX. The result would be long end-zoneclearing for line G. If pilot relaying is used for primaryprotection, increased backup with longer T2 timescould be employed.

Setting zone 3 to cover line HR would providecoverage through several sections HV, VW, WX, andXY requiring a longer T3 setting. Again, the infeedeffect from lines HS and HV probably would notprovide T3 coverage for line HR. This fact reempha-sizes the need for local backup in modern power systems.

4.4.2 Classes of Line Length

Transmission lines are often referred to as short,medium, or long, and relays are recommended on thisbasis. In reality, the length of the transmission line isnot a predominant factor. The significant criterion isSIR (source/line impedance ratio).

The SIR establishes the positive sequence voltagethat will be present at the relay location for a three-phase fault at the far end of the protected line. Whilesolid-state and microprocessor relays require littleenergy to operate, there is a voltage level below whichoperation is not clearly predictable or below which theoperating speed is unsatisfactory. For a distance relay,a fault at the balance point (the point to which therelay is set to reach) is the point at which the operatingvoltage for the relay is zero. For a fault closer to therelay, a positive operating voltage will exist. The lowerthe SIR, the greater the magnitude of this operating

voltage, and thus the more positively the relay willoperate (with the greater speed).

The IEEE/PES/Power System Relaying Committeehas chosen guidelines for line length criteria (IEEEStandard C37.113-1999). They have found the follow-ing to be reasonable:

Short line SIR> 4

Medium line 0.05< SIR< 4

Long line SIR< 0.05

For short lines, there may be little difference involtage and current at the relay location for faultsthat are a considerable distance from one another.The distinction between internal and forward externalfault locations may be difficult. This encourages, forshort lines, the use of relaying systems that establishfault location on the basis of current transformerlocation, such as current differential or phase com-parison.

Long lines allow excellent use to be made ofdistance relays alone or in pilot applications (see thecompanion volume, Pilot Protective Relaying).

For medium lines, many choices of relaying typesexist, including overcurrent or directional overcurrent.Other factors such as the relaying on adjacent linesections, presence of taps on the line, tripping speedrequired, channel availability, etc., must be considered.

Figure 12-31 Distance relays looking into a bus with various lengths of adjacent line sections.


4.5 The Infeed Effect on Distance-RelayApplication

4.5.1 Infeed Effect on Phase-Distance Relay

When there is a source of fault current within theoperating zone of the distance relay, its reach will bereduced and variable. This infeed effect can be seenfrom Figure 12-32, where there are other lines andsources feeding current to a fault at F from bus H. Therelays at bus G are set beyond this fault point to F0.With a solid 0-V fault at F, the voltage for the relay atG is the drop along the lines from the fault to the relay,or

VG ¼ IGZL þ ðIG þ IHÞZH ð12-37ÞSince relay G receives only current IG, the impedanceappears to be

ZG apparent ¼ VG

IG

¼ ZL þ ZH þ IH

IGZH ð12-38Þ

¼ ZL þ ZH

Kð12-39Þ

where K is the current distribution factor (phasor),which equals IG=ðIG þ IHÞ. This apparent impedancecompares to the actual impedance to fault F of

ZG actual ¼ ZL þ ZH ð12-40ÞIf IH is 0 (no infeed), Z apparent equals Z actual. Asthe infeed increases in proportion to IG, Z apparentincreases by the factor (IH/IG)ZH. Since this impe-dance, as ‘‘measured’’ by the distance relay, is largerthan the actual, the reach of the relay decreases. Thatis, the relay protects less of the line as infeed increases.Since the reach can never be less than ZL in Figure12-32 zones 2 and 3 provide protection for the line.

However, remote backup for the adjacent line (s) maybe limited. Since infeed is very common and can bequite large in modern power systems, the trend istoward local backup (see Chap. 13).

Note that the infeed effect varies with systemconfiguration and changes, and that the apparentimpedance may be a maximum under either maximumor minimum system conditions.

For example, for Figure 12-33, assume that the lineimpedances (relay ohms) are 2O from G to the tappoint, 8O from the tap point to H, and 2O from thetap point to R. The zone 1 relay at G is set to reach3.6O. A three-phase fault occurs at 1.0O from the tappoint toward H. Fault-current contributions (relayamperes) are 10.95A from bus G and 14.61A from R.The voltage drop from G to the fault is

2610:95þ 1625:56 ¼ 47:46Vðrelay sideÞSince the current through the relay at G is 10.95A,

Zapparent ¼ 47:46

10:95¼ 4:33O ðrelay sideÞ

The relay apparent impedance of 4.33O is higher thanthe relay actual setting of 3.6O, i.e., the relay will notoperate on this fault, i.e., underreach.

However, if the fault-current contribution is chan-ged to 11.56A from G and 11.56A from R, due to thesource impedances change, the voltage drop from G tothe fault would be

261:56þ 1611:56 ¼ 34:69V ðrelay sideÞSince the current through the relay at G is 11.56A,

Zapparent ¼ 34:69


the relay will see the fault. Hence, the reach of the

Figure 12-32 Effect of infeed on impedance measured by

distance relays. Figure 12-33 Typical infeed effect on a multiterminal line.

260 Chapter 12

distance relay varies as a function of fault-currentdistribution, as well as fault location.

4.5.2 Infeed Effect on Ground-Distance Relay

The descriptions and results in Section 4.5 cannot bedirectly used for ground-distance relay application.For ground-distance units, the operating current, e.g.,the phase A ground unit, is

IA þ Z0L � Z1L

Z1LI0

Therefore, when applying Eq. (12-38) the currents IHand IG should be replaced with

IH ¼ IH þ Z0L � Z1L

Z1L

� �I00 ð12-41Þ

IG ¼ IG þ Z0L � Z1L

Z1L

� �I0 ð12-42Þ

where IH and I00 are phase and zero sequence currentsin the tap, IG and I0 are phase and zero sequencecurrents at the relay. (For more detail, see Appendix Cto this chapter.)

4.6 The Outfeed Effect on Distance-RelayApplications

When a tap has no source except a tie line to a remotebus, fault current can flow out from this tap terminalfor an internal fault near the remote bus (Fig. 12-34).Although the fault is shown on bus H, it could be nearor at the breaker on the GH line. With no source at Rother than line RH, current flows out of R and overRH to the internal fault on line GH, reducing theapparent impedance.

For example (a) (Fig. 12-34a), assume that the lineimpedances (relay ohms) are 2O from G to the tappoint, 8O from the tap point to H, and 2O from thetap point to R, and the tie line between RH is 2O. Thezone 1 relay at G is set to reach 3.6O (90% of line GR).A three-phase fault occurs on bus H. Fault-currentcontributions (relay amperes) are 10.42A from bus G,3.47A from the tap point toward bus H, and 6.95Afrom the tap point out to bus R. The voltage dropfrom G to the fault along line GH is


Zapparent ¼ 48:6


The relay apparent impedance of 4.66O is higher thanthe relay actual setting of 3.6O, i.e., the zone 1 relay Gis not affected by this outfeed current.

However, in example (b) (Fig. 12-34b), if the lineimpedance from the tap point to R is 6O, the zone 1relay at G would have to be set to 7.2O (90% of lineGR). The fault-current contributions (relay amperes)would be 8.67A from bus G, 4.33A from the tap pointtoward bus H, and 4.33A from the tap point out tobus R. The voltage drop from G to the fault along lineGH is


Zapparent ¼ 51:78


The relay apparent impedance of 5.99O is less than therelay actual setting of 7.2O, i.e., the relay will operateon this fault, overreach. The setting must be lowered.

4.7 Effect of Tapped Transformer Bank on RelayApplication

The effect of a transformer bank tapped off a line mustbe considered, although it often presents no problem.The typical case is shown in Figure 12-35. The infeed

Figure 12-34 Typical outfeed effect on a multiterminal line.


fault current from bus H causes the relay at bus G to‘‘see’’ an apparent impedance for faults in thetransformer and low-side system circuits. As describedabove, for a fault at F, relay G sees

ZG apparent ¼ VG

IG¼ nZLIG þ ZTðIG þ IHÞ

IG

Since IF ¼ ðIG þ IHÞ and IG ¼ KIF, where K is thecurrent distribution factor for the current throughrelay G,

ZG apparent ¼ nZL þ ZT

Kð12-43Þ

Since K is always less than 1 when IH > 0, ZG apparent isalways greater than the actual impedance to the faultðnZL þ ZTÞ.

Similarly, relay H would see a higher apparentimpedance of

ZH apparent ¼ ð1� nÞZL þ ZT

1�Kð12-44Þ

The zone 1 relay at G must be set for 80 to 90% of thesmallest value of the actual impedances, ZL orðnZL þ ZTÞ, and not the apparent impedances. Other-wise, zone 1 may overreach either the transformer orbus H, which would result in miscoordination. WhenZT is less than ð1� nÞZL, relay G cannot protect asmuch of the line as it could without the tap. Usually,the tap bank is relatively small, so that ZT is largecompared to ZL.

Zone 2 must be set greater than ZL to protect theline. When ZT is less than ð1� nÞZL for the relay at Gor the setting for zone 2 is greater than ðnZL þ ZTÞ,then zone 2 requires coordination with the relays in thelow-voltage system. As long as H is in service, the

apparent infeed will shorten the reach. When H isopen, however, the infeed effect disappears.

A complex relationship affects the reach of single-phase-type distance relays through star-delta or delta-star banks. To these relays, a phase-to-phase fault onone side of the bank appears as a phase-to-groundfault on the other, and vice versa. Using the principledescribed in Figure 12-22, however, all phase-to-phasefaults on one side of a wye-delta or delta-wye bankappear to the relays on the other side to be a distanceZT away. In other words, the phase shift does notaffect their reach, as it does for single-phase relays.

The above discussion of the transformer tap on theline assumed no source of fault power on the low-voltage side. If such a power source exists, it producesan apparent impedance and relay underreach for linefaults: on the ð1� nÞZL section of the line for the relaysat G, and on the nZL section for the relays at H. Again,zone 1 relays G and H must be set for the actual, ratherthan apparent, impendances. Zones 2 and 3 must be setfor the maximum apparent impedance to cover the linesection. When the tap source is open, of course, thereach will be greater. This situation presents asignificant problem for multiterminal lines.

4.8 Distance Relays with Transformer Banks atthe Terminal

The reach of a distance relay is measured from thelocation of the voltage transformers; directional sensingoccurs from the location of the current transformers.Voltage and current transformers are usually atapproximately the same location for most applications.

4.8.1 Connections Using IH and VH Quantities

When the line includes a transformer bank without abreaker on the line side, there are several ways ofapplying distance relays (Fig. 12-36). Using IH and VH

for the line-distance relays is preferred, since the reachis a function of nZL only.

Figure 12-35 Apparent impedance for low-side faults on a

tapped transformer bank.

Figure 12-36 A line terminating in a transformer bank.

262 Chapter 12

4.8.2 Connections Using IL and VH Quantities

Alternatively, IL could be used with VH, which wouldalso make the reach a function of nZL only. In settingthe relay with high-side primary ohms nZL, the currenttransformer ratio RC of Eq. (12-23) must include theratio of the power transformer ½RC ¼ RCL 6ðKVL=KVHÞ� and ratio RVH of the high-side voltagetransformers. Taps on the power transformer willchange this ratio and, therefore, the reach. Unless therelay setting can be adjusted each time the taps arechanged, zone 1 must be set at 90% of the minimumsecondary ohms using ZL. Zone 2 must be set at morethan 100% of the maximum secondary ohms. The firstrequirement prevents zone 1 from overreaching theremote bus. The second requirement provides end-zone coverage.

If the power transformer is a wye-delta bank, eitherthe low-side current transformers must be connected indelta or auxiliary current transformers used. In thisway, the relay current will be equivalent to the wyecurrent that would be measured on the line or highside. Figure 12-37a shows the connections for the deltaon the high side, and Figure 12-37b those for the wyeon the high side.

The advantage of this arrangement is that,although distance is measured from VH and includesonly nZL, the relay will operate for some faults in thetransformer ZT. These faults fall within the nZL

setting, since the relay is directional from the CTlocation. The impedance for transformer faults isapparent: The voltage is a function of the currentfrom the remote end, while the current flows from thenear end. For the system shown in Figure 12-36,assume a solid three-phase fault of total value Ibetween the breaker and transformer, with K per unitflowing through the low-side current transformer and(1�K) per unit flowing from the far bus to the right.Then

VH ¼ ð1�KÞIZT;

and

Zapparent ¼ ð1�KÞIZT

KI

¼ 1�K

K

� �ZT ð12-45Þ

Normally, ZT would be larger than nZL for zone 1applications, limiting instantaneous protection. Zones2 or 3, however, could be on the order of ZT.

4.8.3 Connections Using IL and VL Quantities

A more common method is to use VL since it may notbe economical to provide high-side potential. It is moreconvenient with VL and IL to use primary ohms on thelow-side (VL) base. With single-phase-type zone-dis-tance phase (not for ground units) relays set throughwye-delta banks, it is necessary to shift both thecurrents and voltages to provide high-side quantitiesequivalent to those that would be measured at IH andVH. Otherwise, for high-side faults, the distance relaywould ‘‘see’’ a complex impedance, which would be afunction of the line transformer and source.

For the KDAR three-phase-type relays, however,this shift is not necessary. The conventional wye-currentand line-to-line voltage connections of IL and VL willprovide phase-fault protection in the trans-former and on the line. The reach is a function of ZT

and nZL only, a distinct advantage. Since the impedanceof the power transformers is known accurately, the zone1 relay should be set through the transformer bank as

ZC for zone 1 ¼ 0:99ZT þ 0:90ZL ð12-46ÞAgain, the power transformer taps must be taken intoaccount, as discussed above.

This method has the disadvantage of limiting lineprotection when ZT is large compared to ZL. Forexample, if ZT ¼ 10O and ZL ¼ 1O then Z1 is set for

9:9þ 0:9 ¼ 10:8O

Subtracting the ZT of 10O leaves only 0.8O of the lineprotected, or 80% rather than 90%, if low-side voltageVL were used.

Taps on the bank can change the value of ZT andthe reflected value of ZL such that setting zone 1 tonever overreach on any tap can result in very little orno line protection on another tap. This assumes thatthe settings are not changed with the taps.

Zone 2 should always be set for the maximumapparent ohms so as to protect the line for any tap.This may cause considerable overreach for other taps.

4.8.4 Connections Using IH and VL Quantities

The fourth possible arrangement is to use VL and IH.There is little advantage in this method, since the ohmsare still measured from the low-side bus. Directionalsensing would be from the line-current transformerlocation, rather than the bus-side current transformers.

When the line terminates in a transformer bank atthe remote end, zone 1 can be set into the bank toprovide 100% high-speed line protection. Only with the


three-phase KDAR type design relays can zones 2 and3 be set through the bank with an accurate balancepoint of ZL þ ZT þ ZLV system for all phase faults.With single-phase-type relays, the reach through wye-delta banks is variable.

Light internal faults in the transformers willprobably not produce enough variation in currentand voltage to operate the remote distance relays.Consequently, transformers should have individualprotection. Such protection does present a problem in

Figure 12-37 Connections for low-side ct’s to provide equivalent of high-side CT’s

264 Chapter 12

that remote breakers must be tripped to clear the faultwhen the local transformer relays operate. A transfertrip system should be used.

4.9 Fault Resistance and Ground-DistanceRelays

Two additional factors must be considered for groundfaults that are not present with phase faults: towerfooting resistance and ground wires. Tower footingresistance can vary from less than 1O to more than200O. This resistance term, multiplied by 3, must beadded to the relay reach equations. As describedbefore, the infeed to the fault from the remoteterminals further magnifies this fault resistance andcan cause the ground-distance relays to over- orunderreach. Although, in theory, this apparent reac-tance effect can be quite significant, relatively fewproblems have been encountered in many years of fieldexperience.

Overhead ground wires substantially reduce the linezero sequence impedance and tower footing resistancecomponent. To the relay, however, the effect isanything but a resistance component. For faultcalculations, the ground wire is assumed to be parallelwith the earth. In practice, its impedance is parallelwith the earth through the tower footing resistance ateach tower.

The effect of arc resistance is shown in Figure 12-38for a 15-mile, 138-kV line with a source at each end.Calculations were made assuming no angle betweenthe two voltage sources and, hence, no angle betweenthe current distribution factors. In practice, there isalways an angle between the current distributionfactors. The single phase-to-ground fault is assumedto occur at the balance point 90% of the line from busG. The zero sequence impedances were calculated inthe conventional manner, using r ¼ 100m-O. For the5000 ft on either side of the fault, however, the mutualand self-impedances of the ground wires were sepa-rated for each span, a tower footing resistance of 10 Oadded, and a modified zero sequence impedancecalculated.

The apparent impedance ‘‘seen’’ is given by Eq. (12-47) and plotted in Figure 12-38:

Z0L ¼ nZ1L þ 3ðRSG þRTFÞ

K1 þK2 þ Z0L

Z1LK0

¼ ZC þ Z0S þ Z0

TF ð12-47ÞFrom the study, Z0

TF is 2:8ff51�O. Although the ground

wires have reduced the tower footing effect from 10 to2.8 O, there is a significant angle that causes problemswith all types of distance-relay characteristics. Asignificant number of tree faults on EHV and HV lineshave shown that, when fault resistance is significant,distance relays are less sensitive than ground over-current relays and often will not respond properly.

4.10 Zero Sequence Mutual Impedance andGround-Distance Relays

Mutual induction from a parallel line will affect thereach of all ground-distance relays, but rarely causeserious problems. The mutual induction effect canbe studied from parallel lines bussed at both endsas shown in Figure 12-39. Systems with morecoupled lines and/or different terminating stations,although more complex, can be analyzed on thesame basis.

Figure 12-38 Example of effect of ground wires on tower

footing and arc resistances and on reach of distance ground

relays


Figure 12-39 also shows the zero sequence network,including mutual effect. An equivalent zero sequenceimpedance to the fault can be obtained by reducing thedelta to an equivalent wye and combining it with thetwo source impedances. By working back, the variousdistribution factors can be calculated. These factors areshown in Figure 12-39.

Consider the ground relays at breaker A. When theparallel line current flows in the same direction (C toD) as the fault current for fault F, the mutuallyinduced voltage is added to the faulted line voltage.This phenomenon causes an apparent impedancegreater than the line impedance. That is, mutualinduction causes the relay to underreach, unlesscompensated for by using the parallel line currentgiven in Eq. (12-37).

However, if the current in the parallel line flowsfrom D to C, the mutual effect is added to the flow ofthe faulted line current. The result is a lower apparentthan actual impedance, causing the relay to reachfarther. The current flowing in line CD is

P0 þ n� 1

2

� �I0 ¼ Z0U=ðZ0S þ Z0UÞ þ n� 1

2

� �I0

ð12-48Þ

When [Z0U/(Z0SþZ0U)þ n] is less than 1, the currentreverses and flows from D to C, causing overreach. Forzone 1, the critical area is that for faults around thebalance-point setting. For faults close to the relay at A,the current would flow from D to C. The effect,however, is of no importance. For end-zone faults at Bor on bus H, current will always flow from C to D.Consequently, zone 1 cannot overreach the end of theline because of the mutual effect.

When breaker B opens for faults at F, the currentthrough the parallel line flows from D to C, causing anoverreach of relay A. In this case, overreach isdesirable, since it will often cause zone 1 to operatesequentially for 100% of the line.

The response of ground-distance relays undervarious system conditions is given in Figure 12-40.The three curves for different values of K0/K1 showthe reach of relay A on line AB. The mutual inductioneffect is shown as a function of the P0/K0 currentdistribution ratio. Curves A to E are superimposed,showing the system constraints for various values of

Figure 12-39 A parallel line section and its zero sequence

network.

Figure 12-40 Zone 1 reach without mutual compensation

for the system of Figure 12-39.

266 Chapter 12

zero sequence source and line impedance relation-ships.

The circled points in Figure 12-40 show thesequential reach after breaker B opens. The regionbetween A and C represents the practical area ofoperation corresponding to zone 1 and reachesbetween 70 and 88%, when nominally set for 85% ofthe line. If there is no zero sequence source at bus H,then curve D shows a reach of 68% with K0/K1¼ 0.5.

Since mutual induction can reduce the reach of theground-distance relay, zone 2 should be set to providea minimum of 100% line protection. Figure 12-41illustrates the setting required for a variety of systemvariables. A setting of 150% will provide goodprotection, including an adequate margin for themajority of systems.

After breaker C opens, mutual induction can extendthe reach of zone 2 at A (Fig. 12-39) for faults at ornear C on line CD. The fault current flows from A to Band from D to the fault. This same condition can causezone 1 at D to extend its reach up to 100% of the line.If zone 2 is not set greater than 150% of the line, thereach of zone 2 at A will coordinate with that of zone 1at D.

The above discussion applies to most relays withoutmutual compensation. Although ground-distancerelays can use the parallel-line current to cancel themutual induction effect, this method is generally notrecommended, particularly for zone 1.

In summary, for the parallel-line cases shown inFigure 12-39:

1. Without mutual compensation, a zone 1 dis-tance relay set for 85% of the line will coverfrom 70 to 88% if the breakers are all in, andfrom 85 to 100% of the line after the far breakeropens.

2. For most applications, a zone 2 distance relaywithout mutual compensation will providecomplete end-zone coverage when set for150% of the line.

3. Compared with an uncompensated relay,mutual compensation usually increases zone 1coverage with the breakers all in, but decreasessequential coverage.

4. Mutual compensation must be used with cau-tion when ‘‘looking into’’ a weak source. Inthese cases, K0I0 flows from A toward B; andfor faults near C, the large mutual compensa-tion current from the parallel line can causemisoperation.

To use mutual compensation, the parallel line mustterminate in the same station in order to have itscurrent available. This cross-connection is complexand increases the possibility of an incorrect connectionor testing mistake.

If currents flow in the source direction in all thelines, additional parallel lines can cause greater relayunder-reach for end-zone faults. In this situation, therelay must be mutually compensated with all parallelcurrents. However, mutual compensation with three ormore lines is inordinately complex. If the compensa-tion of a given line is necessary, it should therefore belimited to the insertion of current from just one parallelline. Although this arrangement minimizes complexity,as long as mutual compensation is employed, over-reach hazards still exist.

5 LOOP-SYSTEM PROTECTION

5.1 Single-Source Loop-Circuit Protection

5.1.1 Using Directional Overcurrent Relays

A loop circuit with a single source is shown inFigure 12-42. For the purposes of the followingdiscussion, all breakers in this circuit will be consideredclosed during operation, at least for a significantamount of time. (Should a breaker open, the systembecomes radial.)

1. Relay application rules are similar to those forradial circuits. Typical uses of the phase overcurrent

Figure 12-41 Required zone 2 setting to provide complete

coverage of the protected line.


relays for the single-source loop circuit shown inFigure 12-42 are summarized in Table 12-5.

2. Nondirectional overcurrent relays can be usedat 1 and 10, since no current flows through theselocations for faults in the source system and on bus G.(This application is indicated by the double-arrow linebelow the breaker.)

3. At all other locations, fault current can flow ineither direction through the relays for faults to theright of bus G. These relays will operate when thepickup current is above the setting, but only if thecurrent (either load or fault) flows in the direction ofthe arrows, which, in each case, is into the line from thebus.

4. In practice, directional relays are usuallyapplied at all locations to accommodate any futuresystem changes.

5. For this single-source system, lines HG and TGmay be protected by directional instantaneous over-current relays that can be applied to 2 and 9 and setvery sensitively. Since the fault current through relays 2and 9 goes to 0 as the fault location reaches bus G, the

directional instantaneous overcurrent relays will notoverreach. For these same faults, however, the currentthrough relay 1 or 10 will be maximum. Either theinverse-time relays or, if applicable, instantaneous-tripunits will operate in minimum time. When breaker 1opens, for line HG faults, the current increases throughrelay 2 in the tripping direction. Thus, for line HGfaults close to bus G, breaker 2 opens sequentially afterbreaker 1. Similarly, for line TG faults close to bus G,breaker 9 opens sequentially after breaker 10.

6. In principle, the coordination procedure forthese relays is the same as for a radial circuit.Coordination is based on maximum fault current onthe remote bus. It must be assumed that the loop isopen at one end, because when the loop is openedbetween the fault and source bus, the current in anybranch will increase. Instantaneous-trip units, whenapplicable, must also be set on the assumption of open-loop conditions that yield maximum relay currents forremote end faults.

5.1.2 Using Inverse-Time and Distance Relays

With single-source loop circuits (Fig. 12-42), theinverse-time distance-relaying scheme offers distinctadvantages. The scheme is especially advantageous forlong loops with many sections, in which the relays atthe source end breaker would require a long time delayfor a far-end fault.

Figure 12-43, the inverse-time distance scheme,consists of a zone 1 distance relay (21) along with asimilar zone 2 distance relay (21) that torque-controls atwo-unit, inverse-time overcurrent relay (51). Thereach of the distance relay is independent of sourceimpedance variations; it can be set below and isindependent of load current.

Zone 1, set for 90% of the line impedance, protects amuch larger portion of the line than an instantaneousunit. Zone 2 is set through the next adjacent section to

Figure 12-42 A single source loop circuit and its protection.

Table 12-5 Phase Overcurrent Relays for Single-Source

Loop-Circuit Protection

Breaker

locations

Relays if

instantaneous-

trip units are

not applicable

Relays if

nondirectional

instantaneous-

trip units are

applicable

Relays if

directional

instantaneous-

trip units are

applicable

1 and 10 Inverse time Inverse time

with

instantaneous

trip

—

2 to 9 Directional

inverse

Directional

inverse with

instantaneous

trip

Directional

inverse with

directional

instantaneous

tripFigure 12-43 Distance controlled overcurrent scheme.

268 Chapter 12

provide end-zone and adjacent line backup protection.Since the two-unit overcurrent relay is torque-con-trolled, it will not operate unless the zone 2 relay hasoperated.

The principle of application is shown in Figure 12-43. Ninety percent of the line is tripped at high speed.For the remaining 10% of the line, the operating timeof the 51 unit can be made comparatively fast (equal toor less than the coordinating time intervals) sincecoordination is with the next zone 1 instantaneousrelay, rather than the time overcurrent unit.

The 51 unit may be set on a tap whose value is lessthan the maximum load current, since the controllingzone 2 distance unit will not operate on load current.This arrangement provides faster end-zone faults andbackup protection for the adjacent section thandirectional overcurrent relays with instantaneous-tripunits.

5.1.3 Protecting Loop Circuits with Tap

Circuits with load taps present coordination problemsif the tap impedance to its bus has the same order ofmagnitude as the impedance from the tap point to theremote line terminal. In such cases, time overcurrentrelays must be coordinated with the protection at andbeyond the tap bus, as well as with that at the remotebus and beyond.

With the inverse-time distance scheme, the zone ofinstantaneous protection can be quite limited if the tapis near the relay terminal, particularly if zone 1 must beset so as not to operate for faults protected by the fuse(Fig. 12-44). Generally, zone 1 would be set into butnot through the transformers. This arrangementimposes no limitation as long as ZT is greater thanZL. Fast reclosing and the subsequent lockout of zone1 permit the fuse to clear transformer faults.

The zone 1 setting must be made on the basis ofactual ohms, since the infeed effect disappears if theloop is open at some point. Also, the zone 1 reach mustremain short of both the remote end and the low-voltage bus.

If zone 2 reaches through the transformer, the 51relay must coordinate with the fuse and low-sidebreakers. For faults in and on the low side of thetransformer, the current from the remote end tends tomake the distance relay underreach. The zone 2 settingmay also be based on actual ohms unless the zone-2-controlled 51 relays are acting as backup to thetransformer secondary main-breaker relays. In thiscase, the effect of far-end contribution requires that the

zone 2 setting be determined on the basis of totalapparent impedance including the transformer.

5.2 Multiple-Source Loop Protection

5.2.1 Using Directional Overcurrent Relays

In general, directional relays are necessary to protectloop circuits with multiple sources. The coordinationproblems for such systems are very complex andfrequently require compromises in protection. Imaginea system that is similar to the one shown in Figure 12-42, except that there are sources at each of the buses(instead of only at bus G). In this case, relays 1 and 10must be directional, and two coordination loops willexist. Relay 1 must be coordinated with 3, 3 with 5, 5with 7, 7 with 9, and 9 with 1. Similarly, relay 10 mustbe coordinated with 8, 8 with 6, 6 with 4, 4 with 2, and2 with 10. In addition, each relay must be coordinatedwith any other circuits or loops connected to thatrelay’s remote bus.

These multiple-source loops tend to close on eachother, so that there is no specific starting point in thecoordination procedure and the last relay inevitablydoes not properly coordinate with the first relay. Atrial-and-error process is necessary to adjust thesettings. This laborious procedure must be performedfor both maximum and minimum operating condi-tions, as well as the various lines in or out of service.

A good starting point for the coordination processis at the largest sources, where the close-in fault currentwill be large and the relay time correspondingly short.Many computer coordination programs have been

Figure 12-44 Protection of line with fused transformer tap.


developed and used. They are invaluable aids in thisprocedure.

For illustration, an example for coordination on amultiple-loop system is described in Appendix D ofthis chapter.

5.2.2 Using Inverse-Time Distance Relays

As discussed in Section 5.1.2, inverse-time distancerelays can provide high-speed protection for 90% ofthe line section, as well as improved protection for theremaining 10% of the line section and adjacent lines.

The relays are used in the same way as described inSection 5.1.2. If the protected line is tapped, the relayapplication and settings should be modified asdescribed in Section 5.1.3.

In multiple-source systems, backup protection maybe limited, since the fault-current infeed from theremote bus to adjacent line faults can reduce the reachof the zone 2 distance relays. The infeed also affects thefault levels for overcurrent relays. Nevertheless,inverse-time distance relays are generally easier toapply, permit more sensitive settings of the inverse-time overcurrent units (settings below load), andimprove fault clearing times. Selecting an inverse-time characteristic compatible with the characteristicsof other similar relays in the system simplifiescoordination.

6 SHORT-LINE PROTECTION

6.1 Definition of Short Line

Many problems associated with so-called short linesare actually related to the SIR value. SIR, the sourceimpedance ratio, is the ratio of the source impedanceto the line impedance. As the SIR value increases, sodo application complexities. A 10-mile line with a lowSIR value may be considered a ‘‘long’’ line, whereas a100-mile line with a high value may face many of theproblems associated with ‘‘short’’ lines.

6.2 Problem Associated with Short-LineProtection

In the past, many short lines were protected by acurrent-only scheme. Today, distance relays are beingused more than ever before. The difficulty associatedwith short-line protection is the zone 1 overreachproblem. Overreach may be caused by the followingfactors:

1. Current and/or voltage transformer inaccuracy2. Ratio of the source impedance to line impe-

dance3. Relay sensitivity4. Voltage transformer transient problem

6.3 Current-Only Scheme for Short-LineProtection

A current-only scheme does not require voltageinformation, and this greatly reduces the complexityof protection. Also, it provides better coverage on arcand fault resistance than the distance schemes. How-ever, most of the time, the application of current-onlyschemes on a short line may be a problem in terms ofdirectionality, and also the application is a function ofthe SIR. Table 12-6 shows that the coverage is limitedas the source impedance is increased. Table 12-6’s dataare based on the following assumptions, which can becalculated from Eq. (12-2):

Source voltage ¼ 69:3V

Line impedance ¼ 0:5O

IF1; IF2 ¼ faults at 0 and 100% of the line section;

respectively

Instantaneous� trip unit setting ¼ 1:3 ðIF2Þ

6.4 Distance Relay for Short-Line Protection

6.4.1 Sensitivity

In general, as the reach setting of a distance relay isreduced, the fault current required to operate the relayincreases and the operating time of the relay alsoincreases. Therefore, for short-line application, morefault current is required for the relay operation.

Table 12-6 Three-Phase Fault Currents

Zs

(O) SIR

IF1(A)

IF2(A)

Instantaneous-

trip unit

setting

Coverage

(O) (%)

0.5 1 138.6 69.3 90.1 0.27 53.8

1.0 2 69.3 46.2 60.1 0.15 30.6

1.5 3 46.2 34.6 45.0 0.04 8.0

2.0 4 34.6 27.7 36.0 No No

270 Chapter 12

For many short-line applications, the sources arequite strong and the fault currents high; however, forthose systems with high source impedance ratios, thecurrent sensitivity of the relay must be considered.

6.4.2 Effect of Source Impedance Ratio

Table 12-7 shows the per unit values of VR, the faultvoltage at the relay location, and VR-IRZC, the relayoperating voltage, for the simple system of Figure12-45 with a fault applied at 85% of the relay reach.The relay is set for 90% of the line, i.e., ZC¼ 0.9ZL.

The relay operating quantity, VR-IRZC, becomesvery small when the SIR value increases. For example,on SIR¼ 30, the voltage that can be applied to therelay VR and the relay operating quantity are less than3 and 0.5%, respectively. When the signals are so small,any error in the voltage or current can be substantialrelative to the theoretical values.

6.4.3 Current Source

Current transformers used with distance relays shouldnot saturate for faults occurring at the balance point.Limited saturation for faults inside the operating zonepresents no problems for the relays, as long as the relaycurrent is not reduced or shifted enough to cause theimpedance phasor to fall outside the operating zone.However, since this determination requires a complexcalculation, it is desirable to use good-quality currenttransformers for short-line applications.

The dc offset component of the fault current maycause a transient overreach for a distance relay. Thistendency is particularly important for zone 1 applica-tions. If the compensators of the distance relay are air-gap-type transformers, the transient overreach will benegligible.

6.4.4 Potential Source

Most conventional voltage transformers are adequatefor use with distance relays. The subsidence transientof some capacitor voltage devices, CCVT, requires aspecial setting consideration or an added time delay forstatic zone 1 relay applications.

As the source impedance ratio increases, the faultvoltage at the relay location decreases. The loweraccuracy of the potential source at this lower voltagesmay limit the usefulness of a zone 1 unit in short-lineapplications.

The transients associated with the capacitive cou-pling devices used to obtain the line voltage for therelay systems have been known to cause problems instatic relay applications. In general, there are twoareas, directionality and transient overreach, where thesubsidence transient behavior of the CCVT cansignificantly affect static relay performance.

1. Directionality The transient error of a CCVTis greatest for zero-voltage faults because anyoutput is pure error. Therefore, the most severecondition for the directionality of a distancerelay would be a zero-voltage reserse fault.Transiently, the voltage seen by the relay is afunction of the design of the particular CCVT,the fault initiation angle, the fault location, andthe burden connected to it. A typical waveformis shown in Figure 12-46. For this case, thepolarity of the voltage from the CCVT is out ofphase with the prefault voltage in the secondhalf-cycle after the fault occurs. For a distancerelay, a trip output is produced when theoperating signal (VR-IRZC) is out of phasewith the polarizing signal Vpol. For a relaywithout memory action, this can result in areversal of the polarizing quantity relative to theoperating quantity (VR-IRZC) that will result ina misoperation. The use of memory action inthe polarizing circuit will insure proper direc-

Table 12-7 Fault and Relay Operating Voltages for System

in Figure 12-45

SIR VR VR-IRZC

0.25 0.754 0.133

1.00 0.433 0.076

10.00 0.071 0.012

30.00 0.025 0.004

100.00 0.007 0.001

Figure 12-45 Sample system for short line relaying.


tional action for this case if it can ride over theCCVT transient error. However, misoperationmay still occur, if the magnitude of the IR ZC

signal is less than that of the CCVT transient.This misoperation is caused by the phasereversal of the IRZC operating signal, not thepolarizing signal; therefore, the use of cross-polarization or memory voltage will not preventmisoperation. This problem is most evidentwhen the magnitude of the IR ZC signal in therelay is small. Thus, it can occur when the faultcurrent IR is low and/or the reach of the relayZC is small.

2. Transient overreach In general, all relaysinclude memory action for a short time afterthe fault occurrence. This memory action isimportant to the performance of a mhodistance relay regardless of the transient under-or overreach. As the location approaches theend of the line, the magnitude of the operatingsignal VR-IRZC approaches 0. As the magni-tude of VR-IRZC is reduced, the effect of anyerroneous voltage, such as the CCVT transient,is increased. If the output voltage is momenta-rily lower than the true value, the relay mayoverreach. For example (Fig. 12-45), with anexternal fault F2 on the remote bus with zone1 set for 90% of the line, the current andvoltage, as well as the Vpol and VR-IRZC relaysignals, are as shown in Figure 12-47. Notethat in the second half-cycle after the start ofthe fault, the CCVT transient has caused theVR-IRZC signal to reverse polarity with respectto the polarizing voltage. This is the operatingcondition for the unit. The problem is mostevident when the magnitude of the IRZC signalin the relay is small. This can occur when the

fault current is low and/or the reach of therelay is small. It can be the situation when therelay is applied to a high source impedanceratio condition.

6.4.5 Arc/Fault Resistance

As the source impedance ratio increases, the voltagedrop in the arc becomes a significant percentage of thevoltage at the relay location. With a source impedanceratio of 100, the arc voltages are greater than—severaltimes over—the voltage drop in the line. The apparentarc impedance seen by the relay is thus greater than theline impedance, again several times over.

Because the magnitude of the apparent arc impe-dance is greatest at high source impedance ratios, theperformance of both phase- and ground-distancefunctions will be similarly affected under thoseconditions.

The infeed from another source will not affect themagnitude of the apparent arc impedance seen by therelay, but the angle of the apparent impedance. Thechange in angle of the impedance could cause adistance function to overreach, or underreach, orhave no effect at all, depending on the design of therelay.

If faults involve ground, the total resistance in thefault is composed of arc resistance plus faultresistance that can be very large depending on thecomponents involved in the fault. For example, a treefault, or a fault to ground through a fire, can have avery large resistance component relative to arcresistance. The effect of infeed on this component isto magnify the resistance, as well as shift it in phaseangle.

Figure 12-46 Typical CCVT transient voltage for a zero

voltage fault. Figure 12-47 Distance relay transient overreach caused by

subsidence transient voltage.

272 Chapter 12

7 SERIES-CAPACITOR COMPENSATED-LINEPROTECTION

7.1 A Series-Capacitor Compensated Line

Transmission lines are inherently inductive. Thepurpose of a series capacitor is to tune out part or allof the transmission-line inductance. In a networkwithout series capacitors, faults are inductive incharacter and the current will always lag the voltageby some angle. Commonly used types of line protec-tion can detect a fault and by operating circuitbreakers clear it fast and selectively. With the seriescompensation of the transmission line, capacitiveelements are introduced, and the network will nolonger be inductive under all fault conditions. Thedegree of this change is dependent on the line andnetwork parameters, extent of series compensation,type of fault, and fault location.

The capacitive or apparent capacitive nature of thefault current may cause the line protection to fail tooperate, or to operate incorrectly, unless carefulmeasures are taken to acknowledge this problem.Due to the capacitive nature of the fault loop, acomplication with respect to protection may arise bothon the compensated line as well as adjacent lines.

Series-capacitor banks are equipped with spark-gaps that bridge the capacitor and often with metaloxide protective devices. The spark-gaps are set toflash over at a voltage two to three times the nominalvoltage of the bank. When the spark-gaps flash over,the network is restored to an inductive nature. In spiteof this, the protection complications remain. Spark-gaps flash instantaneously when breakdown voltage isreached, but following a fault, time is required to reachthis level and some faults will not cause the gaps to fireat all. The time to gap-flashing is often longer than theoperating time of high-speed line protection. The effectof capacitive reactances must be evaluated even forfaults that flash the spark-gaps. Adding to thesecomplications is the fact that transients are generateddue to the presence of the series capacitor at theoccurrence of the fault, as well as at the instant ofspark-gap flashing.

The use of metal oxide devices in parallel with theseries capacitor introduces another element of concern.These units are never removed unless they themselvesare jeopardized. Their level of conduction is approxi-mately 1.5 times the rated peak voltage of the seriescapacitor. When voltage in excess of their conductionlevel appears across the metal oxide device, theirimpedance is reduced markedly, causing the series

capacitor to be partly bypassed. However, when thevoltage decreases to a level below the threshold, theimpedance of the device becomes very high, and thecapacitor is effectively reinserted. This action providesanother level of transient generation, but, in general, itcauses a softer impact on protective relaying thansimple spark-gaps.

The metal oxide devices are bridged with triggeredspark-gaps to limit the energy generated in the deviceduring fault conditions. Therefore, the protectiverelays then must be able to handle the effect of boththe metal oxide device and spark-gaps.

7.2 Relaying Quantities Under Fault Conditions

The effect of series compensation on transmission-lineprotection depends on the location of the capacitorsand degree of compensation. Figure 12-48a is anexample of a one-line diagram of a series capacitor andtransmission line. Figure 12-48b is the steady-stateR-X diagram. Because of the fact that the capacitorbypass protective equipment may be conducting ornot, the apparent impedance as viewed from locationA for a fault at B may appear vastly different.

Figure 12-48 Apparent impedance as viewed from station

‘‘A’’ for a fault at B.


Figure 12-49 illustrates the influence of a nearbyseries-capacitor bank. Faults nearer the capacitor-line junction as viewed from location A will have avery large negative reactance character. This nega-tive reactance is actually due to a reversal of thevoltage at the relaying point or, under certainconditions, reversal of the current through the seriescapacitor.

Voltage reversal occurs at the bus if the negativereactance of the series capacitor is greater than thepositive reactance of the line section to the faultlocation. Current reversal occurs if the negativereactance of the series capacitor is greater than thesum of the source reactance and line reactance tothe fault location. Figure 12-50 depicts this condi-tion.

‘‘Current reversal’’ or ‘‘outfeed’’ can also occur insome applications where a fault is at the capacitor-line junction and a parallel line exists between busesA and B. Figure 12-51 describes for a typical case thevariation of voltage to be expected at variouslocations in the power system. Note that there is novoltage inversion in this case. ‘‘Current inversion’’occurs at 2. Current at 4 also falls in the directionopposite to that for the same case without seriescapacitors. Whether line-side or bus potentials areused for the relays makes no difference in establishingthe direction for this fault.

As can be seen in Figure 12-51, the zero voltagepoint in the system can be moved farther back as aresult of multiple lines contributing to a fault near thecapacitor-line junction. The negative reactance of thecapacitor is enlarged compared to the positive reac-tance of the adjacent lines. This can result in zerovoltage occurring on lines that are located far awayfrom the series capacitor. The voltage can be 0 only ina network with negligible resistances. In a realnetwork, the remaining voltage is so small that it canbe regarded as 0.

Figure 12-49 Apparent impedance at 60Hz under fault

conditions.

Figure 12-50 Voltage and current reversal.

274 Chapter 12

7.3 Distance Protection Behavior

Series compensation of a network will affect thedistance protection on both the compensated line andadjacent lines connected to buses where a voltagereversal can occur. Generally, the most severe pro-blems occur with the relaying associated with theadjacent line.

The following problem areas can be identified:

Determination of direction to a faultLow-frequency oscillationTransients caused by flashing of bridging gapsTransfer of capacitor reactance to resistance by a

metal oxide element bridging the capacitorZone reach measurementFalse voltage zeros

There will be difficulties with distance protection indetermining the correct direction of a fault in a stationwhere a voltage reversal can occur. When directpolarization (polarization voltage from the faultyphase) is used, the protection on both the faulty andhealthy lines may see the fault in an improperdirection. This false determination of direction will

take place with both mho relays and plain directionalelements.

To overcome this and achieve correct directionalmeasurement, polarization quantities from the healthyphases are utilized. Healthy phase quantities will notbe reversed and a correct directional measurementachieved for all unsymmetrical faults for an unlimitedtime.

Cross-polarized mho relays would under some faultconditions overtrip for faults on adjacent lines becauseof the use of a single comparator for both the directionand reach measurement, and therefore additionalmeasuring criteria are required.

In the case of the three-phase fault, where all phasevoltages reverse, only memorizing the prefault polariz-ing voltage can correctly determine direction. Nor-mally in distance protection, memory voltage is usedonly when the voltage is reduced to some percentage ofthe nominal voltage. These criteria cannot be usedwhen a voltage reversal occurs. The use of memoryvoltage must be controlled by general nondirectionalthree-phase fault criteria.

The time the memory voltage can be used must belimited to approximately 100 msec. Today, memorycan be made very accurate, but in the case of a three-phase fault, the prefault condition should only beextrapolated for a limited time after the fault. Thenetwork is in a changing state and will run out ofsynchronism with the memory. Therefore, directionalmeasurements have to be sealed in after the time thememory becomes unreliable.

The transient caused by flashing of bridging gapswill jeopardize the security of the relaying system.Also, line-energizing transients are high frequency incharacter and could cause some relays to operate. Toavoid unwanted tripping, low-pass filtering of themeasuring quantities is necessary.

The problems above require that bandpass filteringbe used on the measuring quantities. The requirementof bandpass filtering exists in all distance protection,but is much more pronounced in applications invol-ving series-compensated networks to avoid unwantedoperation.

With an increase in current through the capacitorbank and an increased ‘‘conducting angle’’ of theparallel metal oxide element, the capacitive react-ance will start to diminish and the combination willhave a resistive component as seen in Figure 12-48.When setting impedance relays on the compensatedline, allowance for this apparent resistance isnecessary to assure tripping at all fault-currentlevels.

Figure 12-51 Typical voltages and currents for fault at

capacitor-line function.


7.4 Practical Considerations

The diversity of problems associated with the intro-duction of series capacitors to transmission lines makesthe selection of relaying systems and principlesdifficult. Voltage-related problems are the main reasonfor which current-only systems, like REL-350, arepreferred for protecting lines equipped with seriescapacitors.

For high-speed relaying of a series-compensatedtransmission line, the use of a pilot system isunavoidable. If directional comparison systems areused, distance elements provided with an acceptableduration of ‘‘memory’’ and very special logic should beavailable. A reverse-looking unit with memory actionis used to block high-speed tripping.

If phase-comparison systems are used, like REL-350, a relaying channel is required to transmit theinformation of the currents from side to side such thata phase comparison of the currents could take place.The simplicity of the logic is attractive, and the well-proven concept has an advantage over directionalcomparison systems with extra logic. Directional andphase-comparison systems are inherently dependent onthe channel.

In REL-350, provisions have been made to includeas backup both time-delayed zone 2 and zone 3distance units. Both zones are composed of onephase-to-phase unit and three phase-to-ground units.

The phase-to-phase unit, described before and in theappendix, is inherently directional and will sense faultsin the forward direction only. Therefore, the seriescapacitor(s) will always be in the protective zoneregardless of a flashing-gap state.

The phase-to-ground units have their limitations.The units have a forward and reverse reach that isselectable. The unit is described in the appendices. Themain purpose of the intentional reverse reach is toinclude negative reactance in the operating area of theunit.

These zone 2 and zone 3 protective zones beingnondirectional, the time delay introduced into the unitsshould coordinate with any step-distance zone outsidethe transmission line. Generally, zone 2 should notoverreach zone 1 of any adjacent line as zone 3 shouldnot overreach zone 2 if possible.

The apparent impedance to the distance relaydepends on the state of the surge protective devicesin parallel with the series capacitor. It is important fora zone 2 application that the transmission line betotally covered. For this reason, zone 2 and zone 3settings should be calculated for a totally uncompen-

sated line. This way, the settings will make sure thatthe line is fully covered under all operating conditions.Figure 12-52 illustrates the coverage of a zone 2application.

The reverse-reach characteristic of the groundunits enables the proper coverage of the negativereactance of the series capacitor as shown inFigure 12-52. It is unfortunate that the unit is notinherently directional since it will operate for reversefaults; however, for a time-delayed backup zone, thisis acceptable.

In applying stepped-distance protection to a series-capacitor environment, coordination may be a pro-blem since the reach of the zones depends on theconducting state of the surge protective equipment forthe capacitor. Some compromises will result from theapplication and they may include the time delaying ofthe compensated line protective zones to also time-coordinate with adjacent lines.

8 DISTRIBUTION FEEDER PROTECTION

For the purpose of relay application, a feeder isconsidered to be radial if, at a particular relaylocation, the maximum backfeed (fault current in the

Figure 12-52 Series-capacitor line protection.

276 Chapter 12

nontrip direction) is less than 25% of the minimumfault current for which the protective relay mustoperate.

8.1 Relay Coordination with Reclosers andSectionalizers on a Feeder

Figure 12-53 shows a typical feeder circuit using acircuit breaker, recloser, sectionalizers, and fuses. Thethree reclosures of breaker G should be time-delayedto allow clearing of faults beyond recloser H. The firstreclosure can be instantaneous, however, if theinstantaneous-trip units of the relays can be set shortof H, and the reclosing relay can lock out subsequentinstantaneous-trip operations after the first reclosure.The recloser at H can be set for either one or twoinstantaneous reclosures; the other two or three shouldbe time-delayed.

Reclosures are circuit-interrupting devices, similarto circuit breakers, that include automatic tripping andreclosing facilities. Normally, there are four reclosuresbefore lockout: one instantaneous and three adjusta-ble, time-delayed. Three types of controls are used:series trip, relay trip, and static control. Series-trip andstatic-control reclosers have adjustable time character-istics over a wide range of minimum pickup and curveshapes. To simplify coordination with protectiverelays, the relay-trip recloser can be equipped withany of the inverse and instantaneous-time overcurrentrelays. Whenever possible, relays with the same typesof time characteristic should be used. Adjustment ofthe time curve shape for the reclosers usually simplifiescoordination with relays. There may be problems,however, in cases requiring coordination with relays,reclosers, and fuses, in that order.

Sectionalizers are usually single-pole devices, whichdo not have fault-current-interrupting capability but

can sectionalize a distribution feeder during a perma-nent fault.

The sectionalizer is opened by an integrator that, inturn, is operated by the fault-current pulses resultingfrom the initial fault and subsequent opening andreclosing cycle of a recloser, or by the reclosing ofcircuit breakers ahead of the recloser. The integratorcounts the number of current pulses and opens thesectionalizer after the count has reached a preset valueand the circuit is dead.

Sectionalizers simplify the coordination of reclosersand fuses, since, for currents above the recloser’sminimum trip, the sectionalizer can be set to open forany 0 current point in the reclosing cycle. Thissequence ensures that the fuse is not subject to anyadditional fault current.

8.2 Coordinating with Low-Voltage Breaker andFuse

Low-voltage breakers, used for circuits of 600 V andbelow, have built-in solid-state overcurrent trip devicesthat actuate a solenoid trip mechanism. Trippingenergy is obtained from the primary fault current,rather than a separate station battery. The time-current characteristics of these devices may be differentfrom those for the time-overcurrent relays.

As shown in Figure 12-54, there are in some casesfour different characteristics: a long delay phase withan inverse characteristic (top right section of thecurves), short delay phase (middle section), phaseinstantaneous with no intentional time delay (lowerright section), and ground (left curve). The settings arecontinuously adjustable and can be easily set andtested in the field with a portable test set. An operatingband rather than a curve is used to describe theircharacteristics.

A typical application of these breakers at a 480-Vsecondary unit substation and load center is shown inFigure 12-55. A typical application with coordinationcurves is shown in Figure 12-56. The key systemcurrent data are plotted first as an aid in coordinat-ing and setting the various devices. These are (1)motor starting, (2) transformer full load, (3) trans-former magnetizing inrush estimated at 8 to 12 timesthe full load at 0.1 sec, (4) the three-phase faultcurrent (19,600- and 4800-A) backfeed from themotor loads, and (5) the total load center busthree-phase fault current (20,000A). This exampleassumes several motor feeders, although only one isshown. Circuit breakers C and E coordinate with B.Figure 12-53 Typical distribution feeder protection.


On double-ended substations where a normally opentie breaker (E) is used, the incoming (C) and tie (E)breakers have duplicate settings, except for the short

time delay. The incoming line breaker (C) is set tooperate one time interval longer than the tie breaker(E).

Figure 12-54 Low voltage air breaker time characteristics (type DS).

278 Chapter 12

The primary protection for the high-voltage side ofthe secondary unit substation normally is furnished byhigh-side current-limiting fuses. As illustrated inFigure 12-56, with a high capacity source, thecurrent-limiting fuse is current-limiting for faults aslow as 33% of a maximum fault. The fuse curves (D)provide very adequate phase-fault protection, but notfor ground faults. These are normally restricted andnot isolated by a fuse.

Figure 12-55 Distribution at a typical secondary unit

substation and load center.

Figure 12-56 Protection and coordination for a typical secondary unit substation and load center.


This fuse protection is very inadequate for trans-former secondary faults located between the transfor-mer and main secondary breaker (C). The maximumsecondary fault would take approximately 1 sec toblow the fuse. For a more probable 50% fault, the fusewould take between 15 to 20 sec. This illustrates that ahigh-side fuse is very good for primary maximumphase-to-phase and three-phase faults, but very inade-quate for the more probable high-side ground faultsand restricted secondary faults.

The best protection and coordination with thesecondary-unit substation secondary protection areprovided by a step-time characteristic that approachesthe low-voltage breaker characteristic very closely(Fig. 12-56). The relay consists of a long time-overcurrent unit and two instantaneous-trip units,one of which operates through a timer that can beadjusted from 0.25 to 3.0 sec. Normally, the IT unithas a range of 10 to 40A; the second unit (IIT) has arange of 20 to 80A. As illustrated in Figure 12-56, therelay will recognize a transformer secondary faultdown to 25% of the maximum value and operate in0.6 sec.

Extremely inverse relays could be applied and set asshown as an alternative. Usually, it is very difficult tocoordinate this characteristic between the breakers’time interval (CTI). Long operating times will occurfor light secondary faults as compared to the step-timeprotection

In areas of high load density, the trend insecondary-unit substations is toward larger and fewerunits to transform the distribution voltage down to autilization voltage of 600 V or less.

Associated with the larger units are higher inter-rupting requirements that may necessitate much larger-frame breakers than the load requirements dictate.When larger-frame breakers are required, smaller-frame integrally fused circuit breakers may be econom-ically applied.

A fused breaker is a standard breaker with specialcurrent-limiting fuses (limiters) to extend the upperlimit of interrupting capability to possibly 200,000symmetrical rms A. This breaker-fuse combinationmay consist of integral or separately mounted appa-ratus.

The current-limiting fuse restricts the peak ‘‘let-through’’ current on the first cycle to a value that iswithin the air breakers’ interrupting capability. Thislimiting is illustrated in Figure 12-57. Figure 12-58shows the instantaneous peak currents for various fuseratings as a function of the available short-circuit rmssymmetrical current and system x/r ratio of 6.6. Note

that the peak let-through values obtained from thecurves take into account the 1.414 factor from thepeak-to-rms ratio and a factor of 1.62 for themaximum offset effect. The total ratio is, therefore,1.4146 1.62¼ 2.29.

Data on the use of current-limiting breakers andminimum recommended fuse sizes are available frommanufacturers. Loads to which these breakers areapplied are protected against single phasing by aninterlock device that trips the breaker when any onefuse blows.

Figure 12-57 Peak let through current of a current limiting

fuse.

Figure 12-58 Characteristics of current limiting fuses for a

system X/R ratio of 6:6.

280 Chapter 12

APPENDIX A: EQUATION (12-2)

Let

n¼ per unit of line section length protected bythe instantaneous unit

Ki ¼ instantaneous unit pickup current; IIT

maximum far-end fault current; IF

SIR¼ source impedance ratio

¼ source impedance; Zs

protected line impedance; ZL

Since

IF ¼ 1

ZS þ ZLIIT ¼ 1

ZS þ nZL

Ki ¼ IIT

IF¼ ZS þ ZL

ZS þ nZL¼ SIRþ 1

SIRþ n

Solve for n:

n ¼ SIRð1�KiÞ þ 1

Kið12-2Þ

APPENDIX B: IMPEDANCE UNITCHARACTERISTICS

B.1 Introduction

There is no topic in protective relaying more challen-ging and interesting than high-speed relaying usingdistance concepts. It is also true that the operation ofdifferent impedance units remains a mystery for mostengineers and has generated several misconceptions onimpedance relay applications. In most cases, as itshould for any application, the characteristics of thedistance elements are not an issue. Most of us tend todisregard the influence that particular system para-meters have on the performance of the units.

The purpose of this appendix is to illustrate theperformance and characteristics of the different dis-tance elements found in ABB relays. More thanstressing the operating characteristics of the units, thereader will notice that the equations and derivations tobe presented are nothing more than academic exercises.Perhaps, as will be described, the simple mathematicalmodels of the units are overshadowed by the superiorperformance of the distance elements in real life.

This appendix will introduce step by step thevarious factors influencing the performance of distanceelements. Symmetrical components analysis is impor-tant and will be used throughout the development ofthe equations.

For a good understanding, the basic idea of acomparator is introduced first and a general proceduredescribed. Some of the common units in ABB relayswill be described as an illustration of the use ofcomparators. Finally, a brief discussion of derivedcharacteristics will complement the contents of thisappendix.

B.1.1 Basic Idea of a Comparator

Phasors are fundamental quantities in the analysis ofac systems. A comparator is a design element used inrelays to compare two phasors either in magnitude orphase. A distance relay will always have a phasecomparator or magnitude comparator regardless of thetechnology used, i.e., electromechanical, solid-state,and microprocessor-based relays.

Protective relaying is a binary science; either it is a‘‘go’’ or a ‘‘no go’’ condition. The diverse units used inany discipline of protective relaying determine that thesystem is normal or abnormal. A comparator will givethe relay system an output when the conditions foroperation are satisfied. Since phasors are expressed inmagnitude and phase, there are two types of compara-tors: phase and magnitude.

Phase Comparators

In Figure 12B-1, given two arbitrary phasors, S1 andS2, the output of a phase comparator is a logic ‘‘1’’ (thecomparator has operated) if

S1

S2¼ Me+j90� ðB-1Þ

as a limiting condition, and

ð�90�Þ < ffS1 � ffS2 < ðþ90�Þ ðB-2Þ

Figure 12B-1 Phase comparator concept.


if we define the operating characteristic of the phasecomparator, as shown in Figure 12B-2.

The quantity

M ¼ jS1jjS2j

is any arbitrary magnitude for the condition to be metand it does not affect the operation of the phasecomparator.

The characteristic angle of +908 determines asymmetric characteristic. This does not mean, how-ever, that other limits have not been used, as will bedescribed later.

Magnitude Comparators

Given two arbitrary phasors, SA and SB, the output ofa magnitude comparator is a logic ‘‘1’’ (the comparatorhas operated) if

SA

SB¼ Ce jr ðB-3Þ

This is a boundary condition, and

jSAj > jSBj ðB-4Þdefines the operating characteristic of the magnitudecomparator (see Figs. 12B-3 and 12B-4). The quantityC is a constant and r any arbitrary angle. Generally, Chas a value of 1.

Relationship Between a Phase and MagnitudeComparator

The characteristics of a phase comparator andmagnitude comparator are totally different. However,the inputs to a magnitude comparator, SA and SB, can

be related to those to a phase comparator, S1 and S2,so that the outputs of both comparators are equiva-lent. The relationships between (SA and SB) and (S1and S2) are

SA ¼ S1 þ S2 ðB-5ÞSB ¼ S1 � S2 ðB-6Þ

Equations (B-5) and (B-6) imply that a phasecomparator with inputs S1 and S2 will provide thesame output as a magnitude comparator if the inputsSA and SB have the values shown above. Refer toFigure 12B-5.

It is also true that if an equivalent phase comparatoris to be derived from a magnitude comparator, theFigure 12B-2 Phase comparator operating characteristics.

Figure 12B-3 Magnitude comparator concept.

Figure 12B-4 Magnitude comparator operating character-

istics.

282 Chapter 12

following relationships are equivalent:

S1 ¼ ðSA þ SBÞ2

ðB-7Þ

S2 ¼ ðSA � SBÞ2

ðB-8Þ

B.1.2 Generalized Use of Phase Comparators

Phase comparators are used widely in distance relaydesigns. The input phasors are generally a combinationof voltages and currents. From these inputs, the ratioV/I, or impedance, is proportional to the distance tothe fault, and it is indeed the quantity of interest.

Most distance relays do not measure Z¼V/Idirectly, but the operating characteristic of the R-Xdiagram is derived from the characteristics of thecomparator.

In general, the inputs to a phase comparator willhave the following format:

S1 ¼ k1Vþ k2I ðB-9ÞS2 ¼ k3Vþ k4I ðB-10Þ

where V and I are the voltage and current of interest toderive Z, or the unit characteristic of the R-X diagram.Constants k1, k2, k3, and k4 are design constants thatmay be complex and introduce a phase shift.

In this section, a general procedure to derive theimpedance characteristic of the comparator on theZ¼V/I plane will be developed. The procedure will beused later to derive a variety of distance unitcharacteristics used in ABB relays.

Using Eq. (B-1), (B-9), and (B-10) for a phasecomparator, we get

S1

S2¼ Me+j90�

¼ k1Vþ k2I

k3Vþ k4I

¼ k1

k3

ðV=IÞ þ ðk2=k1ÞðV=IÞ þ ðk4=k3Þ

¼ k1

k3

Zþ ðk2=k1ÞZþ ðk4=k3Þ ðB-11Þ

In most applications, k1 and k3 are real numbers;therefore, Eq. (B-11) can be simplified, as shown in

Figure 12B-5 Equivalent operation of a magnitude and a phase comparator.


Eq. (B-12), for any value of M1:

Z� a

Z� b¼ M1e

+j90� ðB-12Þ

where

a ¼ � k2

k1ðB-13Þ

b ¼ � k4

k3ðB-14Þ

The quantities a and b are vectors and do have thesame units as Z, i.e., they are impedances as well. Ingeneral, a and b will be sufficient to define theoperating characteristics of the unit.

Equation (B-12) defines the operating characteristicof the phase comparator. If a number of impedances Zcould be found on the Z¼V/I plane that satisfy Eq.(B-12), an impedance locus can be determined, definingthe operating and nonoperating regions of thecomparator in the R-X diagram.

Vectors a and b are fixed, but can be thought asreference points in the R-X diagram.

B.1.3 Generalized Use of MagnitudeComparators

Conceptually, magnitude comparators, as well as thegeneral procedure to derive the comparator character-istic on the R-X diagram, are simpler than those for aphase comparator.

If the inputs SA and SB are expressed in terms ofimpedances, either the power-system or relay settingimpedances, then by using Eq. (B-3), with C¼ 1, theoperating characteristic of the comparator in the R-Xdiagram can be defined.

B.2 Basic Application Example of a PhaseComparator

The basic steps outlined above will be illustrated in thissection. For this purpose, let V and I be the voltageand current input to the relay and the inputs to thephase comparator be

S1 ¼ V� ZcI ðB-15ÞS2 ¼ V ðB-16Þ

where Zc is the relay setting. From Eq. (B-9) and(B-10), it follows that k1 ¼ 1, k2 ¼ �Zc, k3 ¼ 1, and

k4 ¼ 0. Using Eqs. (B-13) and (B-14), we get

a ¼ Zc ðB-17Þb ¼ 0 ðB-18Þ

For the purpose of illustration only, different con-straints will be analyzed.

1. For �90� < ffS1 � ffS2 < þ90�,

Z� a

Z� b¼ M1e

+j90� ðB-19Þ

Referring to Figure 12B-6, we see that an infinitenumber of impedance vectors could be found thatsatisfies Eq. (B-19). The requirement is that theprojection of the vector from a to Z, (Z� a), satisfythe +90� angle difference to the vector from b to Z,(Z� b), and its projection. In Figure 12B-6, two Zvectors have been projected that satisfy the anglerequirement. Notice that the � 908 requirement is metto the right of Zc, and the þ 908 requirement to the left.A characteristic angle of + 908 and the S1 and S2inputs define the popular mho impedance locus. Noticethat the plot is symmetric.

2. A characteristic angle of less than 908 (Fig.12B-7) distorts the mho circle and modifies it to acharacteristic that is called ‘‘tomato’’ due to itssimilarity to the fruit.

The characteristic is really composed of two circles:one for ffS1 � ffS2 ¼ þa, and the other forffS1 � ffS2 ¼ �a. The centers of both are displacedaway from and perpendicular to the middle of Zc. Thecharacteristic in Figure 12B-7 is for a ¼ 45�.

3. A characteristic angle greater than 908 (Fig. B-8)distorts the mho circle and modifies it to a ‘‘lens.’’

Again, the characteristic is composed of two circles:one for ffS1 � ffS2 ¼ þa, and the other forffS1 � ffS2 ¼ �a. The centers of the circles are on aline perpendicular to the middle of Zc. The character-istic in Figure B-8 is for a ¼ 135�.

The conditions discussed above are practical andhave been used for many purposes in relaying.Although a simple description of the procedure hasbeen presented where a¼Zc and b¼ 0, these constantscan take on different values.

A practical phase comparator that when modifiedcan provide the characteristics just discussed isillustrated in Figure 12B-9.

Basically, the timer T determines the characteristic.For 60-Hz systems, 4.16msec equals 908 of the power-system cycle. Therefore, a coincidence of at least4.16msec determines the familiar mho characteristic.

284 Chapter 12

B.3 Basic Application Example of a MagnitudeComparator

Let V and I be the voltage and current input to therelay, and the inputs to the magnitude comparator be

SA ¼ 2V� IZc ðB-20ÞSB ¼ �IZc ðB-21Þ

where Zc is the relay setting.For a magnitude comparator, the characteristic is

determined by

SA

SB¼ e jy

2V� IZc

�IZc¼ e jy

or

V

I¼ Z ¼ ðZc � Zce

jyÞ2

ðB-22Þ

The characteristics of this magnitude comparator are

determined by Eq. (B-22) and plotted in the R-Xdiagram of Figure 12B-10. The diagram illustrates thecenter of Zc/2 and radius of Zc/2. That defines thecharacteristic mho circle.

B.4 Practical Comparator Applications inDistance Relaying

The above discussion has helped us in understandingthe use of comparators for distance relaying. The R-Xdiagrams above corresponded to the Z ¼ V=I plane, orthe impedance seen by the relay due to the voltage andcurrent inputs. It is generally accepted that the R-Xplots are representative of the positive sequence lineimpedance to the fault (Z11). Most distance unitsoperate on a combination of voltages and currentsfrom the power system and have a definite purpose inoperating for certain types of faults only. Distributionfactors and load flow that influence the differentrelaying units will be reviewed, and the response ofimpedance units to different types of faults described.

Distance relays are used in the detection of phasefaults (phase-to-phase and three-phase) and ground

Figure 12B-6 R-X plot for ffS1 � ffS2 ¼ 90�.


Figure 12B-8 R-sX plot for ffS1 � ffS2 ¼ 135�

Figure 12B-7 R-X plot for ffS1 � ffS2 ¼ 45�

286 Chapter 12

faults (single-phase-to-ground). Detecting phase-to-phase-to-ground faults is achieved by the operationof the phase and/or ground units, depending on thezero sequence impedance (Z0) of the system. Anotheruse of comparators is for the familiar blinder lines forout-of-step detection.

B.4.1 List of Symbols to Be Used

Throughout the remainder of this section, the follow-ing list of symbols and abbreviations will be used:

Vs, V0s Source voltages

VA, VB, VC Line-to-ground voltages at the relaylocation

VAm, VBm, VCm Prefault line-to-ground voltagesat the relay location (memoryvoltage)

VA1, VA2, VA0 Phase A positive, negative, and zerosequence voltages at the relaylocation

VB1, VB2, VB0 Phase B positive, negative, and zerosequence voltages at the relaylocation

VC1, VC2, VC0 Phase C positive, negative, and zerosequence voltages at the relaylocation

VF1, VF2, VF0 Phase A positive, negative, and zerosequence voltages at the faultlocation

IA, IB, IC Line currents at the relay location inthe tripping direction

IA1, IA2, IA0 Phase A positive, negative, and zerosequence currents at the relaylocation in the tripping direction

IF Total fault currentIF3ph Three-phase fault currentIL Prefault load flowZ1s, Z0s Positive and zero sequence source

impedancesZ01s, Z

00s Positive and zero sequence source

impedancesZ001s, Z

000s Positive and zero sequence source

and line impedancesZ1l, Z0l Positive and zero sequence line

impedancesZ01l, Z

00l Positive and zero sequence line

impedancesZc Relay impedance-reach settingZ1c, Z0c Positive and zero sequence relay

impedance-reach settingsZ1cF, Z0cF Forward-positive and zero sequence

relay impedance-reach settingsZ1cR, Z0cR Reverse-positive and zero sequence

relay impedance-reach settingsX1c, X0c Positive and zero sequence relay

reactance-reach settingsZ001l, Z

000l Positive and zero sequence reverse-

looking impedanceZRs Zero sequence to positive sequence

source impedance ratio Z0s/Z1s

ZR00s Zero sequence to positive sequence

source impedance ratio Z000s/Z

001s

ZRl Zero sequence to positive sequenceline impedance ratio Z0l/Z1l

ZRc Zero sequence to positive sequencereach ratio Z0c/Z1c

Figure 12B-9 A basic phase comparator.

Figure 12B-10 Mho unit derived from a magnitude

comparator.


ZRcF Zero sequence to positive sequenceline forward-reach ratio Z0cF/Z1cF

ZRcR Zero sequence to positive sequenceline forward-reach ratio Z0cR/Z1cR

XRc Zero sequence to positive sequencereactance-reach ratio X0c/X1c

ZR00l Zero sequence to positive sequence

impedance ratio Z000l/Z

001l

PANG Positive sequence line impedanceangle

RT Blinder settingS1, S2 Phase comparator inputsSA, SB Magnitude comparator inputsK1 Positive sequence current

distribution factorK2 Negative sequence current

distribution factorK0 Zero sequence current distribution

factork1, k2, k3, k4 General constants of a phase

comparator equationsa Complex operator, eþj120

a, b Reference vectors

B.4.2 Phase-to-Phase Unit

The operating characteristic of this unit can be betterunderstood and derived using the magnitude com-parator concept. It is primarily used to detect phase-to-phase faults, for which Figure 12B-11 applies, for aradial system. It has been used in relaying systems suchas the KDAR (KD, KD-4, KD-10, KD-41, KD-11,etc.), Uniflex (LKD), LDAR (LZM, LDM, LDMS,LZ, etc.), and MDAR (ff unit). The inputs to themagnitude comparator are

SA ¼ VA2 � IA2Zc ðB-23ÞSB ¼ VA1 � IA1Zc ðB-24Þ

SA is composed of negative sequence quantities only,and SB positive sequence quantities only, for whichthis unit is not responsive to load or out-of-stepconditions.

The magnitude comparator operating characteristicis defined by Eq. (B-3)

SA

SB¼ VA2 � IA2Zc

VA1 � IA1Zc¼ e jy ðB-25Þ

Using Figure 12B-11, we obtain

SA

SB¼ IA1Z1s � ð�IA1ÞZc

½IA1ð2Z1l þ Z1sÞ� � IA1Zc¼ e jy

or, in the positive sequence R-X plane, the unit’scharacteristic equation is

Z1l ¼ ðZc � Z1sÞ2

þ ðZc þ Z1sÞ2

e�jy ðB-26Þ

Equation (B-26) defines the locus of impedances (Z1l)for the magnitude comparator to change its output.Figure B-12 illustrates the R-X diagram of the unit onthe positive sequence impedance (Z1l) plane.

B.4.3 Three-Phase Unit

This unit can be better understood by using themagnitude comparator concept. Its main purpose is todetect three-phase faults; therefore, the simple

Figure 12B-11 Sequence network for a BC fault.

Figure 12B-12 Phase-to-phase unit operating characteristic

on the Z1l plane.

288 Chapter 12

sequence network connection of Figure 12B-13applies. It has been used in the KDAR (KD, KD-4,KD-10, KD-41, KD-11, etc.) and Uniflex (LKD)relaying systems. The inputs to the magnitudecomparator are

SA ¼ � Zc

2

� �IA1 ðB-27Þ

SB ¼ VA1 � Zc

2

� �IA1 ðB-28Þ

Since for a three-phase fault all quantities are positivesequence, the operating characteristic of the unit isdefined by

SA

SB¼ �ðZc=2ÞIA1

VA1 � ðZc=2ÞIA1¼ e jy

¼ �ðZc=2ÞIA1

Z1lIA1 � ðZc=2ÞIA1¼ e jy

Then,

Z1l ¼ Zc

2

� �� Zc

2

� �e�jy ðB-29Þ

Equation (B-29) defines the locus of impedances (Z1l)for the magnitude comparator to operate. Figure 12B-14 illustrates the R-X diagram of the unit on thepositive sequence impedance (Z1l) plane.

B.4.4 Ground Units

The implementation of phase-to-ground units hasbeen the most difficult. The impedance characteristicsof these units have been derived for phase-A-to-ground faults and, again, the R-X diagrams aredetermined on the positive sequence impedance (Z1l)plane.

For the purpose of generality, since a few groundunits will be analyzed, Figure 12B-15 defines thequantities of interest for a radial system.

The units to be analyzed in more than one wayrequire a positive sequence impedance setting and zerosequence impedance setting, although most of the time,this requirement is not evident at first glance.

It is also worth emphasizing that the R-X plots to bederived are based on the positive sequence lineimpedance

Z1l ¼ VA1 � VF1

IA1

Type SDG, SDGU, and LDG Ground Units

The principle has been used in the SDGU family (SDGand SDGU) and Uniflex (LDG) relays. It can be betterunderstood and its characteristics are defined in amagnitude comparator.

With the quantities defined in Figure 12B-15 for aphase-A-to-ground fault, the inputs to the magnitudecomparator are

SA ¼ VA0 � IA0Z0c ðB-30ÞSB ¼ ðVA1 þ VA2Þ � ðIA1 þ IA2ÞZ1c ðB-31Þ

For a magnitude comparator, the characteristics areFigure 12B-13 Sequence network for a three-phase fault.

Figure 12B-14 Three-phase unit operating characteristic on

the Z1l plane.


defined by

SA

SB¼ VA0 � IA0Z0c

ðVA1 þ VA2Þ � ðIA1 þ IA2ÞZ1c

¼ e jr ðB-32ÞFrom Figure 12B-15, the following relationships are

true:

VA1 ¼ VS � IA1Z1s

¼ IA1ð2Z1l þ Z1s þ Z0s þ Z0lÞ ðB-33ÞVA2 ¼ �IA1Z1s ðB-34Þ

Therefore, simplifying Eq. (B-32), we obtain

SA

SB¼ �IA1Z0s � IA0Z0c

½IA1ð2Z1l þ Z0s þ Z0lÞ� � 2IA1Z1c

¼ e jr ðB-35Þ

The relay settings are chosen, such that

ZRl ¼ ZRc ¼ Z0c

Z1c¼ Z0l

Z1lðB-36Þ

Therefore, Eq. (B-35) can be simplified to

Z1l ¼ 2Z1c � Z0s

ð2þ ZRlÞ �ðZ0s þ ZRcZ1cÞ

ð2þ ZRlÞ¼ e�jr ðB-37Þ

which defines the impedance characteristic of thecomparator on the Z1l plane. Figure 12B-16 illustratesthe characteristic. Notice that the reach Z1c is fixed forr ¼ 180�.

Quadrature-Polarized Ground Unit

This unit has been successfully used in the LDAR(LZM, LDMS, etc.) and MDAR (Z1G, Z2G, Z3G,and PLTG) relaying systems. The operation of thisunit takes advantage of the presence of unfaultedvoltages as the reference or polarizing quantity. Thequadrature-polarized unit is dependent on systemparameters, as will be found next, and has proven tobe a reliable and sensitive unit:

S1 ¼ VA � IA þ Z0c � Z1c

Z1cIA0

� �Z1c ðB-38Þ

S2 ¼ jðVC � VBÞ ðB-39ÞThese two expressions need to be modified. For this

Figure 12B-15 Sequence network for a phase-A-to-ground

fault.

Figure 12B-16 Characteristics of type SDG, SDGU, and

LDG units.

290 Chapter 12

purpose, refer to Figure 12B-15. Then

S1 ¼ ðVA1 þ VA2 þ VA0Þ

� IA1 3þ Z0c � Z1c

Z1c

� �Z1c

¼ VA1 þ ð�IA1Z1sÞ þ ð�IA1Z0sÞ� IA1ð2Z1c þ Z0cÞ þ ðVF1 � VF1Þ

¼ ðVA1 � VF1Þ � IA1ð2Z1c þ Z0c

þ Z1s þ Z0sÞ þ ð�VF2 � VF0Þ¼ ðVA1 � VF1Þ � IA1ð2Z1c þ Z0c þ Z1s

þ Z0sÞ þ IA1ðZ1s þ Z1l þ Z0s þ Z0lÞ¼ ðVA1 � VF1Þ

� IA1ð2Z1c þ Z0c � Z1l � Z0lÞ¼ ðVA1 � VF1Þ � IA1½Z1cð2þ ZRcÞ

� Z1l � Z0l� ðB-40Þ

Working on the other input, we get

S2 ¼ j½ðVC1 þ VC2 þ VC0 � ðVB1 þ VB2 þ VB0Þ�¼ j½VA1ða� a2Þ � VA2ða� a2Þ�¼ �

ffiffiffi3

pðVA1 � VA2Þ

¼ �ffiffiffi3

p½VA1 � ð�IA1Z1sÞ � VF1 þ VF1�

¼ �ffiffiffi3

p½ðVA1 � VF1Þ þ IA1Z1s

þ IA1ðZ1s þ Z1l þ Z0s þ Z0lÞ�¼ �

ffiffiffi3

p½ðVA1 � VF1Þ

þ IA1ð2Z1s þ Z0s þ Z1l þ Z0lÞ�¼ �

ffiffiffi3

p½ðVA1 � VF1Þ

þ IA1ðZ1l þ Z0l þ Z1sð2þ ZRsÞ� ðB-41Þ

For the phase comparator to operate, using Eqs. (B-40) and (B-41), we obtain

S1

S2¼

ðVA1�VF1ÞIA1

� ½Z1cð2þ ZRcÞ � Z1l � Z0l�� ffiffiffi

3p ðVA1�VA1Þ

IA1þ ½Z1sð2þ ZRsÞ þ Z1l þ Z0l�

n oZ1l � ½Z1cð2þ ZRcÞ � Z1l � Z0l�

� ffiffiffi3

p fZ1l þ ½Z1sð2þ ZRsÞ þ Z1l þ Z0l�gðB-42Þ

Therefore, Eq. (B-42) can be expressed as

S1

S2¼ Z1lð2þ ZRlÞ � Z1cð2þ ZRcÞ

� ffiffiffi3

pZ1lð2þ ZRlÞ �

ffiffiffi3

pZ1sð2þ ZRsÞ

ðB-43Þ

If we follow the general procedure for a phase

comparator, the general constants are

k1 ¼ ð2þ ZRlÞ k2 ¼ �ð2þ ZRcÞZ1c

k3 ¼ffiffiffi3

pð2þ ZRlÞ k4 ¼ �

ffiffiffi3

pZ1sð2þ ZRsÞ

The relay settings are chosen such that

ZRl ¼ ZRc ¼ Z0l

Z1l¼ Z0c

Z1c

Therefore, the reference vectors in the R-X diagramare

a ¼ Z1c ðB-44Þ

b ¼ �Z1sð2þ ZRsÞð2þ ZRlÞ ðB-45Þ

The reference vectors are identified in Figure 12B-17,and the locus of impedance vectors (Z1l) is found usingthe techniques discussed for a phase comparatorpreviously. A characteristic angle of +908 is used forthe phase comparator to obtain the circular character-istic.

The influence of the source impedance is evident inthe b vector. The unit can accommodate more faultresistance for short lines, in which the sourceimpedance is large compared to the impedance of theprotected line.

Self-Polarized Mho Ground Unit

This unit has been implemented in the REL 350 (Z2G,Z3G) relay system to clearly define forward andreverse reach. This type of unit utilizes faulted phasequantities for the operating quantity and the restraint

Figure 12B-17 Forward impedance characteristic of the

quadrature polarized ground unit on the Z1l plane.


quantity for which it is called a self-polarized mho unit.This unit is better understood using a phase compara-tor that has the following inputs:

S1 ¼ VA � IA þ Z0cF � Z1cF

Z1cFIA0

� �Z1cF ðB-46Þ

S2 ¼ � VA þ IA þ Z0cR � Z1cR

Z1cRIA0

� �Z1cR

� �ðB-47Þ

These two expressions need to be modified. Referringto Figure 12B-15, we obtain

S1 ¼ ðVA1 þ VA2 þ VA0Þ

� IA1 3þ Z0cF � Z1cF

Z1cF

� �Z1cF

¼ VA1 þ ð�IA1Z1sÞ þ ð�IA1Z0sÞ� IA1ð2Z1cF þ Z0cFÞ � VF1 þ VF1

¼ ðVA1 � VF1

� IA1ð2Z1cF þ Z0cF � Z1L � Z0LÞ ðB-48Þ

and

�S2 ¼ ðVA1 þ VA2 þ VA0Þþ IA1ð2Z1cR þ Z0cRÞ þ VF1 � VF1

¼ ðVA1 � VF1Þ þ ð�IA1Z1sÞþ ð�IA1Z0sÞ þ IA1ð2Z1cR þ Z0cR þ Z1s

þ Z1L þ Z0s þ Z0lÞ¼ ðVA1 � VF1Þ

þ IA1ðZ1l þ Z0l þ 2Z1cR þ Z0cRÞ ðB-49Þ

For this phase comparator to produce an output, withthe use of Eqs. (B-48) and (B-49), the characteristic isdefined by

S1

S2¼

ðVA1�VF1ÞIA1

� ð2Z1cF þ Z0cF � Z1l � Z0lÞ� ðVA1�VF1Þ

IA1� ðZ1l þ Z0l þ 2Z1cR þ Z0cRÞ

ðB-50Þ

Therefore, Eq. (B-50) can be reduced to

S1

S2¼ Z1lð2þ ZRlÞ � Z1cFð2þ ZRcFÞ

�½Z1lð2þ ZRlÞ þ Z1cRð2þ ZRcRÞ�¼ M1e

+j90� ðB-51Þ

Equation (B-51) has the general format for a phasecomparator and to plot its characteristic on the Z1l

plane.

For this phase comparator, we get

k1 ¼ ð2þ ZRlÞ k2 ¼ �Z1cFð2þ ZRcFÞk3 ¼ �ð2þ ZRlÞ k4 ¼ �Z1cRð2þ ZRcRÞ

The relay settings are chosen such that

ZRl ¼ ZRcF ¼ ZRcR ¼ Z0l

Z1l

¼ Z0cF

Z1cF¼ Z0cR

Z1cRðB-52Þ

Therefore, the reference vectors defining the character-istic of the unit on the Z1l plane are

a ¼ Z1cF ðB-53Þb ¼ �Z1cR ðB-54ÞThe a and b vectors are identified in Figure 12B-18

and the locus of impedance vectors, Z1l, that satisfyEq. (B-51) is illustrated using the techniques discussedbefore.

Notice that the Z1cR setting defines explicitly thecharacteristics of the unit and the source impedancehas no influence. In general, any self-polarized unit willnot be dependent on the source impedance; on theother hand, as seen before, any unit that uses acombination of the unfaulted phases will be dependenton the source impedance.

Reactance Ground Unit

This unit has been implemented in the type KDXGreactance ground relay. It is better understood with a

Figure 12B-18 Forward impedance characteristic of the

self-polarized mho ground unit.

292 Chapter 12

phase comparator that has the following inputs:

S1 ¼ VA � IA þX0c �X1c

X1cIA0

� �jX1c ðB-55Þ

S2 ¼ j IA þX0c �X1c

X1cIA0

� �X1c ðB-56Þ

These expressions need to be simplified. Referring toFigure 12B-15, for a phase-A-to-ground fault, we get

S1 ¼ VA1 þ VA2 þ VA0

� IA1 3þX0c �X1c

X1c

� �jX1c

¼ VA1 þ ð�IA1Z1sÞ þ �ðIA1Z0sÞ� jIA1ð2X1c þX0cÞ þ VF1 � VF1

¼ ðVA1 � VF1Þ � IA1½ jð2X1C þX0CÞ� Z1L � Z0L� ðB-57Þ

and

S2 ¼ jð2X1c þX0cÞIA1 ðB-58ÞEquations (B-57) and (B-58) determine the impedancecharacteristic of the comparator. It follows from theprocedure discussed above that

S1

S2¼ Z1l � ðj2X1c þ jX0c � Z1l � Z0lÞ

jð2X1c þX0cÞ ðB-59ÞS1

S2¼ Z1lð2þ ZRlÞ � jX1Cð2þXRcÞ

jX1Cð2þXRcÞ¼ M1e

+j90� ðB-60ÞThe settings of the relay are made such that

ZRl ¼ Z0l

Z1l¼ X0c

X1c

Finding the general constants for a phase comparatorin Eq. (B-60), we obtain

k1 ¼ ð2þ ZRlÞ k2 ¼ jX1cð2þXRcÞk3 ¼ 0 k4 ¼ jX1cð2þXRcÞ

It follows that

a ¼ jX1c ðB-61Þb ¼<+e j90

� ðinfinity at 90�Þ ðB-62Þ

Equation (B-62) might not be a rigorous mathema-tical expression, but it identifies the location of one ofour reference vectors. It can be thought, therefore, thatðZ� bÞ is always perpendicular to the reactance lineand the phase comparator characteristic angle require-

ment every time met. Figure 12B-19 illustrates theconstruction and characteristics of the reactance unit.It should be mentioned that this unit was designed tocover more ground fault resistance; however, it couldnot work by itself since it is nondirectional and wouldeven operate for load. Therefore, it had to besupervised by some other unit, as will be illustratedlater.

B.4.5 Blinder Units

Blinders are impedance elements that are used for out-of-step relaying and also supervising impedance unitson load current encroachment. These elements havebeen employed in REL-300 and REL-350 systems.Implementation is similar to that for the reactanceunit, and the phase comparator approach makes iteasier to understand. A typical implementation in aphase comparator has inputs

S1 ¼ �VA þ IARTejðPANG�90�Þ ðB-63Þ

S2 ¼ þIAejðPANG�90�Þ ðB-64Þ

Since blinders are used for out-of-step conditions, itmonitors three-phase conditions only; the abovequantities are all positive sequence line current andvoltage inputs.

Following the procedure yields

S1

S2¼ Z1l �RTe

jðPANG�90�Þ

�e jðPANG�90�Þ ¼ M1e+j90� ðB-65Þ

Equation (B-65) appears in the standard format so far

Figure 12B-19 Reactance unit characteristics.


used to derive phase comparator characteristics.Therefore,

k1 ¼ 1 k2 ¼ �RTejðPANG�90�Þ ðB-66Þ

k3 ¼ 0 k4 ¼ �e jðPANG�90�Þ ðB-67Þa ¼ RTe

jðPANG�90�Þ

b ¼ �<e jðPANG�90�Þ ¼ <e jðPANGþ90�Þ

Equation (B-67) may not be a rigorous mathematicalexpression, but it provides a reference to infinity forour b vector. This way, the magnitude does not matter,just the angle to infinity (PANGþ 90). If we refer toFigure 12B-20, (Z� b) can be thought of as alwaysperpendicular to the line impedance Z1l that the unit isbeing applied to.

B.5 Reverse Characteristics of an ImpedanceUnit

In the previous section, the forward-looking character-istics of several impedance units were analyzed. Itshould be stressed that in the above R-X plots, the factthat the impedance locus may go through the third andfourth quadrant does not imply that the unit isnondirectional or operate for a reverse fault. Theprevious plots are all forward-looking.

To investigate the directionality of an impedanceunit, the reverse-looking characteristics need to befound. The idea is to find the locus of reverse

impedances ðZ001lÞ for which the unit will operate. For

an impedance unit to be secure and directional, thereverse-looking impedance ðZ00

1lÞ locus should lie on thethird quadrant (negative R and negative X).

For the purpose of studying the reverse character-istics of the different units derived, the circuit inFigure 12B-21 is assumed. The source impedance,denoted Z00

1s, is the composite of the line beingprotected and the source impedance ‘‘looking’’ for-ward. The reverse-looking impedance, Z00

1l, is theimpedance in the reverse direction and the purposeof this study is to find the locus of impedances Z00

1l forwhich the units will operate.

This circuit is again a radial circuit so that theinfluence of different distribution factors can bedisregarded. The directions of the currents are differentfrom before. The plots now will be done on the reverse-looking line impedance Z00

1l.


Following the same approach used before, for a phase-to-phase fault, Figure 12B-22 describes the connectionof sequence networks for a phase-to-phase fault.

The direction of the currents in Figure 12B-22indicates the proper direction of the sequence compo-nents that the unit sees. The inputs to the magnitudecomparator are again


and the impedance characteristic is defined by

SA

SB¼ VA2 � IA2Zc

VA1 � IA1Zc¼ e jy ðB-70Þ

Figure 12B-20 Typical blinder unit characteristics.

Figure 12B-21 Radial system for studying the reverse

characteristic of distance units.

294 Chapter 12

and substituting voltages from the figure yields

SA

SB¼ �ðIA1Z

001sÞ � ð�IA1ZcÞ

�IA1ð2Z001l þ Z00

1sÞ � IA1Zc¼ e jy ðB-71Þ

or on the positive sequence R-X plane, the character-istic is

Z001l ¼ � 1

2ðZ00

1s þ ZcÞ þ 1

2ðZ00

1s � ZcÞe�jy ðB-72Þ

Figure 12B-23 illustrates the impedance characteristicof the unit on the R-X diagram of Z00

1l, the reverse-direction impedance.

B.5.2 Three-Phase Unit

Figure 12B-24 illustrates a three-phase fault in thereverse direction. The voltage and current (VA1 andIA1) directions are the references that the distance unit

‘‘sees.’’ The inputs to the magnitude comparator are

SA ¼ �Zc

2IA1 ðB-73Þ

SB ¼ VA1 � Zc

2IA1 ðB-74Þ

Therefore, the characteristic of the device is defined by

SA

SB¼ � Zc

2IA1

VA1 � Zc

2 IA1

¼ e jy ðB-75Þ

Substituting voltages with currents, we get

SA

SB¼ � Zc

2 IA1

ð�Z001lIA1Þ � Zc

2IA1

¼ ejy ðB-76Þ

The operating characteristic of the unit is defined by

Z001l ¼ �Zc

2þ Zc

2e�jy ðB-77Þ

Figure 12B-25 illustrates the characteristic of the uniton the reverse positive sequence impedance Z00

1l plane.

B.5.3 Ground Units

To develop the reverse-looking characteristic of thedifferent ground units, Figure 12B-26 will be used. Itillustrates the connection of the sequence networks foran A-to-ground fault. The directions of the currentsare those that the relay actually ‘‘sees.’’

SDGU Impedance Unit

The inputs to the magnitude comparator are:

SA ¼ VA0 � IA0Z0c ðB-78ÞSB ¼ ðVA1 þ VA2Þ � ðIA1 þ IA2ÞZ1c ðB-79Þ

Figure 12B-22 Network connection for a phase-to-phase

fault. The direction of the currents indicate the actual

component currents seen by the distance units.

Figure 12B-23 Phase-to-phase unit characteristic on the

reverse direction.

Figure 12B-24 Network for a three-phase fault.


The characteristic is defined by

SA

SB¼ VA0 � IA0Z0c

ðVA1 þ VA2Þ � ðIA1 þ IA2ÞZ1c

¼ e jr ðB-80Þmaking the appropriate substitutions from Figure12B-31

SA

SB¼ ðIA0Z

000sÞ � IA0Z0c

�IA1ð2Z001l þ Z00

0l þ Z000sÞ � ð2IA0Z1cÞ ¼ e jy

or

SA

SB¼ �Z00

0s þ Z0c

2Z001l þ Z00

0l þ Z000s þ 2Z1c

¼ e jy ðB-81Þ

Then, the equation that defines the characteristic onthe Z00

1l plane is

Z001l ¼ �ð2Z1c þ Z00

0sÞð2þ ZR00

l Þþ Z1cZRc � Z00

0s

ð2þ ZR00l Þ

e jy ðB-82Þ

This characteristic is illustrated in Figure 12B-27.

Quadrature-Polarized Ground Unit

The inputs to the phase comparator are

S1 ¼ VA � IA þ IA0Z0c � Z1c

Z1c

� �Z1c ðB-83Þ

S2 ¼ jðVC � VBÞ ðB-84ÞThe directions of the currents and voltages are thoseshown in Figure 12B-26. The above equations can be

Figure 12B-25 Reverse characteristic of the three-phase

unit.

Figure 12B-26 Sequence network connection for a reverse

A-to-ground fault. Figure 12B-27 Reverse characteristic of the SDGU relay.

296 Chapter 12

modified to

S1 ¼ ðVA1 þ VA2 þ VA0Þ � IA16 3þ Z0c þ Z1c

Z1c

� �Z1c

¼ VA1 þ ðIA2Z001sÞ þ ðIA0Z

000sÞ � IA1ð2Z1c þ Z0cÞ

þ ðVF1 � VF1Þ¼ ðVA1 � VF1Þ þ IA1ðZ00

1s þ Z000s � 2Z1c � Z0c � Z00

1l

� Z000l þ Z00

1s þ Z000sÞ

or

S1

IA1¼ �Z00

1l � Z000l � 2Z1c � Z0c

¼ �Z001lð2þ ZR00

l Þ � Z1cð2þ ZRcÞ ðB-85Þ

Also, working on S2, we get

S2 ¼ jðVC1 þ VC2 þ VC0 � VB1 � VB2 � VB0Þ¼ j½VA1ða� a2Þ � VA2ða� a2Þ�¼ �

ffiffiffi3

pðVA1 � VA2Þ

¼ �ffiffiffi3

p½VA1 � ðIA1Z

001sÞ � VF1 þ VF1�

¼ �ffiffiffi3

p½ðVA1 � VF1Þ � IA1Z

001s þ VF1�

or

S2

IA1¼ �

ffiffiffi3

p½�Z00

1l � Z001s � ðZ00

1l þ Z001s þ Z00

0l � Z000sÞ�

¼ �ffiffiffi3

p½�2Z00

1l � 2Z001s � Z00

0l � Z000s�

¼ �ffiffiffi3

p½Z00

1lð2þ ZR00l Þ þ Z00

1sð2þ ZR00s Þ� ðB-86Þ

The operating characteristic is defined by

S1

S2¼ �Z00

1lð2þ ZR001 Þ � Z1cð2þ ZRcÞ

� ffiffiffi3

p ½Z001lð2þ ZR00

l Þ þ Z1sð2þ ZR00s Þ�

¼ M1e+j90� ðB-87Þ

If we follow the procedure for a phase comparator,the general constants are

k1 ¼ �ð2þ ZR00l Þ

k2 ¼ �Z1cð2þ ZRcÞk3 ¼

ffiffiffi3

pð2þ ZR00

l Þk4 ¼

ffiffiffi3

pð2þ ZR00

s ÞZ001s ðB-88Þ

The reference vectors are therefore

a ¼ � ð2þ ZRcÞð2þ ZR00

l ÞZ1c

and

b ¼ �ð2þ ZR00s Þ

ð2þ ZR00l Þ

Z1s ðB-89Þ

Figure 12B-28 illustrates the reverse characteristics ofthe quadrature-polarized relay.

Self-Polarized Ground Unit

The inputs to the phase comparator are

S1 ¼ VA � IA þ Z0cF � Z1cF

Z1cFIA0

� �Z1cF ðB-90Þ

S2 ¼ � VA þ IA þ Z0cR � Z1cR

Z0cRIA0

� �Z1cR

� �ðB-91Þ

Modifying the two expressions using Figure 12B-26and following the same procedure as for the quad-rature-polarized distance unit, we obtain

S1

IA1¼ �Z00

1lð2þ ZR00l Þ � Z1cFð2þ ZRcFÞ ðB-92Þ

and

S2

IA1¼ �Z00

1lð2þ ZR00l Þ þ Z1cRð2þ ZRcRÞ ðB-93Þ

Figure 12B-29 illustrates the reverse characteristics ofthis unit. It is assumed that ZRcF ¼ ZRcR ¼ ZR00

l .

Figure 12B-28 Reverse characteristics of the quadrature

polarized relay.


B.6 Response of Distance Units to DifferentTypes of Faults

Distance units are designed specifically to operate forcertain faults; a phase A ground distance unit shouldrespond to phase-A-to-ground faults. However, alldistance units do have a characteristic for other typesof faults. A phase A ground distance unit will have aresponse to phase-B-to-ground, phase-C-to-ground,phase-to-phase, etc., since the unit will essentially stillbe receiving the input quantities from the powersystem. It needs to be remarked that, in general,distance units responding to other faults will have ashorter reach for other types of faults.

In the next sections, the text will concentrate on thephase-to-phase, quadrature-polarized ground unit andself-polarized ground unit only.


Response to Other Phase-to-Phase Faults

In Section 4.2, the response-to a BC phase-to-phasefault was investigated. If the same approach is used forother phase-to-phase faults, the analysis will yield thesame result: For all phase-to-phase faults, this unit hasthe characteristic shown in Figure 12B-12 in thisappendix. This means that we can use only one phase-to-phase unit to detect all types of phase-to-phasefaults. There is no need for three units for all phase-to-phase faults.

Response to a Three-Phase Fault.

The inputs to the magnitude comparator are


for a three-phase fault SA ¼ 0. Therefore, this unit isnot responsive to three-phase faults and it is notaffected by normal load flow at all. This is an excellentcharacteristic that makes this unit unique.

Response to Phase-to-Ground Faults.

The unit shows the same response for all forwardphase-to-ground faults. If a figure similar to 12B-15 isused and Eqs. (B-23) and (B-24) (SA and SB) areapplied, the locus of impedances on the Z1l plane isdefined by the equation

Z1l ¼ Zc � Z1sð1þ ZRsÞð2þ ZRlÞ � ðZ1s þ ZcÞe�jr

ð2þ ZRlÞ ðB-96Þ

The characteristic for the response of the phase-to-phase unit to all phase-to-ground faults is shown inFigure 12B-30.

B.6.2 Quadrature-Polarized Ground-DistanceUnit

In all microprocessor-based relays, there will be threeground units to detect all the ground fault types. Theunits are constantly receiving the inputs from thepower system and they do have a definite response to

Figure 12B-29 Reverse characteristics of the nondirectional

ground unit.

Figure 12B-30 Response of the phase-to-phase unit to all

phase-to-ground faults.

298 Chapter 12

other types of faults. In this sense, the basic approachused thus far will still help us in the analysis of theresponse of the phase A ground-distance unit to allother types of faults.

Response to Other Phase-to-Ground Faults

If other phase-to-ground faults are applied to thephase A unit, it will tend to have a response mainly dueto the IA0 term in S1. For analysis, a connection similarto Figure 12B-15 should be used, but always byconsidering the phase shifts (a or a2) introduced whenusing other phases as a reference. In this sense, thereference a and b vectors can be calculated.

For a phase-B-to-ground fault, we have

a ¼ �Z1sð1� aÞ þ ð1� a2ÞZRs

ð2þ ZRlÞ

þ Z1cð1þ aÞ þ a2ZRc


and

b ¼ �Z1sð1þ aÞ þ ZRs


For a phase-C-to-ground fault, we have

a ¼ �Z1sð1� a2Þ þ ð1� aÞZRs

ð2þ ZRlÞ

þ Z1cð1þ a2Þ þ aZRc


and

b ¼ �Z1sð1þ a2Þ þ ZRs


The above reference vectors for a phase comparatordefine the characteristics of the phase A ground unitfor phase-B-to-ground and phase-C-to-ground faults.Figure 12B-31 illustrates the characteristics of thephase A unit for all phase-to-ground faults.

Response to a Three-Phase Fault.

For a three-phase fault, the actual inputs to the phasecomparator for the phase A unit are

S1 ¼ VA � IAZ1c ðB-101ÞS2 ¼ jðVC � VBÞ ðB-102Þ

If we use Figure 12B-13 and the general procedure for

a phase comparator, the reference vectors are

a ¼ Z1c ðB-103Þb ¼ 0 ðB-104Þ

Notice that the above equations imply that the reachfor a three-phase fault for the phase A unit is the sameas its setting for phase-to-ground faults. This impliesthat this same unit can be used to detect three-phasefaults. In general, the operation of all three-phase-to-ground units indicates a three-phase fault. It isunfortunate, however, that the unit itself is notdependent on the source impedance, and a fault rightat the bus will not be detected because there is noreference quantity, S2 ¼ 0. To make up for the lack ofpolarizing quantity for a bus fault, REL-300 uses theprefault voltage as a reference. This concept is called‘‘memory voltage,’’ and it is used to increase thecoverage of the unit. This concept is used in manyother designs to avoid the voltage inversion problem inlines with series-capacitor compensation that wasdescribed in the transmission-line protection chapter.Hence, memory voltage increases the coverage and isused to protect series-capacitor lines.

If memory voltage is employed, then the inputs tothe phase comparator are

S1 ¼ VA � ðIAÞZ1c ðB-105ÞS2 ¼ jðVCm � VBmÞ ðB-106Þ

S1 needs no simplification. However, S2 can be

Figure 12B-31 Characteristic for all phase-to-ground

faults.


modified to

S2 ¼ jða� a2ÞVAm

¼ �ffiffiffi3

pðVsÞ

¼ �ffiffiffi3

pIA1ðZ1l þ Z1sÞ ðB-107Þ

Therefore, following the general procedure for a phasecomparator, we get

S1

S2¼ VA � IA1Z1c

� ffiffiffi3

pIA1ðZ1l þ Z1sÞ

¼ M1e+j90� ðB-108Þ

or

S1

S2¼ Z1l � Z1c

� ffiffiffi3

p ðZ1l þ Z1sÞ¼ M1e

+j90�

The reference vectors a and b are

a ¼ Z1c ðB-109Þb ¼ �Z1s ðB-110Þ

Therefore, if the unit has memory voltage, the forwardreach will not be modified, but the coverage of thephase A ground-distance unit for three-phase faultswill be dependent on the source impedance. It isevident that now bus faults can be detected easily.Figure 12B-32 illustrates the response of the quad-rature-polarized ground unit to forward three-phasefaults with and without ‘‘memory’’ action.

Response to Phase-to-Phase Faults

It can be found that by using a figure similar to 12B-11and applying the general procedure for a phasecomparator to the S1 and S2 inputs in this unit, the

reference vectors are as follows. For a BC fault,

a ¼ �Z1s ðB-111Þb ¼ 0 ðB-112Þ

For a CA fault,

a ¼ � Z1sðaþ 1Þ2

þ ða� 1ÞZ1c

2

� �ðB-113Þ

b ¼ � Zs1ð1� aÞ2

� �ðB-114Þ

For a AB fault,

a ¼ � Z1sða2 þ 1Þ2

þ ða2 � 1ÞZ1c

2

� �ðB-115Þ

b ¼ � Zs1ð1� a2Þ2

� �ðB-116Þ

The response of the quadrature-polarized phase A unitto all forward phase-to-phase faults is shown inFigure 12B-33.

B.6.3 Self-Polarized Ground Unit

Response to Other Phase-to-Ground Faults

For analysis, a connection similar to Figure 12B-15should be used, but always by considering the phaseshifts (a or a2) introduced when using other phases as areference. In this sense, the reference a and b vectors

Figure 12B-32 Forward characteristic for three-phase

faults. With and without ‘‘memory’’ action.

Figure 12B-33 Characteristics for forward phase-to-phase

faults.

300 Chapter 12

can be calculated. For a phase-B-to-ground fault,

a ¼ �Z1sð1� aÞ þ ð1� a2ÞZRs

ð2þ ZR1Þ

þ Z1cFð1þ aÞ þ a2ZRcF


b ¼ �Z1sð1� aÞ þ ð1� a2ÞZRs

ð2þ ZRlÞ

� Z1cRð1þ aÞ þ a2ZRcR


For a phase-C-to-ground fault,

a ¼ �Z1sð1� a2Þ þ ð1� aÞZRs

ð2þ ZRlÞ

þ Z1cFð1þ a2Þ þ aZRcF


b ¼ �Z1sð1� a2Þ þ ð1� aÞZRs

ð2þ ZRlÞ

� Z1cRð1þ a2Þ þ aZRcR


The above reference vectors for a phase comparatordefine the characteristics of the phase A ground unitfor phase-B-to-ground and phase-C-to-ground faults.Figure 12B-34 illustrates the characteristics of thephase A unit for all phase-to-ground faults.

Response to a Three-Phase Fault

For a three-phase fault, the actual inputs to the phasecomparator for the phase A unit are

S1 ¼ VA � IA1Z1cF ðB-121ÞS2 ¼ �ðVA þ IA1Z1cRÞ ðB-122Þ

If we use Figure 12B-13 and the general procedure fora phase comparator, the reference vectors are

a ¼ Z1cF ðB-123Þb ¼ Z1cR ðB-124Þ

This means that the unit’s characteristic for a three-phase fault will be the same as for phase-to-groundfaults (Fig. 12B-18). In REL 350 the operation of allthe three-phase-to-ground units indicates the occur-rence of a three-phase fault.

Response to Phase-to-Phase Faults

It can be found that by using a figure similar to 12B-11and applying the general procedure for a phasecomparator to the S1 and S2 inputs in this unit, thereference vectors are as follows. For a BC fault,

a ¼ �Z1s ðB-125Þb ¼ �Z1s ðB-126Þ

For a CA fault,

a ¼ � Z1sð1þ aÞ2

þ ða� 1ÞZ1cF

2

� �ðB-127Þ

b ¼ � Z1sð1þ aÞ2

þ ða� 1ÞZ1cR

2

� �ðB-128Þ

For a AB fault,

a ¼ � Z1sð1þ a2Þ2

þ ða2 � 1ÞZ1cF

2

� �ðB-129Þ

b ¼ � Z1sð1þ a2Þ2

þ ða2 � 1ÞZ1cR

2

� �ðB-130Þ

The response of the self-polarized phase A ground unitto all forward phase-to-phase faults is shown inFigure 12B-35.

B.6.4 Double Phase-to-Ground Faults

It is noticeable that the double line-to-ground faultshave been avoided thus far. The reason is that theanalysis of such faults using the general procedureexplained above and maybe other procedures becomestoo complicated. However, in general, it is accepted

Figure 12B-34 Characteristics for all phase-to-ground

faults.


that the analysis of the performance of distance unitscan be made only for phase-to-ground, phase-to-phase, and three-phase faults. Phase-to-phase-to-ground fault detection is achieved by the operationof the units detecting phase-to-phase and three-phasefaults.

Figure 12B-36 shows the sequence connection for aphase-to-phase-to-ground fault. Notice that the zerosequence impedance is present in the sequence con-

nection. The detection of a phase-to-phase-to-groundfault will be done by the phase-to-phase unit if the zerosequence impedance is large. This means that a phase-to-phase-to-ground fault approaches a phase-to-phasefault if the zero sequence impedance is large. This factis evident in Figure 12B-36, and it should be detectedby the phase-to-phase unit.

On the other hand, if the zero sequence impedanceis small, then the phase-to-phase-to-ground faultapproaches a three-phase fault, and the three-phasefault detection units should operate for phase-to-phasefaults.

In many years of operation, considering the phase-to-phase-to-ground fault as a variation of the phase-to-phase or three-phase fault has never been a problem.

B.7 The Influence of Current DistributionFactors and Load Flow

In all the above analyses, a radial system was assumedfor the purpose of simplicity. In real life, systems arenot necessarily radial and they are carrying load. It isthe purpose of this section to describe the role thatdistribution factors and load current have in theoperating characteristics of the impedance units.

Figure 12B-37 illustrates the definition of distribu-tion factors for all the sequence networks. Distributionfactors are complex per unit factors of the totalsequence current at the fault. In the following

Figure 12B-35 Characteristics for all forward phase-to-

phase faults.

Figure 12B-36 Double phase-to-ground sequence network

connection. Figure 12B-37 Distribution factors in all the networks.

302 Chapter 12

paragraphs, distribution factors will be used todescribe the operating characteristics of the phase-to-phase unit, quadrature-polarized ground unit, and self-polarized ground unit; however, a note of cautionshould be given. The analysis is a simple academiccuriosity, and although it accurately describes thegeneral case, the actual impedance characteristic of theunit will be difficult to visualize since distributionfactors (K0, K1, K2) will change, depending on thelocation of the fault in the line.

From Figure 12B-37, we see that the following istrue:

K1 ¼ VsðZ01s þ Z0

1lÞVsðZ0

1s þ Z01lÞ þ V0

sðZ1s þ Z1lÞ ðB-131Þ

K2 ¼ ðZ01s þ Z0

1lÞðZ0

1s þ Z01lÞ þ ðZ1s þ Z1lÞ ðB-132Þ

K0 ¼ ðZ00s þ Z0

0lÞðZ0

0s þ Z00lÞ þ ðZ0s þ Z0lÞ ðB-133Þ

Notice that if there is no load flow, Vs ¼ V0s, then

K1 ¼ K2. It can be also shown that

1�K2

K1

� �¼ IL

IF3phðB-134Þ


Refer to Figure 12B-38 for the analysis of this unit.The inputs to the phase comparator are still

SA ¼ VA2 �K2IA2Zc

¼ K2IA1Z1s þK2IA1Zc ðB-135ÞSB ¼ VA1 �K1IA1Zc

¼ ðK1 þK2ÞIA1Z1l þK2Z1sIA1 ðB-136Þ�K1ZcIA1

Using the general procedure for a magnitude com-parator, we can show the locus of impedances Z1l thatdefines the impedance characteristics for this unit to be

Z1l ¼ Zc � Z1sð1� IL=IF3phÞð2� IL=IF3phÞ

þ ðZs1 þ ZcÞð1� IL=IF3phÞð2� IL=IF3phÞ e�jr ðB-137Þ

Notice that the unit’s reach is still fixed at Zc and itscenter will move depending on the load flow. Itscharacteristic is not dependent on the distributionfactors, but the prefault load flow. The characteristic isplotted in Figure 12B-39.

B.7.2 Quadrature-Polarized Ground Unit

Reference will be made to Figure 12B-40 that showsthe sequence network connection for a phase-to-ground fault. Distribution factors for all the networksmake the inputs to the phase comparator

S1 ¼ VA � IA þK0IA0Z0c � Z1c

Z1c

� �Z1c

and reduce to

S1

K1IA1¼ Z1l 2þK0

K1ZRl � IL

IF3ph

� �

� Z1c 2þK0

K1ZRc � IL

IF3ph

� �ðB-138Þ

Also, working on S2, we get

S2 ¼ jðVC � VBÞ

Figure 12B-38 General case of a phase-to-phase fault. Figure 12B-39 Characteristic for a phase-to-phase fault.


that reduces to

S2

K1IA1¼ �

ffiffiffi3

pZ1l 2þK0

K1ZRl � IL

IF3ph

� ��

þZ1s 2þK0

K1ZRs � 2

IL

IF3ph

� ��ðB-139Þ

If the settings of the unit are made such thatZRc ¼ ZRl, the a and b reference vectors for a phasecomparator are

a ¼ Z1c ðB-140Þ

b ¼ �Z1s

2þ K0

K1ZRs � 2 IL

IF3ph

� �2þ K0

K1ZRl � IL

IF3ph

� �24

35 ðB-141Þ

The characteristic is illustrated in Figure 12B-41 andshows that the reach of the unit Z1c remainsunchanged. It is dependent on the current distributionfactors of the power system and the prefault load flow.

B.7.3 Self-Polarized Ground Unit

Reference will be made to Figure 12B-41 that showsthe sequence network connection for a phase-to-ground fault. Distribution factors for all the networksmake the inputs to the phase comparator

S1 ¼ VA � IA þK0IA0Z0cF � Z1cF

Z1cF

� �Z1cF

and reduce to

S1

K1IA1¼ Z1l 2þK0

K1ZRl � IL

IF3ph

� �

� Z1cF 2þK0

K1ZRc � IL

IF3ph

� �ðB-142Þ

Also,

�S2 ¼ VA þ IA þK0IA0Z0cR � Z1cR

Z1cR

� �Z1cR

and it reduces to

S2

K1IA1¼ �Z1l 2þK0

K1ZRl � IL

IF3ph

� �

� Z1cR 2þK0

K1ZRc � IL

IF3ph

� �ðB-143Þ

If the settings of the unit are made such thatZRcF ¼ ZRcR ¼ ZRl, the a and b reference vectors fora phase comparator are

a ¼ Z1cF ðB-144Þb ¼ �Z1cR ðB-145Þ

The characteristic is the same as that in Figure 12B-18and does not depend on load flow, or distributionfactors. This is true, in general, for units polarized withthe same faulted phase; the characteristic in the R-Xdiagram is static and does not depend on anyparameter but the settings of the unit.

Figure 12B-40 Sequence connection for a general phase-to-

ground fault.

Figure 12B-41 Characteristic for phase-to-ground faults.

304 Chapter 12

B.8 Derived Characteristics

The discussion above showed the different character-istics that can be obtained with either a phase ormagnitude comparator. The characteristics of the unitsderived are smooth and simple.

Designs have been made, however, in whichdifferent characteristics, many comparators, havebeen used to achieve characteristics that are notnecessarily smooth. This is done for many purposes,like load restriction for heavily loaded long lines, andto try to accommodate more fault resistance, althoughactual apparent impedance is not simply a resistancecomponent.

It is not the scope of this section to discuss theadvantages and disadvantages of impedance character-istics. Therefore, a simple description of how thesecharacteristics are obtained will be given throughexamples.

Refer to Figure 12B-42a. It shows an applicationwith mixed reactance and blinder units. To derive thischaracteristic, four units are required. The reactanceunit X determines the X reach, the resistance unit Rdetermines the R reach, and blinders A and B providethe necessary directionality of the unit. For the relay tooperate, the four units should pick up.

In Figure 12B-42c, two comparators are used toachieve the characteristic. In this case, any of the units,offset mho or lens, will provide the output of the entiredistance unit. This characteristic has been used inheavily loaded lines for the starting impedance unit.

Figure 12B-42b illustrates an intentional increase ofthe resistance reach for a mho unit. In this case, onlythe mho unit operates for the þ 908 coincidencerequirement, and the output is left to the operationof all three blinders for the � 908 coincidence require-ment.

Figure 12B-42d illustrates the use of a mho and lensunit to increase the resistance coverage of the mhoelement. The angle of Zc in both units is different andthe output taken from the operation of any of the twocomparators.

B.9 Apparent Impedance

The term apparent impedance is used very commonlyto describe the resultant impedance when the voltagesand currents entering the relay are used to calculatethe impedance to the fault. This expression isconfusing if one confuses it with the ‘‘impedancecharacteristics’’ that have been described in theprevious paragraphs.

In this section, we define apparent impedance as theloop impedance using the voltages and currents thatthe relay receives. In this sense, the following are theapparent impedances for different fault loops. For athree-phase fault,

Zapp ¼ VA

IAðB-146Þ

For a phase-to-phase fault,

Zapp ¼ VB � VC

IB � ICðB-147Þ

For a phase-to-ground fault,

Zapp ¼ VA

IA þ Z0c�Z1c

Z1cIA0

ðB-148Þ

If the apparent impedance Zapp is within theoperating characteristics, those studied above, thenthe unit will operate. Notice that the operatingcharacteristics of the unit only depend on the settingof the unit and, in some cases, the sourceimpedance.

For the phase loops, three-phase and phase-to-phase, mutual and fault resistance have little influence,and the actual apparent impedance Zapp is the same asthe impedance to the fault.

The influence of mutuals and fault resistance ismore severe in the phase-to-ground loop. Faultresistance and zero sequence mutual effects from otherFigure 12B-42 Composite impedance characteristics.


lines tend to make the apparent impedance Zapp

different from the actual impedance to the fault.Taking into consideration these factors, we knowthat the apparent impedance to the fault with faultresistance and zero sequence mutual effects is

Zapp ¼ Z1l þ 3IA0Rf þ I0EZ0m

IA þ Z0c�Z1c

Z1cIA0

ðB-149Þ

where

Rf ¼ fault resistanceI0E ¼ zero sequence current in the parallel lineZ0m ¼ zero sequence mutual between parallel lines

and the rest of the symbols are as described before.It should be noticed that apparent impedance will

change with fault resistance and mutual effects, and ifZapp enters the operating characteristics of the differentground units described, the relay may under- oroverreach.

B.10 Summary

The purpose of this appendix is to illustrate thedifferent properties of distance elements found inABB relays. In microprocessor technology, severalof these units can be found and the detection offaults in power systems is, in general, a combina-tion of the operation of all the phase and groundunits. In some designs, such as MDAR (REL-300)all the distance units are supervised by a phaseselector and/or forward directional unit. This limitsthe interaction of the several units that MDAR hasavailable. In conclusion, although distance elementshave been designed with a definite type of faultdetection, they will respond to other types of faultsas well.

It is an objective of this appendix, as well, to clarifythe meaning of forward and reverse. These conceptsare totally opposite to each other. It is not proper todraw the forward-looking characteristics ðZ1lÞ planeon the reverse-looking characteristics ðZ

1l00 Þ plane. It is

hoped that the reader can now make such a distinction.They are totally different concepts.

The influence of current distribution factors andload is simply an academic exercise since these factorschange as the fault moves within the protected line. Tofind rigorously the operating characteristic withdistribution factors and load flow would be an iterativeprocess. However, the development illustrates thedynamic performance of the different units.

The term apparent impedance has also been defined.It corresponds to the fault-loop evaluation, and if theapparent impedance falls inside the operating char-acteristics of the unit, the unit will operate. Apparentimpedance for ground faults is affected by faultresistance and zero sequence mutual effects.

APPENDIX C: INFEED EFFECT ON GROUND-DISTANCE RELAYS

Because the current of the reference and/or operatingquantities for ground-distance relays are different fromthose for phase-distance relays, therefore, the descrip-tions and results provided in Section 4.5 for phase-distance relays on infeed effect cannot be directly usedfor ground-distance relay application. The generalsolution for infeed effect on ground-distance relays willbe described as follows.

C.1 Infeed Effect on Type KDXG, LDAR, andMDAR Ground-Distance Relays

Refer to Figure 12C-1.

�VA1 þ 2IA1GZ1L

þ 2ðIA1G þ IA1HÞZ1H

� VA0 þ IA0GZ0L þ ðIA0G

þ IA0HÞZ0H � VA2 ¼ 0 ðC-1ÞVAG ¼ VA1 þ VA2 þ VA0

ðvoltage at relay locationÞ ðC-2ÞFrom Eqs. (C-1) and (C-2), we get

VAG ¼ Z1L 2IA1G þ IA0Z0L

Z1L

� ��

þ Z1H

�2ðIA1G þ IA1HÞ

þ ðIA0G þ IA0HÞ Z0L

Z1L

� ��

¼ Z1L IG þ Z0L � Z1L

Z1L

� �IA0G

� �

þ Z1H

�ðIG þ IHÞ

þðIA0G þ IA0HÞ Z0L � Z1L

Z1L

� ��ðC-3Þ

IR ¼ IG þ Z0L � Z1L

Z1L

� �IA0G ðC-4Þ

306 Chapter 12

Equations (C-3) and (C-4) yield

ZAPP ¼ VAG

IR¼ VAG

IG þ Z0L�Z1L

Z1L

� �IA0G

¼Z1L IG þ Z0L�Z1L

Z1L

� �IA0G

h iþ Z1H ðIG þ IHÞ þ ðIA0G þ IA0HÞ Z0L�Z1L

Z1L

� �h iIG þ Z0L�Z1L

Z1L

� �IA0G

¼ Z1L þ Z1H

ðIG þ IHÞ þ ðIA0G þ IA0HÞ Z0L�Z1L

Z1L

� �IG þ Z0L�Z1L

Z1L

� �IA0G

24

35

¼ Z1L þ Z1H 1þIH þ IA0H

Z0L�Z1L

Z1L

� �IG þ IA0G

Z0L�Z1L

Z1L

� �24

35

¼ ðZ1L þ Z1HÞ þ Z1H

IH þ IA0HZ0L�Z1L

Z1L

� �IG þ IA0G

Z0L�Z1L

Z1L

� �24

35 ðC-5Þ

C.2 Infeed Effect on Type SDG and LDGGround-Distance Relays

Refer again to Figure C-1.

�VA1 þ 2IA1GZ1L þ 2ðIA1G þ IA1HÞZ1H

� VA0 þ IA0GZ0L

þ ðIA0G þ IA0HÞZ0H � VA2 ¼ 0 ðC-1ÞVAG ¼ VA1 þ VA2 þ VA0

ðvoltage at relay locationÞ ðC-2Þ

VAG ¼ Z1L 2IA1G þ IA0Z0L

Z1L

� ��

þ Z1H

�2ðIA1G þ IA1HÞ

þðIA0G þ IA0HÞ Z0L

Z1L

� ��

¼ Z1L IG þ Z0L � Z1L

Z1L

� �IA0G

� �

þ Z1H

�ðIG þ IHÞ

þ ðIA0G þ IA0HÞ Z0L � Z1L

Z1L

� ��ðC-3Þ

LDG balances when

VX1 þ VX2 þ VX0 ¼ 0 ðC-6Þ

where VX1, VX2, and VX0 are relay-compensatedvoltages, and

VX1 ¼ VA1 � IA1GZ1C

VX2 ¼ VA2 � IA1GZ1C

VX0 ¼ VA0 � IA0GZ0C ¼ VA0 � IA0GpZ1C;

p ¼ Z0C

Z1C¼ Z0L

Z1L

or

VA1 ¼ VX1 þ IA1GZ1C

VA2 ¼ VX2 þ IA1GZ1C

VA0 ¼ VX0 þ IA0GpZ1C ðC-7Þ

From Eqs. (C-3), (C-6), and (C-7), we obtain

2VX1 þ 2IA1GZ1C þ VX0 þ IA0GpZ1C

¼ Z1L IG þ IA0GZ0L � Z1L

Z1L

� ��

þ Z1H ðIG þ IHÞ þ ðIA0G þ IA0HÞ Z0L � Z1L

Z1L

� ��

Figure 12C-1 Infeed effect on ground distance relays.



Z1L

� ��


Z1L

� ��

Z1C IG þ IA0GZ0L � Z1L

Z1L

� ��


Z1L

� ��


Z1L

� ��

Z1C ¼ Z1L þ Z1H

ðIG þ IHÞ þ ðIA0G þ IA0HÞ Z0L�Z1L

Z1L

� �IG þ IA0G

Z0L�Z1L

Z1L

� �24

35

Z1C ¼ Z1L þ Z1H

IG þ IA0GZ0L�Z1L

Z1L

� �þ IH þ IA0H

Z0L�Z1L

Z1L

� �IG þ IA0G

Z0L�Z1L

Z1L

� �24

35

Z1C ¼ ðZ1L þ Z1HÞ

þ Z1H

IH þ IA0HZ0L�Z1L

Z1L

� �IG þ IA0G

Z0L�Z1L

Z1L

� �24

35 ðC-8Þ

It is important to note that Eqs. (C-5) and (C-8) are thesame.

APPENDIX D: COORDINATION IN MULTIPLE-LOOP SYSTEMS

Each relay in a multiple-loop system is coordinated asdescribed in Section 2.2.2. Each loop must becoordinated within itself and with adjacent relays inother loops. This technique is illustrated through thefollowing example.

D.1 System Information

A typical 23-kV loop system has three-phase bus faultcurrents as shown in Figure 12D-1. The fault locationsare marked with an X. Each fault is indicated with acircled number, followed by the total maximum andminimum (with bracket) fault current in amperes at23 kV. The fault currents that are contributed fromeach circuit are also shown.

The forward/reverse maximum and minimum cur-rents through the breakers and relays can be calculatedfrom Figure 12D-1 and are shown in Figure 12D-2.Load currents for the 5- and 10-MVA transformers are125.5 and 251.0A, respectively. This means that themaximum load current flow on each line section couldbe 125.5A in either the forward or reverse direction.Therefore, all overcurrent units should be set higher

than ð26125:5Þ ¼ 250 primary A, or 6.25 secondary Afor a 200/5 ct application.

D.2 Relay Type Selection

D.2.1 Selection of Directional Time-OvercurrentRelay

Refer to step 1 in Section 3.6.2. Table 12D-1 isprepared for determining the need of directional time-overcurrent units. It shows that either a directional ornondirectional relay could be applied for D and E. If,however, the relay must provide backup protectionbeyond the remote bus, a directional relay should beused. Using a directional relay would also meanavoiding future problems caused by system changes.

D.2.2 Application of Instantaneous-Trip Relay

Follow step 2 in Section 3.6.2. Table 12D-2 is preparedfor determining the possibility of instantaneous-trip

Table 12D-1 Selection of Directional Time-Overcurrent

Relay

Breaker

location IRM IRL 4 (IRM or IRL) �� IFN

Directional

required

A 250 125 1000 750 Yes

B 1000 125 4000 210 Yes

C 1100 125 4400 180 Yes

D 200 125 800 800 Yes/no

E 200 125 800 800 Yes/no

F 1100 125 4400 180 Yes

G 1000 125 4000 210 Yes

H 250 125 1000 750 Yes

IRM¼ contributed maximum reverse fault current.

IRL¼ contributed maximum reverse load current.

IFN¼ contributed minimum forward bus fault current.

Table 12D-2 Application of Instantaneous-Trip Relay

Breaker

location ICM �� 2:5 IFM IFM

Instantaneous-

trip unit

applicable

A 2750 2500 1000 Yes

B 1100 500 250 Yes

C 1000 500 200 Yes

D 3200 2750 1100 Yes

E 3200 2750 1100 Yes

F 1000 500 200 Yes

G 1100 625 250 Yes

H 2750 2500 1000 Yes

ICM¼ contributed maximum close-in fault current.

IFM¼ contributed maximum forward bus fault current.

308 Chapter 12

unit application. It indicates that the instantaneous-trip unit can be set to underreach if applied to all thesebreakers.

D.2.3 Requires Directional Instantaneous-TripRelay

Follow step 3 of Section 3.6.2. Table 12D-3 is preparedfor determining the need of directionally controlledinstantaneous-trip overcurrent units.

D.2.4 Summary of Type of Relay Selection

Table 12D-4 is a summary of relay selection for eachbreaker.

D.3 Relay Setting and Coordination

D.3.1 Coordination Paths

As suggested in Section 5.2.1, a good starting point forthe coordination process is at the largest sources.

Figure 12D-1 Example of relay application and settings for loop system.


Therefore, the coordination paths for the system asshown in Figure 12D-1 can be attempted as below:

1. Coordinate B with H and I; D with B and L; Fwith D and J; and H with K and F.

2. Coordinate G with A and I; E with G and K; Cwith E and J; and A with C and L.

The first step is to define the protections on I and J, setrelays L and K based on the known data from relaysMand N, and then make coordinations for loops a and bin turn.

D.3.2 Protections on I and J

Assume that locations I and J provide transformer(either HU or CA-26 relays) and bus (either KAB or

Figure 12D-2 Maximum and minimum fault currents through each breaker.

Table 12D-3 Requiring Directional Instantaneous-Trip

Relay

Breaker

location IRM�� IFM

Required

directionally

controlled IT unit

A 250 1000 No

B 1000 250 Yes

C 1100 200 Yes

D 200 1100 No

E 200 1100 No

F 1100 200 Yes

G 1000 250 Yes

H 250 1000 No

IRM¼ contributed maximum reverse fault current.

IFM¼ contributed maximum forward bus fault current.

310 Chapter 12

CA-16 relays) differential protections. Their opera-tions should therefore be high-speed.

D.3.3 Protections on M and N

Assume that the phase relaysM and N are type HiLoCO-8 (refer to Fig. 12D-3 for time curves), with 1- to12-A taps, and with a 6- to 144-A IIT unit. Forpreventing operation on cold-load inrush, tap 10 and atime dial of 2 have been predetermined.

In this case, the equivalent fault current at 23 kV forthe time-overcurrent unit at a 10-A setting would be

106200

56

7:2

23¼ 125A ðat 23 kVÞ

Then the following data can be obtained from the timecurve in Figure 12D-3 for a time dial of 2:

Multiples of tap

setting

Primary current

at 23 kV

Time to close

contacts (sec)

1.0 125

2.0 250 4.00

2.5 312 2.38

3.0 375 1.65

4.0 500 1.05

5.0 625 0.78

7.0 875 0.58

10.0 1250 0.46

16.0 2000 0.36

Also, assume that the IIT unit setting of 18A has beenselected after analysis of the 7.2-kV circuits. Theequivalent fault current at 23 kV for a nondirectional

IIT unit at an 18-A setting would be

186200

56

7:2

23¼ 225A ðat 23 kVÞ

Plot this information in Figure 12D-4 as curvesM andIIT. They will be used for coordination with relays L.

D.3.4 Setting for Relay L (and K) to Coordinatewith Relays M and N

Since these relays energize a transformer bank, itsminimum pickup setting should be approximatelytwice the full load, and it should be coordinated withrelayM based on the maximum through-fault current.

The recommended ct ratio for relayL is 200% of thetransformer self-rating, which is 26125 ¼ 250, i.e.,select a ct ratio of 250/5. Therefore, the time-over-current unit should be set above

261255

250¼ 5A

Tap 5 is selected. The maximum through fault current is

12005

250

� �¼ 24A

which is 24/5¼ 4.8 times the tap setting. The operatingtime of N at 1200A, from Figure 12D-4, is 0.47 sec.Then set relay L for 0.47 plus 0.3 (one step of CTI), or0.77 sec. Using the curve from Figure 12D-3 for a 4.8multiple and 0.77 sec, we select time dial 2, which givesa time of 0.83 sec.

Alternatively, the multiple and time requirementscan be specified and the time dial set by actual currentin the relay. This latter method would yield a time dialof approximately 1.75, instead of 2. Using tap 5 (250Aprimary) and time dial 2, the curve points for the time-overcurrent unit are as follows:

Table 12D-4 Type of Relay Selection

Breaker location

Time-overcurrent

unit needing

directional control

Instantaneous-trip

unit can be set to

underreach

Instantaneous-trip unit

requiring directional

control

Type of relay

selected

A Yes Yes No CR with IIT

B Yes Yes Yes IRV

C Yes Yes Yes IRV

D Yes Yes No CR with IIT

E Yes Yes No CR with IIT

F Yes Yes Yes IRV

G Yes Yes Yes IRV

H Yes Yes No CR with IIT


Multiples of tap

setting

Primary current

at 23 kV

Time to close

contacts (sec)

1.0 250

2.0 500 4.00

2.5 625 2.38

3.0 750 1.65

4.0 1000 1.05

5.0 1250 0.78

7.0 1750 0.58

10.0 2500 0.46

Plot these points in Figure 12D-5 as curve L andcopy curveM from Figure 12D-4 to illustratethe setting of relay L to coordinate with relays Mand N.

The IIT instantaneous-trip unit for relay L shouldbe set at the highest of the following: (1) about four tosix times the transformer self-rating; (2) higher thanthe inrush current; or (3) 1.1 to 1.3 times the maximumthrough fault. Condition (3) would be the highest forthis example. For example, use a factor of 1.3 for thesetting calculation; then 1:361200 ¼ 1560A (at23 kV), or 1560=50 ¼ 31:2A (relay side). Set the IITunit at 32A (or 32650 ¼ 1600A at 23 kV). Relay K(on the bottom line of Figure 12D-2) is assumed tohave the same setting as L.

D.3.5 Setting for Relay B to Coordinate withRelays H and I

Based on the assumption in Section D.3.2 that locationI provides transformer and bus differential protection,their operation should be high-speed. However, thesetting for H is unknown and cannot be determineduntil all other relays in the loop are set andcoordinated.

However, since the relayH is at the end of the B,D, F,H loop, at least it will have 4 (CTI)¼ 1.2 sec for thesetting. Consider the benefit of IIT application; at thispoint, 1 sec for line faults out of H can be tried. Thecritical fault forB is (1)maximum, giving 250Aprimary.

The maximum forward-direction load flow currentwill be 125.6A if breakers I and F are opened. Theovercurrent unit at B should be set to 251.2A for thisload current. However, the maximum end-zone faultcurrent is only 250A. Therefore, from the point ofview of protection, operating the system with bothbreakers I and F opened at the same time is notallowed. Based on this conclusion, it can be assumedthat load will not flow through B to A, and since the

relay is directional, then there is no need to set theovercurrent unit at twice the maximum load current. Itcan be set based on continuous rating. The maximumcontinuous load is 125:6=40 ¼ 3:14A (relay side). Thecontinuous rating of any HiLo CO tap is higher thanthis value. Therefore, choosing a 1-A tap should be noproblem. Tap 1 corresponds to 1640, or 40A primary.The critical fault multiple then is 250=40 ¼ 6:26, atwhich value the operating time (CTI plus 1) should be0.3 plus 1, or 1.3 sec. From Figure 12D-3, for a 6.26multiple and 1.3 sec, the time dial is approximately 4. Atime dial of 1 may be used, or the value determined bya test. Using tap 1 (40A primary) and a time dial of 4,we can arrive at the curve points for Figure 12D-6:

Multiples of tap

setting

Primary current

at 23 kV

Time to close

contacts (sec)

1.0 40

3.0 120 3.40

4.0 160 2.10

5.0 200 1.58

10.0 400 0.95

14.0 560 0.80

20.0 800 0.70

The instantaneous-trip unit of IRV-8 is a cylinderelement. It can be set for 1.106 250, or 275A primary(6.88A secondary). A setting of 7 or 280A primary willprovide high-speed, sequential tripping. After A opens,the fault current increases from 420 to 380A, which isabove the instantaneous unit pickup of 280A.

D.3.6 Setting for Relay D to Coordinate withRelays B and L

Since relay L operates instantaneously for the max-imum fault (2), the critical fault is (2) minimum.Critical fault current is 1550A for relay L and 800Afor relay D. Relay D is set for a minimum pickup of26 125, or 250A primary (6.25A secondary). Tap 7 isselected to give a minimum pickup of 76 40, or 280A.The critical fault multiple is 800/280, or 2.85 for relayD. The operating time of relay L for the critical faultvalue of 1550A is 0.65 sec at the 1550/280¼ 5.52multiple. The set point is 0.65 plus 0.30, or 0.95 sec at800A. From Figure 12D-3, for a 2.85 multiple and0.95 sec, the time dial is approximately 1.10. The 1.25value may be used, or the time dial determined by a

312 Chapter 12

Figure 12D-3 Typical time curve of the type CO-8 or IRV-8 relay.


Figure 12D-4 RelayM settings.

314 Chapter 12

Figure 12D-5 RelayL settings.


Figure 12D-6 Relay B settings.

316 Chapter 12

test. Using tap 7 (280A primary) and a time dial of1.25, we can find the curve points for Figure 12D-7:

Multiples of tap

setting

Primary current

at 23 kV

Time to close

contacts (sec)

1.0 280

2.0 560 2.50

2.5 700 1.55

3.0 840 1.00

4.0 1120 0.600

5.0 1400 0.45

7.0 1860 0.35

Finally, the IIT of CR-8 is set to 1.36 1100, or 1430A(36A secondary).

D.3.7 Setting for Relay F to Coordinate withRelays D and J

Based on the assumption in Section E.3.2 that locationJ provides transformer and bus differential protection,their operations should be high-speed.

The maximum forward-direction load flow currentwill be 125.6A if breakers J and B are opened. Theovercurrent unit at F should be set to 251.2A for thisload current. However, the maximum end-zone faultcurrent is only 200A. Therefore, from the point ofview of protection, operating the system on bothbreakers J and B opened at the same time is notallowed. Based on this conclusion, it can be assumedthat load will not flow through F to E, and since therelay is directional, then there is no need to set theovercurrent unit at twice the maximum load current.It can be set based on a continuous rating. Themaximum continuous load is 125.6/40¼ 3.14A (relayside). The continuous rating of any HiLo CO tap ishigher than this value. Therefore, choosing a 1-A tapshould be no problem. Since the IIT unit of the Drelay will operate for fault (3) maximum, the criticalfault for relay F is fault (3) minimum. The overcurrentunit of the D relay will operate in 0.55 sec (1180/280¼ 4.22 multiple).

For relay F, tap 1 corresponds to 16 40, or 40Aprimary. The critical fault multiple is 180/40 or 4.5, atwhich value the operating time should be 0.55 plus 0.3,or 0.85 sec. From Figure 12D-3, for a 4.5 multiple and0.85 sec, the time dial is approximately 2. Using tap 1(40A primary) and a time dial of 2, we can find thecurve points for Figure 12D-8:

Multiples of tap

setting

Primary current

at 23 kV

Time to close

contacts (sec)

1.0 40

2.5 100 2.38

3.0 120 1.65

4.0 160 1.05

5.0 200 0.78

7.0 280 0.58

10.0 400 0.46

16.0 640 0.36

The instantaneous-trip unit of the IRV-8 is acylinder element. It can be set for 1.106 200, or220A primary (5.5A secondary). A 6-A setting or240A primary current will provide high-speed, sequen-tial tripping. After E opens with a fault near E on lineFE, the fault current increases at F to 300A which isabove the instantaneous unit pickup of 220A.

D.3.8 Setting for Relay H to Coordinate withRelays K and F

Since relay K operates instantaneously for the max-imum fault (2), the critical fault is (2) minimum. Criticalfault current is 1550A for relay K and 750A for relayH. Relay H is set for a minimum pickup of 26 125, or250A primary (6.25A secondary) load current. Tap 7 isselected to give a minimum pickup of 76 40, or 280A.

The critical fault multiple is 750/280, or 2.68 forrelay H. The operating time of relay K for the criticalfault value of 1550A is 0.70 sec (1550/280, or a 5.52multiple). The set point for relay H should be 0.70 plus0.30, or 1 sec at 750A. From Figure 12D-3, for a 2.68multiple and 1 sec, the time dial is 1. Using tap 7 (280Aprimary) and a time dial of 1, we can find the curvepoints for Figure 12D-9:

Multiples of tap

setting

Primary current

at 23 kV

Time to close

contacts (sec)

2.0 560 2.1

2.5 700 1.2

3.0 840 0.8

4.0 1120 0.5

5.0 1400 0.38

7.0 1960 0.30

The instantaneous-trip unit of the IRV-8 is acylinder element. It can be set for 1.106 1000, or


Figure 12D-7 Relay D settings.

318 Chapter 12

Figure 12D-8 Relay F settings.


Figure 12D-9 Relay H settings.

320 Chapter 12

1100A primary (27.5A secondary). Settings arerounded off to 28.0A (1120A primary).

D.3.9 Check Relay Settings on Coordinationwith Relay B

The instantaneous unit setting of relay H is 1120Aprimary. This means that relay H will operate for linefaults (2) under both maximum (2750A) and minimum(1410A) conditions. On tap 7, with a time dial of 1, theH relay time unit operates for line faults (2) in 0.25 sec(maximum fault condition) or 0.38 sec (minimum faultcondition). Both values are well within the 1-sec timeused in the coordination of B (Sec. D.3.5).

It should be noted that the major factor slowingdown the H relay setting is the speed of relay K (Fig.12D-9). A better speed may be obtained for the loop ifthe performance of relay K (and L) is improved.

D.3.10 Settings and Coordinations for Relays G,E, C, and A

For this particular example, from Figure 12D-1, thesystem configuration and fault currents are corre-spondingly symmetrical between A, B, C, D and H, G,F, E. Therefore, the settings for relays G, E, C, Ashould be correspondingly the same as those for relaysB, D, F, and H.

D.3.11 Summary

This example has been used to describe an applicationof inverse-time directional overcurrent relays. Thoughelectromechanical relays have been referred to, iden-tical functions and curve shapes are available in solid-state and microprocessor versions.


13

Backup Protection

Revised by: E. D. PRICE

1 INTRODUCTION

Backup relaying, which provides necessary redun-dancy in protective systems, is defined in the IEEEStandard Dictionary as ‘‘protection that operatesindependently of specified components in the primaryprotective system and that is intended to operate ifthe primary protection fails or is temporarily out ofservice.’’

Backup protection includes remote backup, localbackup, and breaker-failure relaying. Breaker failure isdefined as a failure of the breaker to open or interruptcurrent when a trip signal is received.

Backup protection for equipment such as genera-tors, buses, and transformers usually duplicatesprimary protection and is arranged to trip the samebreakers. In the event of a breaker failure, some remoteline protection would isolate the fault.

In the past, backup protection for transmissionlines was provided by extending primary protection toline sections beyond the remote bus. This remotebackup is defined in the IEEE Dictionary as ‘‘backupprotection in which the protection is at a station orstations other than that which has the primaryprotection.’’

With the advent of EHV and increased concernabout both service continuity and possible breakerfailures, local backup, including breaker-failure pro-tection, has become common.

2 REMOTE VS. LOCAL BACKUP

2.1 Remote Backup

Circuit breakers occasionally do fail to interrupt or tripfor various reasons, and the remote terminal relays andbreakers may be able to provide backup for a failedbreaker. A remote backup system is shown inFigure 13-1. The relays protecting line GH at bus Gfor breaker 1 must also overreach and protect all otherlines extending from the remote bus H. That is, therelays at G must operate selectively for faults on linesHR, HS, and HT if the relays or breaker 2, 3, or 4,respectively, fail to clear the associated line fault. Thus,the relays on breaker 1 provide primary protection forline GH, as well as backup protection for lines HR,HS, and HT. Backup systems use time discriminationto detect faults in the remote line sections.

In designing remote backup systems, their selectiv-ity, sensitivity, speed, and application must be con-sidered.

2.1.1 Selectivity

Remote backup provides poor selectivity. It interruptsall tapped loads on the unfaulted line sections. Open-ing breaker 1 at G for faults on a remote section mayinterrupt service to tapped loads on line GH unne-cessarily. For a breaker failure at bus H, all linesfeeding fault power through the defective breaker must

323

be opened at their remote ends with remote backup.Such a scheme interrupts all loads on the lines, as wellas on bus H. Thus, if breaker 2 fails to open for a lineHR fault, breakers at G, S, and T must be opened withremote backup.

2.1.2 Sensitivity

Remote backup provides poor sensitivity. Relays atbus G may not ‘‘see’’ faults near buses R, S, or T. Forthe scheme shown in Figure 13-1, the fault infeed effectat bus H for faults near R, S, or T tends to reduce thecurrent magnitude and ‘‘reach’’ of distance relays atbreaker 1 at bus G. In these cases, relays at otherremote terminals will have to trip first, redistributingthe fault currents and increasing the effective reach ofthe relays at bus G. This can result in sequentialtripping.

2.1.3 Speed

Remote backup must be relatively slow to give theprimary relays in the remote line time to clear theirfault (coordinating time interval in Figs.13-2 and 13-3).As the coordinating time interval is typically 0.3 sec,backup times greater than 20 cycles are common. Ifsequential tripping is necessary, as indicated in thepreceding paragraph, the fault-clearing time for thebreaker for a remote backup must be further increased.

2.1.4 Application

The application and setting of relays for remotebackup require an understanding of fault levels underall possible operating conditions.

2.2. Local Backup and Breaker Failure

There are various reasons for a circuit breaker to fail tointerrupt or trip. The need for remote backup, localbackup, or breaker-failure relaying depends on theconsequences of such failure.

2.2.1 Local Backup

Unlike remote line protection, local backup is appliedat the local station. If the primary relays fail, localbackup relays will trip the local breakers. If the localbreaker fails, either the primary or backup relays willinitiate the breaker-failure protection to trip otherbreakers adjacent to the failed breaker. Although localbackup protection has many advantages and is widelyused, it does not automatically eliminate the need forremote backup.

Figure 13-2 Remote backup at bus G, breaker 1 for line

HR with inverse time overcurrent relays.

Figure 13-3 Remote backup at bus G, breaker 1 for line

HR with distance relays.

Figure 13-1 Remote backup at bus G, breaker 1 requires

these relays to selectively operate for faults on lines HR, HS,

and HT under all operating conditions.

324 Chapter 13

Local backup and breaker-failure protection arecharacterized by fault detection and initiation oftripping at the local terminal. For example, if a faulton line HR (Fig. 13-1) is not properly cleared by theprimary protection system because of a failure in anypart of the system other than the circuit breaker, thelocal secondary relaying system will detect the faultand trip breaker 2. If the fault on line HR is notproperly cleared because of a failure of breaker 2, thenthe primary and/or secondary protective relays willinitiate local breaker-failure backup to open breakers3, 4, and 5 at bus H. Figure 13-4, a block diagram,illustrates the difference between the remote backup,local backup, and breaker-failure schemes.

A protection system involves a number of elements,including protective relays; ac current transformersand wiring; ac voltage transformers, devices, andwiring; dc supply and wiring; circuit breakers or otherdisconnection means; and a communication channelwith pilot relaying. Ideally, a backup protection systemshould duplicate all these elements to provide totalredundancy. In practice, all elements except the circuitbreaker can be and frequently are duplicated in avariety of combinations, depending on the degree ofprotection required. The circuit breakers are not andcannot be duplicated. Many modern breakers have twoindependent trip coils, and breaker-failure protectionprovides a duplicate function.

The maximum practical redundancy is separatingtwo protection systems as shown in Figure 13-5. Thecommon element is the circuit breaker; even so,separate trip coils are shown. If a common stationbattery is used, separate fused leads from the batteryare used for the two protection systems. Quitefrequently, only one voltage source is used, and

separate fuses leads are run from the voltage transfor-mer to the relays.

Instantaneous relays can be considered an indepen-dent protective system. Because these relays do notfully cover the line section, however, a remote end-zone fault would require an additional protectionsystem. Local backup is usually applied only on linesequipped with a primary pilot system backed up by asecond pilot system or nonpilot backup relays, or both.

2.2.2 Basic Schematic and Operation

Figures 13-6 and 13-7 illustrate basic local backup andbreaker-failure schemes for electromechanical andlogic-based relays, respectively. The operating princi-ple is the same for either scheme. In Figure 13-6,operation of one of the protective relay systems tripsits associated breaker and energizes the breaker-failureinitiation relays 62X and/or 62Y, which are auxiliarySG, MG-6, or AR types of relays. Contacts on 62Xand 62Y operate the timer 62BF only if the instanta-neous overcurrent relay 50 indicates that current iscontinuing to flow. This continued current flow,indicating a failure to clear the fault, causes the timerto energize the multicontact 86BF lockout relay,tripping all the adjacent breakers. The 86BF lockoutrelay may also be used to block reclosing, to stop thecarrier with blocking-type pilot relaying so that theremote end can trip, if possible, and to initiate transfer-trip.

Figure 13-4 Block diagram illustrating the differences

between remote backup, local backup, and breaker failure

schemes.

Figure 13-5 Transmission line protection with maximum

practical redundancy.

Backup Protection 325

Timer 62BF should be energized with device 50(Fig. 13-6) rather than through the breaker auxiliarycontacts 52a, since these contacts may be open, whilethe contacts in a damaged breaker are closed.Alternatively, breaker 52a contact can be used togetherwith the 50 relay (Fig. 13-7).

Since transformer faults may not provide sufficientcurrent to operate device 50, a transformer differentialrelay auxiliary contact 86T can be used instead,

supervised by the breaker 52a contact to operate timer62BF (Fig. 13-8).

The three overcurrent units (50) are normallyconnected in phases A and C and to ground, but canbe connected in all three phases if necessary. The phaseunits must be set below minimum fault current and canbe set below load current, if necessary. A setting aboveload current is preferred in the interest of security if theminimum phase-fault current substantially exceeds themaximum load current. Setting above maximum loadhelps to prevent undesirable tripping during testing.

2.3 Applications Requiring Remote Backup withBreaker-Failure Protection

When ring buses or breaker-and-a-half schemes areused, breaker-failure protection does not necessarilyeliminate the need for remote backup: As shown inFigure 13-8, a fault on line HR, for example, requirestripping both breakers 2 and 3 at station H. If breaker2 fails to clear the fault, breaker failure would initiatethe tripping of breaker 5, but would leave line GH stillconnected to the fault. The breaker-failure protectionfor breaker 2 frequently initiates the transfer-trippingof breaker 1 at station G. If transfer-trip is not appliedor is not operative, however, remote backup at breaker1 is still required to clear the fault.

Because of the infeed effect and high apparentimpedances, remote backup from the remote stationsmay be difficult, if not impossible, to achieve when alllines are in service. Opening the breakers around thefailed breaker will, however, remove the infeed effectand permit remote backup coverage. If, for example,breaker 2 fails for a fault on line HR (Fig. 13-8), lineprotection will open breaker 3 and breaker-failureprotection will open breaker 5 to remove all infeedsaround station H, except that from line GH.

Breaker-failure protection would trip both breakers5 and 6 upon the failure of breaker 4 for a line HS fault

Figure 13-6 Simplified dc protection schematic for breaker

failure and local backup protection. (Alternating current

circuit per Figures 13-5, 13-13, or 13-14.)

Figure 13-7 Simplified logic diagram for logic-based

breaker failure and local backup protection. (Alternating

current circuit per Figures 13-5, 13-13, or 13-14.)

Figure 13-8 Remote backup required with breaker failure

at station H where ring bus or breaker and a halfschemes are

used.

326 Chapter 13

(Fig. 13-8). Similarly, the failure of breaker 5 inFigure 13-8 for a line GH fault would result in trippingboth breakers 4 and 6. All other breaker failureconditions of breakers 4 or 5 would require remotebackup, at S or G, respectively, or transfer-trip.

3 BREAKER-FAILURE RELAYINGAPPLICATIONS

Breaker-failure relaying should not be considered asubstitute for good system design and equipmentmaintenance.

Breaker-failure protection should be as fast aspossible without tripping unnecessarily. This criterionis particularly important in EHV lines, where stabilityis critical. Here, breaker-failure timer settings of 100to 250msec (6.5 to 15 cycles on a 60-Hz basis) areused.

These critical operations require dual high-speedsolid-state or microprocessor pilot and breaker-failureprotections. With electromechanical-type relays,breaker-failure timer settings of around 250msec ormore are practical.

In applying breaker-failure protection, it is recom-mended that

One breaker-failure circuit per breaker be applied,regardless of the bus configuration.

All adjacent breakers should be tripped by breaker-failure protection, regardless of fault location.

In all cases, all breakers tripped by the breaker-failure scheme should be locked out.

A remote breaker must also be tripped by either itsown relays or transfer-trip initiated by localbreaker-failure protection. Remote backup clear-ing of the breakers may be preferred over thedirect-transfer trip.

One timer per bus or one timer per breaker may beused.

The latter is recommended, since it provides maximumisolation and flexibility, even though it does involveadditional timers. These methods will be illustrated forvarious bus arrangements.

Circuit-breaker auxiliary switches should not beused to indicate whether or not a circuit breaker iscarrying current unless there is no other way toaccomplish this.

The fact that an auxiliary switch has operated is notsufficient proof that a circuit breaker has interrupted afault. The auxiliary switch may be opened because

1. Its operating linkage is broken.2. It is out of adjustment.3. The breaker mechanism has operated but the

main breaker contacts have failed to interruptthe current.

When protective relays are being tested, the breaker-failure scheme should be properly blocked or isolated toprevent misoperation.

3.1 Single-Line/Single-Breaker Buses

A typical single-line/single-breaker bus is shown inFigure 13-9. Figures 13-10 and 13-11 illustrate the dcschematics or logic diagrams for breaker-failure/localbackup protection using one timer per bus section.

Figure 13-12 shows the schematic for one timer perbreaker. The two methods, one timer per bus (method1) and one timer per breaker (method 2), have thefollowing differences:

Method 1 is less costly than method 2, since fewertimers are required.

Figure 13-9 A single-line single-breaker bus. (The breaker

failure local backup protective scheme is shown in Figures 13-

10 and 13-11.)


Transfer-trip of the remote breaker is easier withmethod 2. With method 1, the common timercannot distinguish which breaker has failed.

An evolving fault may cause incorrect breaker-failure operation with method 1. If a line 1 faultevolves to line 2, with sequential operation of62X, 62Y, and 50 contacts, the common timercircuit may be enrgized long enough to operateand trip all breakers, even though both line 1 and2 breakers trip normally. With method 2, eachtimer is deenergized as soon as the associated linefault is cleared.

The common timer approach of method 1 requiresthat the timer be set for the slowest breakerinterrupting time. Method 2 permits the separatetimers to be set for the interrupting times of theindividual breakers.

3.2 Breaker-and-a-Half and Ring Buses

Typical breaker-and-a-half and ring buses are shownin Figures 13-13 and 13-14. These arrangementsrequire tripping two breakers and paralleling thecurrent transformers for each line, as shown. A currentdetector (50) is provided for each breaker. One timerper breaker is recommended for all these bus config-urations. Breaker-failure protection systems are shownin Figures 13-6 and 13-7. Another set of 62X, 62Yauxiliaries (shown dotted) must be added for thesecond breaker.

The breaker-failure/local backup circuits are thesame for each breaker, except for the application of the86 relay contacts. The 86BF relay operations areoutlined in Table 13-1 for Figure 13-13 and in Table 13-2 for Figure 13-14. Neither table includes reclosinglookout, which may be desired.

It will be noted from Tables 13-1 and 13-2 that alladjacent breakers are tripped, regardless of fault

Figure 13-10 Typical simplified dc schematic for breaker

failure local backup protection using a common timer for a

single-line single-breaker bus of Figure 13-9.

Figure 13-11 Typical simplified logic diagram for breaker

failure local backup using a common timer for a single-line

single-breaker bus of Figure 13-9.

Figure 13-12 Typical simplified dc schematic for breaker

failure local backup protection using one timer per breaker

for the single-line single-breaker bus of Figure 13-9. (The

solid-state logic diagram would be as shown in Figure 13-7

for each breaker.)

328 Chapter 13

location. For simplicity and reliability, breakers thatare already tripped will be retripped. Assume, forexample, that a fault occurs on line 1 of Figure 13-14.The protective relays for line 1 will attempt to tripbreakers 1 and 4 and the remote end of line 1. Assume

breaker 1 fails to clear, but breaker 4 and the remoteline 1 breaker do open. Then breaker failure need notretrip 4 and transfer-trip line 1, as shown in Table 13-2.Similarly, for a fault on line 2, it may be unnecessary toretrip breaker 2 and transfer-trip line 2.

It is simpler, however, to trip all breakers involved,and it also allows reclosing to be blocked throughout.This practice also provides symmetrical protectionaround the bus.

4 TRADITIONAL BREAKER-FAILURE SCHEME

4.1 Timing Characteristics of the TraditionalBreaker-Failure Scheme

Figures 13-6 and 13-7 show the approach of traditionalbreaker-failure schemes. The operating principle ofthis approach as illustrated in Figure 13-15 is that thebreaker-failure timer 62BF is started by the operations

Figure 13-13 Single-line diagram of a breaker and a half

bus.

Figure 13-14 Single-line diagram of a ring bus.

Table 13-1 Breaker-Failure Operations for Breaker-and-a-

Half Bus (Fig. 13-13)a

For local backup or

breaker failure no. 86 Relay operations

1 Trip 2 and all other bus breakers,

such as 4, etc., on bus L. Transfer-

trip line 1.

2 Trip 1 and 3. Transfer-trip lines 1 and

2.

3 Trip 2 and all other bus breakers,

such as 5, etc., on bus R. Transfer-

trip line 2.

aSee schematics for Figures 13-6 or 13-7.

Table 13-2 Breaker-Failure Operations for Ring Bus

(Figure 13-14)a

For local backup or

breaker failure no. 86 Relay operations

1 Trip 2 and 4. Transfer-trip line 1 and

2.

2 Trip 1 and 3. Transfer-trip lines 2 and

3.

3 Trip 2, 4, and 5. Transfer-trip line 3.

4 Trip 1, 3, and 5. Transfer-trip line 1.

a See schematics for Figures 13-6 or 13-7.


of the fault detector 50 and the breaker failure initiatessignal 62X or 62Y.

A time chart for the traditional breaker-failurescheme is given in Figure 13-16. The shaded margintime provides security and should accommodate thefollowing variables:

1. Excessive breaker interrupting time. Accordingto ANSI Standard C37.04, the circuit-breaker inter-rupting time may be one cycle longer for three-cyclebreakers at currents below 25% of the maximumrating. Also, the interrupting time may be longer onclose-open duty.

2. Inconsistency in BFI times. These are mini-mized by static and microprocessor breaker-failureinitiation. However, the wide time range associatedwith electromechanical BFI is primarily a function ofthe variation in dc voltage. For example, the BFI ARcontact output in the SRU relay or ARM module(Uniflex) has an operating time of 2 to 5msec. The BFItelephone relay in the SRU relay has an operating timeof 8 to 16msec. The SG relay 62X/62Y in the KC-4/TD-5 breaker-failure scheme has an operating time of16 to 50msec. The breaker-failure total clearing time,as shown in Figure 13-16, assumes that the pickup ofdevice 50 is equal to the protective relay plus the BFItime. Either 50 or BFI signal delaying will delaystarting the 62BF timer (Fig. 13-15) and the totalbackup breaker-failure clearing time will be longer andthe margin increased.

3. Overtravel of the 62BF timer after the faultdetector reset. Static timers have less than 1msec ofovertravel. In microprocessor relays the overtravelreally depends on the logic and operation design. Theovertravel is usually zero, but the worst case expectedwould be the computation interval time (the timebetween logic calculations).

4. Inconsistency in 62BF timer. The static timer,as contained in the TD-5, TD-50, and SBFU, has arepeatability (including variations of ambient tempera-ture and voltage supply) of +5%, which is +5% msecfor a timer setting of 100msec (six cycles on a 60-Hz

base). Microprocessor timers are generally consistentto within the computation interval time.

5. 62BF timer setting error. Includes humanerror, instrumentation error, and potentiometer reso-lution. The static timer may be set within 2ms. Themicroprocessor timer can generally be set to within theresolution of the logic computation interval time.

6. A safety factor. Because of the widespreadharmful effects of a false 86BF operation, it isrecommended that a generous safety factor be incor-porated in the margin time. The degree of safetyrequired is a direct function of the confidence level ofthe total protective system. Typical values range fromtwo to six cycles (60-Hz base). A two-cycle safetyfactor appears to be adequate, with three cycles being awidely used total ‘‘margin.’’

4.2 Traditional Breaker-Failure RelayCharacteristics

Four types of relays are used in the traditional breaker-failure scheme: current-detector relays, timer relays,auxiliary relays, and multitrip auxiliary relays.

4.2.1 Fault-Current Detector (50BF)

Fault detectors that have high dropout and whosedropout time is minimally affected by ct saturation anddc decay in the secondary circuit should be considered.Examples of this type are cylinder unit relays and staticand microprocessor relays with suitable filtering.Hinged armature and plunger-type relays that canhave a significant dropout delay should be consideredcarefully before use. If such devices are used, theirdropout time could be ascertained under the worstconditions and this time should be considered whensetting the breaker-failure timer. Determining theworst condition may be difficult.

A type KC-4 relay can be used for the fault-currentdetector in an electromechanical breaker-failurescheme. It is a high-speed unit with a 98% or greaterratio of dropout to pickup. The KC-4 relay has three-cylinder-type overcurrent units in an FT case. It isavailable in 0.5- to 2.0-, 1- to 4-, 2- to 8-, 4- to 16-, 10-to 40-, and 20- to 80-A ranges. The pickup isapproximately 19 to 24msec at four times pickupcurrent, and 12 to 16msec at eight times pickup.Dropout times are on the order of 20msec after thecurrent decreases to 0. Various combinations of rangesfor phase and ground applications are available.Circuit shield types 50H and 50B are static relayswith similar characteristics.

Figure 13-15 Block diagram of the traditional breaker

failure scheme.

330 Chapter 13

4.2.2 Timer (62BF)

A direct-current static timer, type TD-5, is used in theelectromechanical scheme with contact input, as shownin Figure 13-6. The range, for the breaker-failureapplications is 0.05 to 0.4 sec. Static timers, types 62Band 62T are generally used with the static faultdetectors.

4.2.3 Auxiliary Relay (62X, 62Y, or BFI)

Operating times are 16 to 50msec for the SG and MGrelays, and 2 to 3msec for the AR relays.

4.2.4 Multitrip and Lockout Auxiliary Relay(86BF)

Either the electroswitch type WL relay or type LORcan be used; their operating times are approximately16msec.

4.3 Microprocessor Relays

The microprocessor breaker-failure function is gener-ally implemented in line relays in two levels. The firstlevel provides only a breaker-failure initiate signal(BFI, 62X, 62Y) simultaneously with the breaker trip.

Figure 13-16 Time chart of the traditional breaker failure scheme (time not to scale).


In this case there is no 62X or 62Y operating time toconsider.

The second level is where the complete breaker-failure tripping function is included in the line relay.Quite often special design considerations—filteringand algorithms—are implemented to assure fast andsecure dropout of the 50-fault detector. These timeswill be specified and generally range from one-half toone cycle. It is generally safe to assume that themaximum dropout time will be one cycle plus thecomputation interval time. That is, if the status of the50-fault detector is computed every one-quarter cycle,then the maximum dropout time will be 1.25 cycles.This assumes the appropriate dc filtering that isexpected in microprocessor relays.

The setting resolution of the 62BF timer will also bedefined by the computation interval time of the 50-fault detector and the frequency with which the timer iscompared with the fault detector.

5 AN IMPROVED BREAKER-FAILURE SCHEME

5.1 Problems in the Traditional Breaker-FailureScheme

5.1.1 Fault-Detector Reset Time Problem

Referring to Figure 13-16, we see that the reset time ofdevice 50 is a critical characteristic in the performanceof the breaker-failure system. It will affect the marginvalue, as well as the 62BF timer setting. Device 50resetting stops the timer; with 62X/62Y seal-in, it is theonly function that stops the timer. It should be notedthat the longer the reset time of device 50, the longerthe 62BF time has to be set. Consequently, it may bedifficult to set the timer securely and still avoid systeminstability.

To prevent the device 50 reset time variables fromaffecting the margin, it is recommended that the 62BFtime delay be determined for maximum device 50 resettime. Any faster 50 reset will then merely add to thesafety margin.

The reset time of the overcurrent fault detector isaffected by several factors:

The reset times are longer when the current afterinterruption is nonzero. Certain types of circuitbreakers are equipped with arcing contacts andshunting resistors. When the breaker maincontact interrupts fault current, the currentdoes not drop immediately to 0, but to a leveldetermined by the shunting resistor. It falls to 0

when the arcing contacts open. The reset time ofdevice 50 on such applications may be longer.

The fault-current level at which the unit is energizedprior to interruption.

The setting of the unit.Current transformer decay. When breaker current is

interrupted at current zero, the current transfor-mer magnetizing current will not be zero. Thiscurrent would then have to decay through theconnected relays, possibly increasing their resettime.

The reset time of a solid-state overcurrent relay isfaster than an electromechanical one.

5.1.2 Fault-Detector Setting Problem

Figure 13-17 illustrates a problem in the traditionalbreaker-failure scheme for breaker A on a ring busapplication where maximum load exceeds minimumfault levels. Because IF1> IF2, it may be difficult to setI< (IF2)/2 in the traditional single-fault-detectorscheme. One solution to this problem (Fig. 13-17) isto use a special two-fault-detector breaker-failurescheme. The low set with 86T (from the transformerdifferential scheme) is for a fault at F2, whereas thehigh set with 62X (from the line protection) is for faultF1.

Figure 13-18 illustrates another problem in thetraditional breaker-failure scheme for breaker B in abreaker-and-a-half system application. If breaker Bfails for a traditional breaker-failure scheme, the 62BFtimer for breaker B may not start (because initi-al¼ 10% may be lower than the fault-detector setting)until breaker A opens (because sequential¼ 100%).This condition results in unduly long backup clearingtime. One solution to this problem is to use a special

Figure 13-17 Traditional BF scheme with two timers.

332 Chapter 13

breaker-failure scheme as shown in Figure 13-18. Itincludes two timers and two overcurrent units. How-ever, the shorter timer setting or lower overcurrent unitsetting may affect the security of the traditionalbreaker-failure scheme.

These problems can be solved by applying the typeSBF-1 breaker-failure relay. It uses an improvedconcept as described in Section 5.2.

5.1.3 Transformer Breaker-FailureConsiderations

A low-level fault internal to the transformer may causethe sudden pressure relay (SPR) to operate. Thecurrent level may be well below a ‘‘low set’’ valuethat the overcurrent relay could detect. For this case,an 86T contact in series with the circuit breaker’s 52a(logically 86T AND 52a) as shown in Figure 13-19must be used to assert and start the timer. In this case if

tripping occurs by the SPR and the breaker fails totrip, breaker failure tripping occurs. The drawback isthat if the breaker trips successfully and the 52acontact does not open, breaker failure clearing stilltakes place.

5.2 The Improved Breaker-Failure Scheme

Many approaches have been devised for improving thetraditional breaker-failure scheme such as using ashorter reset time of device 50 (for example, using astatic relay), or employing separate timers for differentlevels of fault current. These approaches will still beaffected by the reset time of the device 50 and stillpresent difficulties in determining the reset time,especially when information on the shunting resistorcurrent is not available.

Figure 13-20 shows the basic concept of theimproved approach. The unique feature of thisapproach is that the fault detector operates only afterthe 62BF times out. By that time, the breaker-failurecondition has been confirmed; consequently, the resettime of the fault detector will have no effect on thescheme. Figure 13-21 is the time chart for the improvedbreaker-failure scheme.

The improved approach has many advantages overthe traditional ones:

1. The fault detector 50 will not operate before the62BF is timed out; therefore, it will neveroperate when clearing normally and device 50reset time is not a consideration.

2. It permits shorter margin and shorter overallclearing times and will wield a net saving of oneto two cycles over the traditional approach.This can be illustrated as below (compare Fig.13-16 and 13-21): Net saving in clearing time -¼maximum 50 reset time, which representsabout one cycle for static units and two cyclesfor electromechanical units.

3. The overcurrent-fault-detector unit never oper-ates when clearing normally so it can be set

Figure 13-18 Traditional BF with two timers and two fault

detectors.

Figure 13-19 Transformer BF considerations. Figure 13-20 Logic diagram of the SBF-1 relay.


lower than load current, if necessary (forexample, application problems as described inSection 5.1.2).

Figure 13-16 Figure 13-21

Total clearing time

¼Protective relay time

Total clearing time

¼Protective relay time

þ breaker interrupting time þ breaker interrupting time

þmaximum 50 reset time

þmargin

þmargin

þ 86BF time

þ 86BF time

5.3 Type SBF-1 Relay

The type SBF-1 breaker-failure relay uses theimproved concept for breaker-failure detection, asdescribed in Section 5.2. The relay, as shown inFigure 13-22, consists of phase- and ground-faultdetectors, 62BF timer, and a ‘‘control’’ timer. Thecontrol timer is for resetting the SBF-1 relay. Inapplying the SBF-1 relay:

The fault detectors can be set below the loadcurrent.

Figure 13-21 Time chart of the improved breaker failure scheme.

334 Chapter 13

Figure 13-22 Type SBF-1 relay.


The 62BF timer should be set longer than thebreaker normal clearing time, with a margin oftwo to three cycles.

The control timer should be set at least 32mseclonger than the 62BF time setting to provide longenough time for the SBF-1 relay to trip thebreakers. The breakdown of this 32msec is asfollows (refer to Fig. 13-22):

Maximum pickup ofRR relay 1 to 3ms50 unit 3 to 8ms

AR relay 2 to 5ms

8<:

Total 16ms

86BF relay 16ms

Total 32ms

6 OPEN CONDUCTOR AND BREAKER POLEDISAGREEMENT PROTECTION

At voltages of 345 kV and above, the physical size ofthe operating components and the phase spacingrequirements of power circuit breakers have led tothe use of an independent operating mechanism foreach. Also, with the trend toward larger and largerturbine-generator sizes, system stability criteria haveoften dictated the use of independent pole-operatedbreakers at voltages below 345 kV to obviate the three-phase fault, three-pole ‘‘hung’’ breaker fault condition.Rarely will more than one fail to interrupt faultcurrent. Therefore, serious faults (those involving twoor three phases) will always be downgraded if notentirely cleared, reducing the danger of systeminstability.

Field experiences have shown it necessary toconsider the consequences of the unsymmetricaloperation of such a breaker. Electrical or mechanicalfailures have left one phase open when the others areclosed, and vice versa. Since pole disagreement mayoccur under a no-fault condition, the breaker-failureinitiation circuit of the conventional breaker-failurescheme may not be energized. Therefore, the conven-tional breaker-failure relay scheme will not respond ifthe failure occurs under a nonfault condition.

Figure 13-23 shows logic that can be used to detectpole disagreement under a no-fault condition. Thislogic has an output whenever one or more phasescarries current above the IH setting, while one or morephases carries current below the IL setting. The currentinputs IL, IH, and IM must be capable of measuringcurrent accurately in the 1.0 to 100.0mA range todistinguish between no current and charging current.This usually requires special current inputs to the relay.

For breaker-and-a-half or ring-bus configuration,with low-magnitude load current flowing through thebus, odd phase current combinations result fromunequal phase impedances and multiple paths forcurrent flow through the bus; unequal phase currentsflowing through an individual breaker can give thefalse appearance of pole disagreement. This logiccontains a zero sequence voltage comparison circuit(Fig. 13-23) that allows this low-current difference tobe ignored, while permitting tripping when a hazar-dous pole disagreement actually exits.

Simpler methods may be employed that utilizeconventional ct inputs that are only reliable withload flow. Typical minimum current levels that can beaccurately measured range from 0.25 to 0.5A (250 to500mA). These applications involve the measurementof zero or negative sequence current through the poles

Figure 13-23 Pole disagreement logic example.

336 Chapter 13

of the breaker, and may be used to trip or alarm. Thesequence current measured must distinguish betweenthat which is produced by normal unbalances and thatproduced by an open breaker pole. A long time delay isapplied to ride through maximum fault clearing times.

7 SPECIAL BREAKER-FAILURE SCHEME FORSINGLE-POLE TRIP SYSTEM APPLICATION

Special arrangements should be considered for abreaker-failure scheme in a single-pole trip application.

The problem is that the unequal load current duringsingle phasing may pick up the ground overcurrent Iounit (the phase overcurrent units may pick up also) inthe conventional breaker-failure scheme (Fig. 13-24)and trip the scheme falsely. One special arrangement,as shown in Figure 13-24, has been employed success-fully in the field for years for solving this problem. Inthis special arrangement, no ground overcurrent isused in the scheme. Each phase overcurrent unit issupervised by its individual breaker-failure initiatecontact 62Z/A, B, and C. Blocking diodes TRB areused for blocking the sneak path during single phasing.

Figure 13-24 Special consideration on breaker failure scheme for single-pole trip application.


14

System Stability and Out-of-Step Relaying

W. A. ELMORE

1 INTRODUCTION

Since relaying systems must function properly duringsystem swings, it is necessary to understand the effectsof these disturbances on relay performance. Swings arethe oscillations of synchronous machines with respectto other synchronous machines. They are caused bychanges in load, switching, and faults. A swing doesnot necessarily indicate system instability. In somecases, however, the swing is severe enough to causesynchronous machines to go out of step. Beforeexamining the influence of system swings on relayperformance, three factors must be considered: steady-state stability, transient stability, and relay quantitiesencountered during swings.

A system is stable when it is able to developrestorative forces in excess of the disruptive forces towhich it is subjected. Disruptive forces are produced bysuch factors as faults, loss of excitation, or switching oflines, transformers, or generation. Restorative forcesare produced by current flow, voltage increase, orimpedance reduction produced by fault removal and/or line switching.

2 STEADY-STATE STABILITY

The fundamentals of power transmission and stabilityare more easily understood if both system resistance,excluding the load impedances, and machine saliencyare neglected. In this case, the power P transmittedover circuits connecting two portions of the systems is

given by the following equation:

P ¼ VSVR

Xsinf ð14-1Þ

where

VS and VR¼ sending- and receiving-end voltages,respectively

X¼ reactance between VS and VR

f ¼ angle by which VS leads VR

See Figure 14-1.If system resistance is not neglected, different

equations apply for the sending- and receiving-endpower. The variables, however, are essentially the same.If phase-to-phase voltages are used, Eq. (14-1) yieldsthree-phase power. For this discussion, VS and VR aretaken as per unit quantities, and Eq. (14-1) gives perunit power. If VS, VR, and X are held constant in Eq.(14-1), the power flow is changed by varying the anglef.

As the load increases at the receiving end, synchro-nous machines are momentarily slowed down, and themachine-rotor inertia meets the increased load require-ments. That is, an increase in load results in a smallreduction of system frequency until there is a change inmechanical input via the governor or manual action.

Figure 14-1 Simple equivalent system.

339

To restore system frequency, the mechanical input tothe machines must be increased. This input must begreater than the steady-state load requirements, sincethe machines must be accelerated to a new and largerangle. When the new angle f is reached, the mechan-ical input will exceed the load requirements by theamount required to accelerate the machines. Themechanical input must then be reduced to maintainfrequency and required power transfer. Any loadchange therefore results in swings or oscillations asthe system adjusts to the changes. Steady-state stabilityis the ability of the system to adjust to gradual loadchanges.

The extreme unstable condition occurs when f isequal to 908. At this point, increased load conditionscould only be met by increasing VS or VR. An increasein f would cause a reduction, rather than an increase,in power transfer.

3 TRANSIENT STABILITY

Transient stability is the ability of the system toproperly adjust (remain in synchronism) to suddenlarge changes. Again, if we ignore system resistanceand machine saliency, the power transmitted duringthe transient interval P is given by the followingequation:

P ¼ V0SV

0R

X0 sinf ð14-2Þ

In Eq. (14–2), P is three-phase MW if V0S and V0

R areexpressed in kV (phase to phase) and X0 in ohms perphase, where

V0S ¼ voltage behind transient reactance at the

sending endV0

R ¼ voltage behind transient reactance at thereceiving end

X0 ¼ reactance between V0S and V0

R, includingtransient reactances of the machines

f ¼ angle by which V0S leads V0

R

Figure 14-2 shows a power-transfer curve. Note thatthe peak value is inversely proportional to the totalreactance X. Figure 14-3 describes how X is influencedby the presence of a fault, as well as the type of fault.

It is recognized that the quantity to be inserted asXF in the fault representation is dependent on the typeof fault, whether 3f, ffG, ff, or fG. For the 3ffault, XF is 0. For the ffG fault, XF is X2 and X0 inparallel. The positive sequence network is retained

intact without reduction, whereas X2 and X0 are thethevenin impedances in the negative and zero sequencenetworks, respectively. XF for the ff fault is simplyX2. For the fG fault, XF is X2 and X0 in series.

Based on the normal relationship found betweenX1, X2, and X0, the relative order of severity of thevarious types of faults produces the relative power-transfer curves shown in Figure 14-4. The reduction ofthe ‘‘T,’’ of which XF is one leg, to an equivalent ‘‘pi’’produces X0 which is the transfer impedance that, withthe voltage VS and VR, defines the peak value of thepower-transfer curve.

Figure 14-5 describes a representative case of twoparallel lines with generation at each end. The powertransfer curves show that a fault causes an immediatedrop in transmitted power from bus S to bus R becauseof the fact that the ‘‘torque angle’’ m cannot changeinstantaneously, but X0 does. This change in power

Figure 14-2 Power transfer curve.

Figure 14-3 Effect on X0 of fault.

340 Chapter 14

quantity DE. Power DE is the difference betweenmechanical input and electrical output that, therefore,produces acceleration of the rotating mass. With lossesignored in this example, quantity DE is also thedifference between electrical power input and mechanic

power output at the receiving end, causing decelerationthere. Both of these effects increase m.

When the fault occurs (Fig. 14-5), the transmittedpower is reduced to point E because of the increasedeffective X0, and the swing begins along E-F. At pointF, breaker A opens, and the transmitted powerincreases to G. The swing then continues along G-H.

When the fault is cleared at H by breaker Bopening, the sending-end rotor kinetic energy hasincreased by an amount proportional to area I, sincethe mechanical input has exceeded the transmittedelectrical power. As the fault is cleared, the transmittedpower increases to J, exceeding the mechanical inputand causing the sending-end machines to decelerateand the receiving-end machines to accelerate. Since thevelocity of a rotating mass cannot be changedinstantly, the swing continues to K, at which pointthe additional sending-end rotor energy, resulting fromthe fault, is completely absorbed (area II equals area I).

The velocity of the sending-end mass, with respectto the receiving-end mass, is 0 at K. At K, the electricaloutput of the sending end exceeds the mechanicalinput; therefore, the swing reverses, reaching a pointN. At N, the swing reverses again. Voltage regulatorand governor action, as well as system resistance, willdampen the oscillation, until the final operating pointis reached.

If the initial swing went to point L and the sending-end generators still had excess rotor energy (area IIsmaller than area I), the swing would continue in thesame direction. After L was passed, the mechanicalinput of the sending-end generators would againexceed the electrical output, and the swing would beaccelerated, resulting in instability with the machinesoperating out of synchronism with each other. Afterthis, only system separation and resynchronizing of themachines could restore normal system operation.

4 RELAY QUANTITIES DURING SWINGS

With the fault cleared and the system operating out ofstep, extreme variations in voltage and current willoccur throughout the system. Figure 14-6 providesinsight into this phenomenon. A simple system isrepresented with the relay of interest shown at someintermediate point between the two sources. The relayvoltage VR is established by the angle between VS andVU (each assumed here tomaintain their predisturbancevalues) and the distribution of impedances (ZS, ZL, andZU) between these two sources. VR will assume someposition between them. As VS, the accelerating source,

Figure 14-4 Relative order of severity of fault types.

Figure 14-5 Power transfer curves before, during, and after

line-to-line fault.

System Stability and Out-of-Step Relaying 341

moves in phase positionwith respect toVU, the apparentimpedance viewed by the relay (VR/IR) will change withtime, producing a trajectory on an R-X (resistance-reactance) diagram such as that of Figure 14-6c.

With the depressed voltage and large current, thisswing condition (long after the fault has been success-

fully cleared) appears to be a three-phase fault. Thelocation of this apparent fault is at the electrical centerof the system. If this apparent impedance enters theoperating area of a distance relay, it will operate.

If VR and VU are not supported at full value, theirratio is influential in determining the locus on the R-X

Figure 14-6 Relay quantities for OS condition. (Shown for 908 impedance angles.)

342 Chapter 14

diagram as the swing progresses. Also, if impedancesare not pure reactance values as is assumed inFigure 14-6, the effect can be determined using asimilar simple diagram.

Interaction between machines in complex networkscan only be determined by a large-scale digital study.For distance relaying evaluation, actual impedance(using line current and relay voltage) values must bedetermined with respect to time for each pertinentrelay location, and each chosen switching condition,fault location, and type.

5 EFFECT OF OUT-OF-STEP CONDITIONS

5.1 Distance Relays

A distance relay (21) responsive to three-phase faultswill operate if an out-of-step (OS) condition producesa swing locus that falls within its operating area (Fig.14-7).

When swing ohms enter the operating area of a zone1 relay with a circular characteristic, there is a 908angle between the voltages at the points along the lineangle identified by the relay reach. In Figure 14-7, forexample, there must be a 908 phase displacementbetween the voltage phasors at relay location A and apoint at 90% of the line length for a 90% reach zone 1phase relay to operate on a swing. The effectivegenerator voltages will be displaced substantially morethan 908 (angle m in Fig. 14-7). The likelihood of asystem attaining a stable operating condition after

such a swing is virtually nil. In general, zone 1 swing-trips occur only on unrecoverable swings.

Some form of blinding is required to screen over-reaching distance relays against tripping on severeswings from which recovery is possible. Operatingindependently, a phase-distance relay (21) will initiatetripping when the angle between the two systemvoltages is very large and increasing (Fig. 14-7).

Figure 14-8a shows that zone 1 tripping is highlydependent on the locus of the swing ohms andtherefore the distance to the electrical center for thecase involved.

Figure 14-7 Source angle relationship for swing trip on the


Figure 14-8 Effect of OS swings on various line relaying

systems.


5.2 Directional Comparison Systems

Swing ohms entering a protected line area will producesimultaneous high-speed tripping at the two terminals.Only swings entering external to the line area will blocktripping, and then only if local 21S operates and/or theremote 21P does not. For the permissive overreachingtransfer-trip or the unblock system, tripping occurs onOS conditions only if the 21P at both terminalsoperate. If tripping is to be avoided for cases whereoperation takes place, out-of-step blocking must beused.

5.3 Phase-Comparison or Pilot-Wire Systems

Phase-comparison or pilot-wire schemes are solelycurrent-responsive, and since swings produce athrough-current condition, tripping does not occur.

5.4 Underreaching Transfer-Trip Schemes

Out-of-step swings entering the circle of either zone 1relay will cause tripping at both terminals when usingthe underreaching transfer-trip (whether direct orpermissive) scheme.

5.5 Circuit Breakers

With the two system segments 1808 apart at the instantof interruption, a theoretical undamped recoveryvoltage of four times normal is possible. Figure 14-9describes this phenomenon with the breaker at theelectrical center of the system. At current 0, whereinterruption takes place, the voltage on each side of thebreaker must settle at a new value. In the process ofgetting there, overshoot takes place as a result of thepresence of inductance and distributed capacitance inthe system. Recovery voltage is the voltage across thebreaker contacts following current interruption.

Figure 14-10 shows that this identical phenomenonoccurs even though the breaker is as far away from theelectrical center as possible. The extremely largetransient recovery voltage still appears. If the circuitbreaker has insufficient dielectric strength to withstandthis voltage, reignitions will continue until a morefavorable angle is reached. To interrupt at all, abreaker must be capable of attempting interruption,possibly for several seconds, at each current 0. If thebreaker cannot perform such interruptions, tripping

must be initiated at a favorable angle, preferably justbefore the two sources are in phase.

5.6 Overcurrent Relays

Figure 14-6 can be used to illustrate the conditionsencountered by phase-overcurrent units during swings.Assume, for example, that an instantaneous over-current unit set for 2.5 times full load were used in aline connecting Vs and Vu, and that ZSþZLþZU

equals 0.765 per unit on the full-load base. During anOS condition, the instantaneous unit would operatebecause the current reaches at least 2.61 (2/0.765) timesfull load when VU lags by 1808. Swings during stableconditions will also result in higher than normalcurrents, although currents will be considerably lessthan during an OS condition.

5.7 Reclosing

When a fault persists after reclosing, the stability of thesystem will probably be jeopardized. On the otherhand, system stability is greatly improved if the fault is

Figure 14-9 Maximum recovery voltage on OS trip with

breaker at electrical center.

344 Chapter 14

temporary and does not reignite following reclosure.Three-phase faults tend to be permanent more oftenthan other faults. They also have the most severe effecton stability. For this reason, reclosing is often blockedfor three-phase faults and for OS conditions, butallowed for all others.

There appears to be no advantage to the high-speedreclosing of both terminals following an OS trip.Reclosing one terminal, or preferably blocking trip-ping at one terminal, will facilitate system restoration.The second terminal can be reclosed under synchron-ism-check relay supervision.

6 OUT-OF-STEP RELAYING

Ideally, fault relays should clear faults fast enough tomaintain stability. Also, they should not operate onswings from which the system can recover. If a systemdoes go out of step, it should be split by circuitbreakers opening at a few preselected locations, in sucha way that generation and load on each side of the splitare reasonably balanced. The system should not be

split so most of the generation is separated from themajor system load.

In Figure 14-11, breaker A is in a poor location forsplitting system (1) from (2), since it would dump oneunit of load on system (2), which only generates 0.4unit of power. Splitting the system using breakers D orE would offer a more tolerable generation/loadbalance. In this scheme, system (1) need only increaseits generation from 0.6 to 0.66 to maintain frequency.

Figure 14-11 also illustrates that OS tripping isdesirable at some points, but should be blocked atothers. This selective tripping/blocking philosophy isbasic to the intelligent application of OS relaying.

6.1 Generator Out-of-Step Relaying

Generator per unit reactances have steadily increasedover the years, and inertia constants have decreased asmachine ratings have increased. This, in turn, hasreduced critical clearing times and increased the needfor the OS relaying for generators.

Loss-of-field relays, equipped with directional unitsand undervoltage supervision, may provide a measureof OS protection for generators. Viewed from theterminals of a large modern machine, the ohms will, ingeneral, fall within the machine or unit transformerwhen the machine is out of step with the system. If theswing ohms fall within the machine for the systemshown in Figure 14-11, the KLF (or KLF-1) loss-of-field relay (40) will operate if swing ohms stay insidethe characteristic circle for 0.25 sec.

If a loss-of-field relay is used for OS sensing, thetimer must not time out for stable swings. It mustoperate, however, for field failure before damage (orfurther damage) can occur, and it must recognize thefastest realistic swing rate. Generally, all these timeconstraints can be satisfied. Figure 14-12 shows atypical stable swing locus following a severe three-phase fault.

If swing ohms pass through the unit transformer,OS detection may not be possible with either a loss-of-

Figure 14-10 Maximum recovery voltage on OS trip with

breaker at source.

Figure 14-11 Generation and load distribution.


field relay or simple distance relay. Moving thedirectional unit characteristic output to point G onFigure 14-12 would substantially increase the possibi-lity of a false trip on a stable swing, such as GCD.Alternatively, the time criterion could be increased tothe point where the stable swing would not triggerrelay operation, but then the fastest out-of-step swingmay not be recognized.

When the smaller characteristic is used as describedby the dotted circle of Figure 14-12, a higher degree ofsecurity is achieved. However, it occurs at the expenseof making out-of-step recognition impossible, as wellas losing the ability to match accurately, as describedin Chapter 8, the loss-of-field relay characteristic,steady-state stability, and minimum excitation unitlimits for the machine.

In general, devices applied for other functions (faultdetection, loss of field, thermal protection, etc.) areunsuitable for detecting OS conditions. The use ofrelaying, tailored explicitly for OS detection, is the onlydependable approach, unless extensive studies demon-strate that other devices will suffice or instability isimprobable.

This should not be taken to mean that generatortripping is encouraged when OS conditions develop,

but rather that OS detection may be easiest at thegenerating plant. Out-of-step separation would then beaccomplished by transfer-tripping, or other suitablemeans, to maintain a generation/load match, asdescribed above.

6.2 Transmission-Line Out-of-Step Relaying

The prime criterion in OS tripping is to maintain ageneration/load match in the islands created. If such amatch were perfect, no large load shifts and loaddropping would be required. Also, little or nogeneration would be dropped. To even approximatethis ideal would, in all probability, require trip-blocking at some locations and trip initiation at others.

Distance-relay operation on OS conditions tends tooccur at locations where the relay reach settings arelongest. There are two reasons for this phenomenon.First, the minimum system voltage during an OScondition tends to occur in the high-impedancesegments of the system. Second, distance relays withlong reach settings, such as those on long lines, cover alarger area on the R-X diagram and therefore are morelikely to respond to swing conditions. Out-of-steptripping at long-line terminals is not necessarilyconducive to ideal system splitting.

7 PHILOSOPHIES OF OUT-OF-STEPRELAYING

Certain fundamental objectives should influence thedesign of protection systems:

1. Block tripping at all locations for stable swings.2. Ensure separation for every OS condition.3. Effect separation at points that will leave a

satisfactory load/generation balance in eachseparated area. Loads should not be inter-rupted.

4. Block tripping or automatically reclose at oneend of any line that trips because of an OScondition.

5. Initiate tripping while the systems are less than1208 out of phase and the angle is closing inorder to minimize breaker stress.

6. Minimize the possibility of an OS conditionoccurring by

(a) Using high-speed relaying(b) Using a high-speed excitation system

Figure 14-12 Stable swing following clearing of nearby

three-phase fault with the KLF relays (40)

346 Chapter 14

(c) Employing loss-of-field relays to remove aunit that is drawing excessive reactivepower from the system

(d) Providing sufficient transmission capacity(e) Tripping generators upon the loss of

critical transmission lines(f) Applying generator braking resistors or

inserting series capacitors for critical faults(g) Applying fast valving techniques(h) Using an independent mechanism for each

breaker pole to downgrade faults fromthree-phase to phase-to-phase

Although easily stated, these objectives are not soreadily achieved, particularly item 3 above.

7.1 Utility Practice

Utility practice consists of a combination of:

1. Allowing the line-protection relays to initiateOS line tripping

2. Allowing the loss-of-field relays to initiate OSgenerator tripping (when the nature of the loss-of-field relay allows it)

3. Restricting relay-trip sensitivity at the higherpower factors

4. Blocking tripping on OS5. Blocking reclosing on OS6. Initiating tripping using relays designed for OS

tripping

There is no industry standard for protection-systemdesign; however, once the difficult functional decisionsof ‘‘what’’ and ‘‘where’’ are made, there is reasonableconsistency in the ‘‘how.’’

8 TYPES OF OUT-OF-STEP SCHEMES

Some typical systems used in OS relaying are describedhere.

8.1 Concentric Circle Scheme

A concentric circle scheme for OS detection onterminal A is shown in Figure 14-13. Customarily, anOS relay with a characteristic as shown for 68 is addedto a transmission-line relaying system and surroundsan over-reaching element, such as 21P. Because ofrotating apparatus inertia, significant time is requiredfor the torque angle to advance and the swing locus to

pass from 68 to 21P. For a fault within the 21P reach,however, both elements operate essentially simulta-neously. The relaying logic senses the sequence,identifies swing or fault, and initiates the appropriateaction.

The dotted arcs of circles described as a ‘‘tomato’’characteristic represent a popular analog implementa-tion of the ‘‘outer’’ characteristic.

This scheme is appropriate for an OS trip-blockingfunction or reclose-blocking. It is not appropriate forOS tripping, however, unless additional logic is added.If an external phase fault occurs close to the balancepoint of 21P, as at P in Figure 14-13, for example, therelay will respond slowly because of its low energylevel. Device 68, on the other hand, has an appreciablyhigher energy level and operates faster than does 21P.Removing remote terminal infeed on external faultclearing following breaker failure can also producesequential operation of 68 and 21P. If the time in thesensing logic is shorter than this operating-timedifference, a fault at P would be incorrectly identifiedas an OS condition and would cause a false trip at A. Ifthe time is increased to avoid this situation, rapid OSswings would not produce OS tripping. Adding a thirdconcentric circle would allow better perceptive segre-gation of swings and faults, but could introduce loadohm involvement. OS tripping can be achieved withthe two-concentric-circle scheme if the 68/21P operat-ing sequence is checked and the resetting sequence iscorrect.

For the three-terminal applications, infeed canadversely affect OS blocking relays that use theconcentric circle scheme if the sequential clearing of athree-phase fault can occur. The reach-shorteningeffect of the third terminal infeed can cause an internal

Figure 14-13 The concentric circle scheme for out-of-step

detection.


line-end fault to operate 68, but not 21P. Clearing theinfeed may then allow 21P to operate. This sequencecould cause undesirable OS trip-blocking at oneterminal. Out-of-step blocking then should not beused in a three-terminal application unless remoteterminal coverage can be obtained with maximuminfeed.

8.2 Blinder Scheme

One form of a blinder relay has an operatingcharacteristic that parallels the transmission-line plotin an R-X diagram (Fig. 14-14). A single-blinder relay21B, gives the two linear characteristics shown.

An obvious application of this device is for limitingthe coverage of a distance relay in the load area. This isa side benefit of the application of blinder OS relays.

A single-blinder relay plus auxiliary logic can beused for OS tripping. Its use, however, is limited toonly those applications where OS trip-blocking ofphase-distance relays is not required, since swingspassing through the line section (on an R-X plot) willcause operation of the line relays. A single blindercannot distinguish between a fault and an OScondition until the resetting sequence is confirmed.Such a scheme delays OS tripping until the swing iswell past the 1808 position and is returning to an in-phase condition.

The single-blinder OS package is recommended forcausing a system splitting to occur through the trippingof a line that is protected by a phase-comparison relay.

Another application is OS tripping on swings passingbeyond the reach of the line relays (on an R-X plot).Finally, this relay scheme is particularly well-suited forgenerator OS trip applications.

The two-blinder scheme (Fig. 14-15) senses OSconditions by observing the operating sequence of theouter and inner blinders. A fault produces essentiallysimultaneous operation; an OS condition causes theouter blinder to operate first, followed by operation ofthe inner blinder. The two-blinder scheme allows thetrip-area restriction of distance relays, OS trip-block-ing, OS reclose-blocking, or OS tripping, regardless ofnormal load-flow direction.

9 RELAYS FOR OUT-OF-STEP SYSTEMS

9.1 Electromechanical Types

9.1.1 KS-3 (68) OS Blocking Scheme

Figure 14-16 shows the configuration for the type KS-3out-of-step blocking scheme. If ZOS operates approxi-mately 60msec or more ahead of 21-2, the OS relayoperates to block all or selected tripping. The OS relayalso blocks reclosing when some elements in thesystem, such as zone 1 or time trips, are allowed tooperate during OS conditions.

For faults, 21-2 trip contacts or the D0 and I0contacts of the ground relay close to short out the OSunit coil, blocking pickup of the OS unit. This scheme

Figure 14-14 The single blinder scheme for out-of-step

detection.

Figure 14-15 The two blinder scheme for out-of-step

detection. (I¼ inner blinder, O¼ outer blinder.)

348 Chapter 14

is recommended for short-to moderate-length lines. Itshould not be used on long lines, where the load mightoperate the ZOS unit.

9.1.2 KST (68) OS Tripping Scheme

Figure 14-17 illustrates the KST scheme of OStripping. After sensing an OS condition in the sameway as the KS relay, telephone relays T1 and T2 addtwo requirements: The 21-2 relay (for example, theKD-10 phase-distance relay) must operate for 100msec, and 21-2 resets 60msec or more ahead of ZOS.On a swing, ZOS operates first to energize OS. If 21-2does not operate before 60msec, OS operates. Thenwhen 21-2 operates, the AR relay, T1, is energized. Ifboth ZOS and 21-2 remain closed for 100msec, T1operates. As the swing moves out, 21-2 resets first,deenergizes AR, and permits the energization of T2through AR back contacts if ZOS is still closed. If ZOS

does not reset for 60msec, T2 operates to trip andblock reclosure as shown.

A fault that operates ZOS and 21-2 together (orwithin approximately 60msec) will have no effectbecause the short around the OS coil will be

maintained. The zone 2 timer would time out andtrip in the scheme shown.

9.2 Solid-State Types

9.2.1 SDBU-1 (21B), SI-T (50), ARS (94), OSTripping Scheme

For the single-blinder OS tripping scheme (Fig. 14-18),swings from the right to left cause B1 to operate, B2 tooperate, B1 to reset, and then B2 to reset. It is of noconsequence whether B1 is initially operated by loadand B2 does not subsequently reset.

Device 50 (SI-T relay) is sensitively set and operatesat a current level above maximum zero power factorinterchange, line charging, or transformer-magnetizingcurrent. The device operates when a swing begins andprohibits load pickup trip.

Thus, AND 2 operates when B1 and 50 operate withB2 reset to identify the swing origin in the positive Fregion. After 4msec, the feedback circuit holds theupper input AND 2. AND 4 has an output when theswing moves between B1 and B2 to operate bothblinders. If AND 4 output persists for 20msec and theswing moves across B1 to reset it, AND 6 has anoutput. An output from AR occurs 20msec later fortripping and reclose block.

Swings originating to the left of B1 traveling left toright produce identical action through AND 1, AND3, and AND 5. The restricted trip feature preventstripping on recoverable swings. B1 and B2 may be usedto supervise the tripping of a phase-distance relay.

Figure 14-16 The type KS-3 out-of-step blocking scheme.

Figure 14-17 The KST out-of-step tripping scheme.


Out-of-step block of reclosing is not available with thiscomplement, unless OS trip is used.

The above scheme is recommended for generatorOS sensing, because its logic requires that swing ohmsemerge from the side of the relay characteristicopposite to that from which it entered. That is, theremust be a reversal of power flow as viewed from themachine terminals, and the reversal must occur duringhigh current. These two conditions will not be satisfiedunless the machine is out of step with the system. Alow-current reversal can, however, occur duringmotoring.

This scheme, or its equivalent, supervised by anover-current or distance relay, is the most secureavailable for generator OS tripping. A severe but stableswing, such as shown in Figure 14-12, cannot causemisoperation, regardless of the timing involved in thetransient.

9.2.2 Lens Scheme

One significant variation of the blinder scheme uses alens and single-blinder line characteristic as describedin Figure 14-19 for out-of-step tripping. It is alsoequipped with a reactance-type characteristic torestrict the reach of the system to the desired extent.Various areas are established (inside the lens, right ofthe blinder, left of the blinder, above the reactance line,below the reactance line) and, by the addition of

Figure 14-18 The SDBU-1 out-of-step tripping scheme.

Figure 14-19 Lens scheme for out-of-step tripping GZX-

104.

350 Chapter 14

sequence and timing logic, can determine the originand termination of swings and the time involved inpassage, thus identifying out-of-step conditions.

9.2.3 Double-Blinder OS Tripping and BlockingScheme

Figure 14-20 contains the logic for out-of-step sensingused in MDAR 2.5. 21BO and 21BI are the outer andinner blinder units, respectively. 3fF is an input from aphase selector identifying the fact that no unbalancedfault exists (not fG, not ff, not ffG). IAL signifiesthat current above a preselected level exists.

The logic differentiates between a swing and faultthrough the sequence of operation of 21BO and 21BI.If it is a swing, the optional blocking (shown asswitches OSB1, OSB2, and OSB3) of each of threedistance units is obtained. The inner blinder supervises(permits) tripping of each of the distance units to avoidany possibility of load trip not resulting from a swingcondition.

Logic is included to permit delayed OS tripping ‘‘onthe way out’’ to minimize the possibility of breakerdamage during tripping or ‘‘on the way in’’ tominimize the possibility of thermal damage to atransmission line. The terms ‘‘way in’’ and ‘‘way out’’refer to the trajectory of the ohmic value seen by the

relay and the point at which tripping is initiated (on theway into the inner-blinder operating region or on theway out of the operating region of the outer blinder).

10 SELECTION OF AN OUT-OF-STEP RELAYSYSTEM

The key ingredients in out-of-step relaying areidentification and strategy. Identification is possibleusing any of the schemes described, though a higherorder of security is mandatory for those schemes to beused for OS tripping. Also, OS identification isdependent on suitable voltage and current relation-ships being present at the point of application of therelaying system to clearly recognize an OS conditionwhen it occurs. Strategy relates to action required oncean OS condition has been identified. The choices areOSB (out-of-step block of tripping), OST (out-of-steptrip), and OSBR (out-of-step block of reclosing).

The electrical center is not a fixed point in a system.Indeed, several electrical centers may be present for agiven swing condition. Further, the electrical centermoves as the number of generators and switchingconditions vary. For the particular unstable case underconsideration, the ohmic value (and angle) manifestedat the relay location must be known with respect to

Figure 14-20 Out-of-step trip logic used in MDAR 2.5.


time to assure that proper recognition is possible. Ingeneral, a system study is required for this.

Out-of-step blocking is less critical than OS trippingin terms of identification of the swing and security. OSblocking must respond only when the blocked deviceresponds. Swings producing ohmic values beyond thereach of, say, a supervised zone 2 phase-distance relayneed not be recognized by the OS detection scheme.Also, a fault appearing on the R-X diagram betweenthe characteristics of an OS relay and a phase-distancerelay that is used with it can falsely identify an OScondition with impunity, whereas OS tripping underthe same circumstances could not be tolerated.

Stated simply, (1) OS blocking should be appliedwhen swing-produced trips are possible but are intol-erable, and (2) OS tripping should be applied whentripping will not naturally occur and tripping must takeplace for an optimum generation-load match following

separation of the two system segments that were out ofsynchronism. Even when ‘‘natural tripping’’ occurs (thedistance relays applied for other functions operate forthe OS condition), OS sensing is required to blockreclosing. Reclosing following a legitimate OS trip isfutile and should not be attempted.

Blinders aid natural tripping by providing a screen-ing effect against undesired tripping in response tolarge power swings that are stable. This is particularlyuseful in long-line applications that require largeimpedance settings on distance relays, making themmore susceptible to tripping on load.

Out-of-step relaying is as much art as science,requiring a realistic appraisal of what is possible, whatis probable, and what is certain. It may, in particularareas of a complex system, be less expensive touniversally apply out-of-step blocking rather than todeduce where it is needed.

352 Chapter 14

15

Voltage Stability

L. WANG

1 INTRODUCTION

Power system instability may occur under two kinds ofdisturbances: (1) gradual variations of system condi-tions such as load, and (2) drastic changes of systemconditions such as faults and loss of important lines orgenerators. According to the reasons causing powersystem instabilities, they can be classified as small-disturbance (SD) voltage instability, SD angle instabil-ity, large-disturbance (LD) voltage instability, and LDangle instability or transient instability. While Chapter14 covers transient instability, this chapter mainlyfocuses on voltage instability.

1.1 Small-Disturbance Instability

During normal steady-state conditions, power systemvoltage magnitudes, voltage angles, and frequency areconstant. The power system is operating at stableequilibrium points. A power system equilibrium pointis composed of a set of independent voltages andcurrents.

As power system operating conditions change, suchas the variations of load conditions or systemconfigurations, the system equilibria will changeaccordingly. Under small variations, the transitionfrom one equilibrium point to the other can beassumed to be instantaneous. At certain conditions,the number of system equilibria will change, which iscalled bifurcation. For example, both the maximumpower point in the power-angle curve and the nosepoint in the Q-V curve are where the number of

equilibria unite, and both are bifurcation points. At thebifurcation point, a system usually loses its stability.The resulting unstable condition is small-disturbanceinstability.

For a stable operating condition, it is helpful toknow how stable that condition is. This translates tothe amount of power a power system can furthersupport from a steady-state condition without losingstability. The study of SD instability is concerned withthe stable degree of a stable equilibrium point. A staticsystem model will be sufficient for the study of SDinstability.

Small-disturbance instability can be further classi-fied as SD angle instability and SD voltage instabilitybased on the reasons initiating the bifurcation orunification of system equilibria. If a bifurcation occursbecause of generator angles increasing to their limits,the instability is called SD angle instability. If abifurcation occurs because of bus voltages decreasingto their limits, the instability is called SD voltageinstability. SD angle and voltage instabilities areexplained in the following two examples.

1.1.1 SD Angle Instability

The mechanism of angle instability can be illustratedby considering a single-machine infinite-bus (SMIB)system (Fig. 15-1). Figure 15-2 shows the systempower-angle curve, which is the trajectory of thesystem equilibrium points with the mechanical powerinput being the variable. Superscripts ‘‘s’’ and ‘‘u’’represent state variables at the stable and unstableequilibria, respectively. For a specific mechanical

353

power input, there exist two equilibrium points. One isstable, and the other is unstable. In Figure 15-2, thecontinuous line in the power-angle curve representsstable equilibria, and the dashed line in the power-angle curve represents unstable equilibria.

As the mechanical power is increased gradually, thestable and unstable equilibria will move closer. Finally,the stable and unstable equilibria will join together andthe system reaches the bifurcation or maximumloading point. At this point, the system will lose thestable operation condition. This instability phenom-enon is the small-disturbance angle instability becausethe unstable condition results from a generator anglereaching its limit (908 for this specific situation).

1.1.2 SD Voltage Instability

Analogous to angle instability, the mechanism of SDvoltage instability can be illustrated by considering agenerator connected to PQ load through a double-circuit transmission line, as shown in Figure 15-3.Figure 15-4 shows the system Q-V curve. The Q-Vcurve can be obtained by increasing the load whilekeeping the source voltage constant. The Q-V curverepresents the trajectory of stable and unstableequilibrium points. The nose point in the Q-V curverepresents the maximum power transfer, which occursat a load impedance corresponding to the Theveninimpedance as viewed from the load bus.

For any load conditions, there exist stable andunstable equilibrium points. The solid line in the Q-Vcurve represents the trajectory of stable equilibriumpoints, and the dashed line in the Q-V curve representsthe trajectory of unstable equilibrium points. A powersystem can never operate at unstable equilibriumpoints, thus unstable equilibrium points only haveanalytical importance and do not have practicalmeaning. With the increase of load, the stable andunstable points will become closer and eventually willjoin together at the nose point of the Q-V curve. Thenose point corresponds to the bifurcation point.

Unlike the situation discussed in the SMIB system,the occurrence of the bifurcation here is caused by the

Figure 15-1 Single-machine infinite-bus system.

Figure 15-2 Power-angle curve of a SMIB system.

Figure 15-3 Two-bus power system.

Figure 15-4 Q-V curve of a two-bus system.

354 Chapter 15

voltage decreasing, and thus the system’s inability tosupport further load increases. Therefore, this instabil-ity is SD voltage instability. At the bifurcation point,the voltage phase angle may be much less than its 908limit.

1.2 Large-Disturbance Instability

Large disturbances move a system condition far awayfrom a stable equilibrium. After the disturbances, if asystem can recover to its stable equilibrium, it is stable;otherwise instability will occur. These kinds ofinstability are called LD instability.

The study of LD instability is concerned with thesystem ability to return to an equilibrium followinglarge disturbances. Based on the dynamic systemtheory, for any stable equilibrium, there exists a regionaround it; if the system’s initial transient states fall inthis region, the system will recover to the stableequilibrium. If the initial states are outside of thisregion, the system will be unstable. This region isdefined as the domain of attraction of the stableequilibrium.

Finding the domain of attraction is a problem onlypertinent to nonlinear systems because linear systems,if they operate at steady states, will remain stable afterlarge disturbances irrespective of the strength of thedisturbances. In other words, the domains of attractionof linear systems are the whole state space.

Figure 15-5 shows the concept of linear systemstability and Figure 15-6 shows nonlinear systemstability. In Figure 15-5a, the ball is at its stableequilibrium; in Figure 15-5b, the ball is away from itsstable equilibrium but will return to it eventuallyirrespective of the initial positions of the ball. For

nonlinear system, the situations are quite different, asshown in Figure 15-6. Depending on the disturbancesor its initial positions, the ball may not return to itspredisturbance stable equilibrium and may neverreturn to any stable equilibrium. Only when the ballfalls within the domain of attraction of a stableequilibrium will it become stable and return to thatequilibrium eventually.

Under large disturbances such as short circuits orloss of important lines or generators, a power systemcannot be described by a linear model. This is mainlybecause the real and reactive power terms include sinefunctions and multiplications of voltages. The domainof attraction of its equilibrium point is not the wholestate space but equilibrium dependent. To determinethe domain of attraction, a dynamic system model willbe required.

Similar to SD instability, LD instability may beclassified as LD angle instability (or transient instabil-ity) and LD Voltage instability. The concepts oftransient instability and LD voltage instability will beexplained using the example systems shown in Figures15-1 and 15-3, respectively.

1.2.1 LD Angle Instability (or TransientInstability)

Figure 15-7 shows the transient responses of the SMIBsystem following two large disturbances, which set theinitial states of the postfault system at ðd;oÞ ¼ ðd1; 0Þand ðd;oÞ ¼ ðd2; 0Þ. Angle d represents the angledifference between the two generators, and angularvelocity o represents the angular speed differencebetween the two generators. For simplicity of illustra-tions, we assume the angular velocity differences arezero at the two initial states. For the initial state ðd1; 0Þ,

Figure 15-5 Linear system stability: one stable equilibrium.

Voltage Stability 355

the postdisturbance system is stable and finally setdown at ðds; 0Þ, while for the initial state ðd2; 0Þ, thepostdisturbance system is unstable. The state ðd1; 0Þ iswithin the domain of attraction of ðds; 0Þ, but ðd2; 0Þ isnot. The resulting instability is the LD angle instabilityor transient instability.

1.2.2 LD Voltage Instability

Shown in Figure 15-8 are the transient responses of thetwo-bus system illustrated in Figure 15-3 following twolarge disturbances. The disturbances set the initialstates of the postfault system at ðV; _VVÞ ¼ ðV1; 0Þ andðV; _VVÞ ¼ ðV2; 0Þ, where V and _VV represent the load busvoltage and the rate of change of the bus voltage,respectively. For simplicity of illustrations, we assumethe voltage change rates are zero at the two initialstates. The postdisturbance initial state ðV; _VVÞ ¼ðV1; 0Þ is within the domain of attraction of ðVs; 0Þ,

but the state ðV2; 0Þ is not within this region. Theunstable condition originating from the initial stateðV2; 0Þ is the LD voltage instability because thedomain of attraction is affected by the bifurcationpoint, which is mainly caused by a decrease in voltage.

Three mechanisms play a vital role in voltagecollapses; tap changer dynamics, load dynamics(particularly those involving large percentages ofinduction motors), and generator field excitationdynamics. These mechanisms are critical in determin-ing whether a system becomes unstable following afault condition.

1.3 Voltage Instability Incidents

Many voltage instability incidents have been reportedover the past years. Following are some examples:

Figure 15-6 Nonlinear system stability: multiple stable equilibria.

356 Chapter 15

French system: December 19, 1978; January 12,1987

Swedish system: December 27, 1983

Japanese system: July 23, 1987New York system: September 22, 1970; July 13,

1977Florida system: December 28, 1982; May 17, 1985TVA system: August 2, 1987

Among these incidents, some were caused by gradualchange of loads, while others were caused by loss ofimportant lines or generators.

2 VOLTAGE INSTABILITY INDICES

Voltage stability studies have attracted great attentionin recent years, and various indices have beendeveloped for SD voltage instability detection. Someindices require the knowledge of system state variables(bus voltage magnitudes and phase angles) only at acurrent operating condition, and others require statevariables at stressed operating conditions. The formerneeds less computation time, but the latter can providepower margins to voltage collapse, which is very usefulto system operators. This section discusses typicalmethods in each category.

2.1 Indices Based on Current OperatingCondition

Corresponding to a combination of load demands andgenerator outputs, stable and unstable equilibria can

Figure 15-7 Transient responses of a SMIB system.

Figure 15-8 Transient responses of a two-bus system.


be found by solving load flow equations. A stableequilibrium is the condition at which the power systemoperates, while an unstable equilibrium is just amathematical solution of power flow equations. Apower system cannot operate at an unstable equili-brium, but the unstable equilibrium provides a helpfulconcept in defining voltage stability indices.

Voltage instability indices based solely on stableequilibrium and based on both stable and unstableequilibria have been proposed. The indices requiringthe knowledge of unstable equilibrium conditions takemore computational time because the procedure offinding the correct unstable equilibrium is morecomplicated than that of finding a stable one. In thissection, we limit our discussions to the method whichonly requires state variables (voltage magnitudes andphase angles) at a stable equilibrium.

2.1.1 Q Angle Method

Simply speaking, voltage instability (at least for a lotof cases) is due to too much power being transmittedto the grid. In increasing power to the grid, a pointwill be reached beyond which the grid cannot sustain.By looking at the power flow solutions mathematicallyat this point, the load flow Jacobian matrix becomessingular. The load flow Jacobian matrix reflects thesensitivity of power flow to voltage variations.Whenever the load flow Jacobian matrix is at or closeto singular condition, even a very small increase inpower demand or power supply will cause the powersystem to move away from its stable operationconditions.

Detecting the singularity of a load flow Jacobianmatrix is a widely used method for detecting voltagecollapse. The proximity of a Jacobian matrix tosingularity indicates how close an operating conditionis to the voltage collapse point.

To explain the Q angle method, let’s start from asimple two-bus system as shown in Figure 15-9. Thesystem may represent a Thevenin equivalent of asystem as seen from a load. In this system, a load withreal and reactive power, P and Q, is supplied by aninfinite system via a single lossless transmission line.The series reactance and shunt admittance of thetransmission line are denoted by jX and jB1, respec-tively. The source voltage V1 is kept constant at 1:0ff0,and the load bus voltage is V2ffd.

This system represents a simple load flow problem:given power demands, find voltages. To solve for thevoltage magnitude V2 and voltage angle d at the load

bus, the load flow equations are expressed by

� V1V2

Xsin d ¼ P ð15-1aÞ

V1V2

Xcos d� 1

X� B1

� �V2

2 ¼ Q ð15-1bÞ

In the equations load active power P and reactivepower Q are knowns; only V2 and d are unknowns.Solving Eq. (15.1) will provide bus voltage. This istypically done by using the load flow Jacobian matrix,which represents the sensitivity of power flow tovoltage variations. By taking partial derivatives ofactive power P and reactive power Q with respect tovoltage magnitude V2 and phase angle d, the load flowJacobian matrix for this simple system is found as

J ¼qPqV2

qPqd

qQqV2

qQqd

" #

On the other hand, with V2 and d as variables andwith active power P and reactive power Q as constants,Eq. (15.1a) or (15.1b) can be viewed as representing aplane curve. The intersection points of the two planecurves gives the bus voltage magnitude V2 and voltagephase angle d.

Figure 15-10 shows the real and reactive powercurves, where the active load P is fixed (1.0 p.u.) and thereactive power Q takes three different values. Since P isfixed, the load flow solutions vary along the constant Pcurve as Q is changed. When the reactive powerdemand is increased, the load bus voltage magnitudedecreases and its angle (absolute value) increases. It isalso found that when the reactive load becomes heavy,the angle between the gradient vector HQ and thetangent vector of the constant P curve at the feasibleload flow solution point (higher voltage magnitude) willincrease. For ease of discussion, we define a1 as theangle between the gradient vector HQ and the tangent

Figure 15-9 One-line diagram of a two-bus test system.

358 Chapter 15

vector of the constant P curve at the feasible load flowsolution. When Q¼ 1.08 p.u., the two curves aretangent to each other and the angle a1 equals 908.

The tangent point is the bifurcation point ormaximum load point, where two solution points arecoincident. Beyond his point, voltage collapse occurs.Thus, the angle a1, which is the Q angle, can indicatehow close an operation condition is to voltage collapse.

Figure 15-11 shows the curves for the case where thereactive load Q is fixed (1.0 p.u.) and the active power Ptakes different values. The load flow solutions are alongthe constant Q curve. Similar observations as seen inFigure 15-10 can also be obtained from Figure 15-11.

2.1.2 Application to Multibus Systems

The method described above can be extended togeneral multibus power systems. Consider a powersystem, and let n be the total number of buses minusthe swing bus. Allow the ðnþ 1Þth bus to be thereference bus, and assume mþ 1 to be the number ofgenerator buses. At the system buses, the real and

reactive power balance equations for the power systemmay be expressed by

Xnþ1

j¼1

ViVjYij cosðdi � dj � yijÞ � piðViÞ ¼ Pi

ði ¼ 1; . . . ; nÞ ð15-2aÞXnþ1

j¼1

ViVjYij sinðdi � dj � yijÞ � qiðViÞ ¼ Qi

ði ¼ 1; . . . ; n�mÞ ð15-2bÞ

where

ðPi;QiÞ ¼ constant part of the net powerentering bus i

ðpiðViÞ; qiðViÞÞ ¼ voltage dependent part of the netpower entering bus i

Viffdi ¼ ith bus complex voltageYijffyij ¼ (i, j)th element of the network

admittance matrix

Figure 15-10 Real and reactive power curves in the state space for the two-bus system (P is fixed).


Based on the assumptions on the number of busses,load flowEq. (15.2) includes 2n�m algebraic equationswith 2n�m unknowns. From a geometric point ofview, every equation in (15.2) represents a space surface,and the remaining ð2n�mÞ � 1 simultaneous equationsrepresent a space curve in a ð2n�mÞ-dimensionalspace. The intersection points between the space surfaceand the space curve give the voltage magnitudes andphase angles corresponding to a specified load level.

Figure 15-12 illustrates the relative movementbetween the space surface and the space curve as theinjection power in the space surface equation isincreased from Q to Q0. Also shown in the figure arethe gradient vector of the space surface and the tangentvector of the space curve calculated at their intersec-tion point.

The angles corresponding to the gradient vector ofthe space surfaces expressed by the active powerbalance equations are defined as P angles; and in asimilar way, the angles corresponding to the gradientvector of the space surfaces expressed by the reactivepower balance equations are defined as Q angles.

It can be seen that at the voltage collapse point bothQ angles and P angles are equal to 908. Consequently,the closeness of these angles to 908 is an indicator forthe voltage instability detection.

2.2 Indices Based on Stressed SystemConditions

The indices based on a current operating conditionusually take less time to compute and thus are suitablefor online application. But, a MVA margin of anoperating condition to voltage collapse is sometimesrequired. To compute the margin, a load and genera-tion variation pattern need be assumed. Along theassumed direction, the system is stressed until thevoltage collapse point. In a stressed condition, espe-cially near the voltage collapse point, the conventionaliterative power flow may have difficulty in converging;thus the continuation power flow method has to beused. The solutions from the continuation method canbe displayed in the form of P-V or Q-V curves.

Figure 15-11 Real and reactive power curves in the state space for the two-bus system (Q is fixed).

360 Chapter 15

P-V curves and Q-V curves are widely used byelectric utilities to determine voltage instability. Onboth P-V and Q-V curves, the nose points correspondto the voltage collapse point. These curves can begenerated by increasing the load demands at a singlebus or at the buses of a specific area according to a

predetermined load variation pattern. The continua-tion load flow method may be utilized to find thelow-voltage solutions required in plotting thesecurves.

Once the P-V and Q-V curves are drawn, systemvoltage stability can be evaluated based on the distance

Figure 15-12 Relative movement between the space surface and the space curve, Q<Q0.


between a current operating point and the noise pointsof these curves. The drawback of the method is theheavy computations required to execute the largenumber of load flows needed to create the curves.

2.3 Summary

This section reviewed themethods for small-disturbancevoltage instability detection. The methods based on acurrent operating condition are computationally fast,while the methods based on stressed system conditionsrequire more computational time. Although the lattercan calculate a MWmargin to voltage collapse point, afuture load variation is required in the calculation.

3 VOLTAGE INSTABILITY PROTECTION

While the voltage instability indices described in thelast section are academically sound and useful forsystem planning and operations, they have to besimplified for protective relaying applications becausethey require extensive computations and system-wideinformation. These indices provide excellent guidancein designing more practical voltage instability protec-tion tools. This section focuses on industry practices inprotection of voltage instability problems.

The critical issue in voltage instability protection isto identify and predict the occurrence of voltageinstability problems. In most cases, this is difficult todo without access to system-wide voltages throughcommunications. Compromises have to be made whendetecting voltage instability using localized voltages.

3.1 Reactive Power Control

Voltage instability is mainly caused by system inabilityto provide sufficient reactive power. Control of thereactive power sources is one of the effective methodsin preventing voltage collapses.

Capacitors can deliver reactive power to the powersystem and keep system voltage within tolerable limits.To ensure that capacitor bank control actions areactivated only during voltage collapse conditions, notduring fault conditions, capacitor banks switching issometimes supervised by zero sequence overvoltagerelays or overcurrent relays. This purpose can also beachieved by delaying switching actions, provided thatfaults can be cleared within a short period of time, say,0.5 sec. These schemes have been applied by a utility tosubtransmission systems at 161 kV and distributionsystems at 46 kV.

For transmission systems over 230 kV, fast capa-citor switching is required due to the potential impactof prolonged low-voltage conditions. Similar to sub-transmission systems, capacitor switching is activatedonly after faults have been cleared. Programmable-logic-controller (PLC)-based high-speed capacitor con-trol schemes have been used for fast capacitor switch-ing. Low voltages at all three phases are the triggerconditions, and a time delay of 0.2 sec is used to avoidswitching during fault conditions.

Other than mechanically switched capacitors, thereare other reactive power sources which can preventvoltage collapses. The reactive sources includestatic var compensators (SVCs), static condensers(STATCONs), synchronous condensers, and genera-tors. They are usually controlled by solid-state switchesso they provide varied reactive power in a continuousmanner. When the reactive power sources reach theirmaximum limits and voltages at main buses arestill in a dangerous state, other remedial means suchas load shedding will have to be used.

3.2 Load Tap Changer Blocking Schemes

Load tap changers (LTCs) provide an economic meansto help keep voltages at satisfactory levels. Duringnormal conditions (other than voltage collapses),LTCs will adjust the tap to a lower level wheneverthe controlled voltage is higher than a desired upperlimit; similarly LTCs will adjust the tap to a higherlevel whenever the controlled voltage is lower than thedesired lower limit.

Under low-voltage conditions that are a result ofpossible voltage collapses, the normal LTCs operationscannot help but deteriorate the situations. This is trueespecially when load reactive power demand is verysensitive to voltage levels. Under such a situation, thevoltage rise by LTCs will cause a significant increase inreactive power demand and thus widen the gapbetween reactive power supply and demand. There-fore, it is recommended that LTCs be blocked fromtrying to raise voltages whenever there are indicationsthat there is an impending system voltage collapse.

3.3 Load Shedding

Load shedding causes forced power interruptions tocustomers and results in high costs to power suppliersand consumers, therefore undervoltage load sheddingis usually the last option in preventing voltagecollapses. It is only applied when all other optionshave been taken but still without success.

362 Chapter 15

3.3.1 General Considerations

Undervoltage load shedding is implemented in stagesand a time-delayed fashion. In general, following arethe factors that need to be considered when imple-menting undervoltage load shedding schemes:

Determination of the amount of load to be shedDetermination of time steps of load sheddingSelection of load to be shedDetermination of voltage levels to shed loads

The selection of appropriate time steps and theamount of load to be shed at each step is important ineffective undervoltage load shedding. The proper timestep depends on system load characteristics, generationsource, and rate of change of voltage. The amount ofload being shed at each stage and the time delayassociated with each stage are usually predeterminedbased on extensive voltage stability studies. It is notallowed to shed more load than is necessary because ofunnecessary electricity interruptions. On the otherhand, insufficient load shedding does not preventvoltage collapses or alleviate the system conditions.Shedding load in stages with proper time delay cansolve this dilemma. Some utilities in the United Stateshave elected to shed 15% of peak load in three stepsduring an undervoltage condition:

1. 5% of area load at voltage 10% below the lowestnormal voltage with 3.5 sec time delay

2. 5% of area load at voltage 8% below the lowestnormal voltage with 5 sec time delay

3. 5% of area load at voltage 8% below the lowestnormal voltage with 8 sec time delay

All kinds of load, including industrial, commercial, andresidential, play an important role during voltagecollapse conditions. They are normally close to their

peak values at voltage collapse, so shedding any of thoseloads would be helpful to alleviate voltage conditions.

Constant power loads such as motors are the mostharmful to voltage conditions during voltage collapse,these loads may be considered first before sheddingother types of loads. Also, power factors should beconsidered when determining what loads should beshed because it is the reactive power shortage thatcauses voltage drops and voltage collapse. Thus,shedding loads with lower power factor will be moreeffective in maintaining system voltages.

When considering tripping lines, choose the oneswhich are less critical to maintain system integrity.Radial lines or lines with heavy tapped loads but lightthrough loads are better choices than other lines wouldbe. Regarding the voltage level for initiating loadshedding, it is a common practice to set the level in therange of 0.85 to 0.97 p.u. of the lowest normal voltage.

3.3.2 A Centralized Load Shedding Scheme

Undervoltage load shedding can be implemented eitherbased on local measurements or on measurements fromseveral substations. This section shows a centralizedundervoltage load shedding scheme. In the scheme, theload shedding decision is made based on voltagemeasurements and reactive power measurements fromthree substations, two at 230 kV and one at 500 kV.

Figure 15-13 shows the implementation, wherestaged load shedding is initiated when voltages inmore than one key bus are lower than settings and theoutput of reactive power sources is near theirmaximum Mvar capacity. Three stages are shown inthe figure with three time delay constants and threeload blocks. In the example, voltage setting is equal to97% of the lowest normal voltage; each load sheddingblock is equal to 250MW; and the first time delay T1 is10 sec, T2 and T3 are 2 sec.

Figure 15-13 Undervoltage load-shedding. (From IEEE Report, reference 2.)


16

Reclosing and Synchronizing

Revised by: S. WARD

1 INTRODUCTION

The large majority of overhead line faults are transientand can be cleared bymomentarily deenergizing the line.In fact, utility reports show that less than 10% of allfaults are permanent. It is, therefore, feasible to improveservice continuity byautomatically reclosing the breakerafter fault relay operation. For example, automaticreclosing greatly improves service in radial distributioncircuits, where service continuity is directly affected bycircuit interruption. High-speed reclosing on tie lines, ifsuccessful, also assists in maintaining stability. Reclos-ing is generally not used for cables as cable faults areseldom transient. It may, however, be applied on circuitswith both overhead line and cable sections, especially ifthe overhead line/cable ratio is high.

This chapter will describe the application andoperation of reclosing and synchronizing relays.

Microprocessor relays such as feeder protectionsDPU2000R and REF, distance protections REL, andbreaker terminal REB551 all offer optional built-inreclosing and synchronism check devices. These protec-tions implement many of the features and functionsdiscussed below for the individual relay units, and theapplication of reclosing and synchronism check followthe same guide lines as required for a stand-alone unit.

2 RECLOSING PRECAUTIONS

Automatic reclosing can improve continuity of serviceand increase the availability of transmission lines, butcertain precautions must be followed:

1. A generator must never be connected to asystem under conditions that will produce a thermal ormechanical impact that will deprive it of life. The angleof voltages across a breaker in the vicinity of agenerator is an inadequate measure of the possiblehazard associated with closing that breaker. Thesudden change in power in the generator followingclosure is, however, a key indicator. Steady-stateswitching may be appreciably less severe than reclosingfollowing a fault because of the transient forces thatare established by the original fault condition. Achange in power, following steady-state switching, of0.5 per unit will produce negligible loss of life.

The change in power following reclosure maysubstantially exceed this level. For this reason, reclos-ing breakers near a generating plant should beeliminated, restricted in number, be time-delayedbeyond roughly 10 sec, or be carefully controlled.Modern reclosing devices offer fault-selective reclos-ing, so that reclosing may be allowed for the less severesingle-phase-to-ground faults, while reclosing isblocked for the more infrequent two- and three-phasefaults.

2. When a transformer is subjected to a substantial‘‘through-fault,’’ severe forces are developed withinand between windings, which produces motion.Repetitive motion can produce failure of the winding.Cases have been reported of prolonged life oftransformers that are subjected to shorter and/or fewersevere faults. Minimizing the number of reclosures canprovide benefit in this regard. The protection systemshould also be designed so that a transformer fault, ona line-transformer section, does not result in reclosing.

365

3. Bus faults are not common, but when they dooccur they are generally severe and produce high faultcurrent. For this condition, it is not desirable to havethe breakers reclose automatically and a trip from busprotection relays therefore leads to lock-out.

3 RECLOSING SYSTEM CONSIDERATIONS

3.1 One-Shot vs. Multiple-Shot Reclosing Relays

The desired attributes of a reclosing system vary widelywith user requirements. In an area with a high level oflightning incidence, most transmission line breakerswill be successfully reclosed on the first try. Here, thesmall additional percentage of successful reclosuresafforded by multiple operations does not warrant theadditional breaker operations. Single-shot reclosingrelays (those which produce only one reclosure untilreset) are entirely justified.

Subtransmission circuit-reclosing practices also varywidely, depending on the requirements of the loadssupplied. If there are motors or generators in thesystem, the first reclosure may be time-delayed. Mostoften, two or three reclosures are used for subtrans-mission circuits operating radially, and only one or tworeclosures for tie circuits. Approximately half theutilities use some form of circuit checking beforetime-delayed reclosure to assure that either synchron-ism exists or one circuit is dead.

Multiple-shot reclosing relays are warranted ondistribution circuits with significant tree exposure,where an unsuccessful reclosure would generallymean a customer outage. A typical utility experienceon distribution feeders in an area with a large numberof annual thunderstorm days is as follows:

Number of successful reclosures %

Immediate 83.25

Second (15 to 45 sec) 10.05

Third (120 sec) 1.42

Total successful 94.72

Lockouts 5.28

The data show that increasing the number ofreclosures does improve service continuity, but theincremental benefit of each additional reclosure is lessthan for the preceding one.

3.2 Selective Reclosing

The speed of tripping is a significant factor in the successof a reclosure on a transmission circuit. The faster theclearing, the less fault damage and/or degree of arcionization, the less the shock to the system on reclosure,and the greater the likelihood of reenergization withoutsubsequent tripping. The probability of successfulreclosing then is improved if reclosing occurs only aftera high-speed pilot trip. By allowing only pilot trippingto initiate high-speed reclosing, maximum success canbe assured for single-shot reclosing, and many unsuc-cessful reclosures can be avoided. Such a systemeliminates the high probability of unsuccessful reclosureon nonpilot trips, particularly for end-zone faults inwhich clearing occurs sequentially and the deenergizedtime is short. Some modern reclosing devices (e.g.,REB551) include a feature to prolong the dead timewhen the communication link is out of service, thusincreasing the likelihood of a successful reclosure.

3.3 Deionizing Times for Three-Pole Reclosing

When reclosing at high speed, the dead time requiredto deionize the fault arc must be considered. Based ona study of 40 years of operating experience, minimumdead times can reasonably be represented by a straightline, using the following equation:

t ¼ 10:5þ kV

34:5cycles ð16-1Þ

where kV is rated line-to-line voltage. On a 345-kVsystem, for example, this formula would give anapproximate required dead time of 20.5 cycles.

If the inherent minimum reclosing time of thebreaker involved can produce a shorter dead time thanindicated, a reclosing delay must be incorporated. Inthe interests of placing the device in the protectedenvironment of the control house and having a moreaccurate timing device, the time delay should be in thereclosing relay rather than the breaker.

Single-pole tripping and reclosing requires longerdead time because of the fact that the two phases thatremain energized tend to keep the arc conductinglonger.

3.4 Synchronism Check

A synchronism-check relay is an element in thereclosing system that senses that the voltages on the

366 Chapter 16

two sides of a breaker are in exact synchronism. (Anautomatic synchronizer, on the other hand, initiatesclosure at an optimum point when the two systemsegments are not in precise synchronism, but have asmall beat frequency across the breaker contacts.) Thesetting for most synchronism-check relays is based onthe angular difference between the two voltages anddesigned to minimize the shock to the system when thebreaker closes. The angular difference between thevoltages does not, however, determine the transient towhich the system will be subjected upon closure.Rather, the shock to the system is related to the voltageacross the breaker contacts (the ‘‘phasing voltage’’).The phasing voltage is the critical quantity indetermining whether or not a breaker is allowed toclose. Therefore, more advanced synchronism-checkdevices have been developed that also have settings forvoltage difference, phase angle difference, and fre-quency difference.

3.5 Live-Line/Dead-Bus, Live-Bus/Dead-LineControl

Live-line/dead-bus, live-bus/dead-line (LLDB/LBDL)control is frequently introduced in a reclosing systemfor a transmission or subtransmission circuit. Thisscheme allows the breaker to be closed when the circuiton one side is energized at full voltage and the circuiton the other side is dead. The synchronism-check andLLDB/LBDL controls are complementary. The syn-chronism-check unit allows closure when the twovoltages are high but in synchronism; the voltagecontrol allows closure when one voltage is normal andthe other is very low, preferably 0.

Generally, synchronism check is performed on asingle-phase basis, i.e., by comparing one phasevoltage on each side of the breaker. For LLDB/LBDL control, however, it might be advantageous tocheck all three phases. This prevents out-of-synchron-ism closing of the breaker in case of fuse failure. Ifcheck is made in one phase only and the fuse wereblown in this phase, the line (or bus) would be seen as‘‘dead’’ and allow LLDB/LBDL closing.

3.6 Instantaneous-Trip Lockout

On distribution systems, where coordination with fusesis necessary, the fuse is often protected on the first tripwith an instantaneous tripping element on the substa-tion breaker relay followed by a reclose attempt. Thiscontrol is removed after the first trip, allowing the fuse

to blow and preventing a second breaker trip if thefault occurs beyond the fuse. The result is a combina-tion of minimum outage area for permanent lateralfaults and reclosing for temporary faults.

It should be recognized that voltage-sensitive loadson unfaulted lateral circuits may be more seriouslyaffected by a fault when this ‘‘fuse-saving’’ strategy isused. A fault on a fused lateral produces a voltage dipthroughout the circuit, and the severity of the dip isdependent on the proximity to, and nature of, thefault. Since complete loss of voltage to the load on theunfaulted laterals occurs when the substation breakeris opened, some deterioration in service occurs usingthis principle, in exchange for the benefit of not havingto replace a fuse for a temporary fault.

3.7 Intermediate Lockout

Unattended substations that are not equipped withsupervisory control can be controlled more effectivelyby a reclosing scheme that locks out on a permanentfault before exhausting all its reclosing shots. With anattended (or supervisory controlled) substation at theother terminal of the line, a manual reclosure can beattempted after lockout, whenever the operator judgesthat the fault no longer exists. If manual closure issuccessful, a synchronism-check relay will operate—inconjunction with the reclosing relay in the intermediatelockout condition at the unattended station—torestore the second-line terminal to service. Thisreclosing scheme is very effective for multiterminallines in which several unattended stations withoutsupervisory control are disconnected for line faults.

Intermediate lockout is achieved in REL301/512 bya ‘‘hold cycle’’ input; in MRC2 by the ‘‘pause’’ input,and in DPU2000R by the ‘‘ARCI’’ input. These relaysall perform the same function as described above;when the input is asserted, operation ‘‘freezes’’ andwhen deasserted, the reclosing cycle resumes.

3.8 Compatibility with Supervisory Control

All reclosing systems should incorporate some provi-sion allowing circuit breakers to be tripped manuallyor by supervisory control without inadvertent reclosingaction. Such a provision is inherent in any reclosingsystem that has an ‘‘initiate’’ function. In otherreclosing systems, lockout must be accomplished byother means, such as a breaker control-switch slip-contact or the temporary removal of control voltage tothe reclosing relay.

Reclosing and Synchronizing 367

3.8.1 Reset Time

Reclosing relays include a reset timer that starts whenthe final reclosing attempt has been made. If thebreaker opens before this timer has expired, the relaygoes to lockout and further reclosing is blocked until amanual close has been made.

3.8.2 Follow-Breaker Function

The follow-breaker function is used to preventsustained pumping action of a circuit breaker as aresult of a permanent fault, or if the breaker is closedby any other device than the designed reclosing relayaction. If during the open interval time the relay seesthe breaker close, then the relay will step forward in itsprogram and begin the reset timer. Should the breakeropen before the reset time expires, the timing for thenext open interval in the reclosing sequence will begin.

3.8.3 Close Fail Time

Generally, the reclosing relay monitors the breaker 52bcontact and will keep its close pulse asserted until thebreaker has closed. However, should the breaker fail toclose, a ‘‘close-fail-timer’’ will deenergize the closecontact when the set time has expired.

3.8.4 Manual Close

A manual close input to the reclosing/synchronismcheck device can be used to supervise closing bysynchronism check or LLDB/LBDL control. In addi-tion, the close input is used to ensure that the reclosingrelay will be in lockout state for the duration of thereset time.

3.9 Inhibit Control

In some applications, reclosing should be inhibiteduntil further action takes place in other devices. Forexample, if a dual-breaker scheme is used, and one ofthe breakers fails while clearing a fault, a transfer-tripsignal would be sent to the remote terminal to clear thefault contribution from that source. Reclosing of theremote breaker would be prevented until the transfer-trip signal was removed. (There must also be assurancethat removal of the transfer-trip signal is not a result ofchannel failure.) With this additional logic in thereclosing relay, the remote breaker can be reclosedreliably, simply by resetting the local breaker-failurelockout relay after the faulty breaker has been properlyisolated.

3.9.1 Drive to Lockout

When a ‘‘drive-to-lockout’’ input is energized, thereclosing relay will go into lockout from any point inthe sequence. The relay will stay in lockout until theinput is removed and the breaker is closed manually orby supervisory control. Upon removal, the recloser willgo through its reset sequence and return to ‘‘ready’’state.

3.10 Breaker Supervision Functions

Microprocessor reclosing relays provide a number ofbreaker operation functions to prevent excessivebreaker wear and aid in breaker maintenance schedul-ing. The number of maximum allowable reclosures perselected time period and recovery time can be set. Inaddition, counters for cumulative operations willprovide alarm when breaker service is needed.

3.11 Factors Governing Application ofReclosing

The factors governing the application of reclosing aresummarized below:

1. For instantaneous reclosing, the protectiverelay contacts must open in less than thebreaker reclose time. This presents no problemwith high-speed relays but on some slow-speedrelays, it may be necessary to reduce the contactfollow, or ‘‘wipe.’’

2. The breaker latch check (LC) and, whenapplicable, the low-pressure switch (LPC)should be used to avoid operating the breakerif the mechanism is not prepared to acceptclosing energy or gas pressure is inadequate.

3. The breaker should be derated according to thebreaker standards.

4. For instantaneous reclosing, arc deionizing timemust be considered (see Sec. 3.3).

4 CONSIDERATIONS FOR APPLICATIONS OFINSTANTANEOUS RECLOSING

The applications of instantaneous reclosing fall intothree categories:

1. Feeders with no-fault-power back-feed andminimum motor load

368 Chapter 16

2. Single ties to industrial plants equipped withlocal generation

3. Lines with sources at both ends

4.1 Feeders with No-Fault-Power Back-Feedand Minimum Motor Load

Instantaneous reclosing can be applied to these feeders,but care must be taken to avoid reclosing into motorsthat are still rotating, since their internal voltage maybe out of phase with the system voltage.

4.2 Single Ties to Industrial Plants with LocalGeneration

Since these circuits must be opened at the plant beforereclosing, instantaneous reclosing at the utility end isnot practicable unless instantaneous tripping of theplant tie or local generator is assured. Without thisinstantaneous tripping, the local generation, even whenquite small, can maintain the arc for line faults andnegate successful reclosure. Reclosing may also occurout of synchronism.

4.3 Lines with Sources at Both Ends

Simultaneous tripping of the line is necessary ininstantaneous reclosing that invariably requires someform of pilot relaying. Both ends of the line can beinstantaneously reclosed only if there are sufficient tiesbetween the terminals or sufficient inertia in bothsystems to ensure that the two ends will not go out ofsynchronism during the dead time. In the absence ofsufficient ties (or inertia), one end can be reclosedinstantaneously by LBDL reclosing, with the otherthen closed manually or by a synchronism check.

Instantaneous reclosing of both ends of a linewithout any checking is widely practiced for high-voltage transmission lines in the United States. Theselines have multiple parallel circuits and are protectedwith pilot relaying. This reclosing practice is usuallysuspended when the pilot relaying is out of service,since zone 1 phase and ground or ground instanta-neous relaying do not provide 100% instantaneous lineprotection. Sequential tripping and instantaneousunsupervised reclosing will produce unsuccessfulreclosing.

5 RECLOSING RELAYS AND THEIROPERATION

5.1 Review of Breaker Operation

Knowledge of the operation of the breaker and itsauxiliary contacts is essential to understanding howreclosing relays function. The time sequence of eventsoccurring within the breaker and its auxiliary contactsduring a typical instantaneous reclosing cycle is shownin Figure 16-1. The auxiliary contacts are actuateddirectly by breaker main contact travel (a and b) or bythe operating mechanism travel (aa and bb).

5.2 Single-Shot Reclosing Relays

A typical breaker-control scheme is shown inFigure 16-2. Contact 79 provides the reclosingintelligence. In some varieties of reclosing relays, thiscontact is pre-closed, waiting only for the 52bb contactto close to initiate ‘‘closing.’’ This 52bb contact closes

Figure 16-1 Typical circuit breaker instantaneous reclosing

cycle.


as the breaker moves to the open position in responseto protective relay action (or for a ‘‘manual’’ trip). Theclosing of the breaker is indicated to the reclosing relayby the opening of the 52b switch. When it opens, thereclosing relay becomes locked out. If reclosing issuccessful, the 52b stays open, causing an internaltimer to reset the reclosing relay, allowing the sequenceto be repeated at a later time. If reclosing is unsuccess-ful (fault still present), the 52b opens before timing iscomplete, and the reclosing relay stays locked out.

For the manual trip, 101/SC opens and stays openas a result of the movement of the control switchhandle by an operator to the trip position. This contactis closed in ‘‘close’’ and closed ‘‘after close.’’ It is openin ‘‘trip’’ and open ‘‘after trip.’’ It ‘‘slips’’ and istherefore called a slip contact. No reclosing is desiredfollowing a manual trip, and the slip contact supervisesthis.

The breaker-control scheme of Figure 16-2 is onlyone of a number of schemes used, but the funda-mentals of automatic reclosing are similar.

5.2.1 Solid-State Single-Shot Reclosing Relay

The logic diagram for the SGR-51 is shown inFigure 16-3. It requires only a 52b contact to indicatebreaker status. The single-shot function provides anoutput during the closing stroke of the breaker. It has ashort-duration output immediately following a 1 input.The output then reverts to 0, regardless of whether theinput 1 is short or continuous.

With the SGR-51 reset and the breaker closed, the52b switch is open, and a continuous 1 exists at thesingle-shot input with a steady-state 0 output. For thiscondition, the flip-flop outputs are as shown. With thenegated input to the upper amplifier, relay CR iscontinuously energized, providing a closed contact CRin the breaker close circuit.

As the breaker is tripped by protective relays, theonly open contact in this breaker close circuit, 52bb,closes (Fig. 16-3). As the breaker moves, 52b closesto produce a 0 input to the single shot. Then, as thebreaker recloses, 52b opens, putting a 1 on the single-shot input. A short 1 output follows to operate theflip-flop. The upper output of the flip-flop changes to1, putting a 1 on the negated amplifier input anddeenergizing CR relay. This opens contact CR in theclose circuit. The lower output of the flip-flopchanges to 0, operating the amber lamp to indicatea lockout.

If the breaker stays closed (52b open), the two 1inputs to AND permit an output to the reset timer.If this condition continues for the reset interval(adjustable from 3 to 30 sec), the lower input (c)to the flip-flop is energized. This resets the flip-flop,turns off the amber light, and energizes the CRrelay, making it ready for the next automatic recloseoperation.

If the breaker retrips before the reset timer timesout, the closing of 52b removes the 1 from AND tostop the timer and prevent the reset. Further action isblocked until the breaker is closed manually.

Figure 16-2 External schematic of the SGR-12 reclosing relay.

370 Chapter 16

5.2.2 Solid-State SGR-52 Relay Operation

SGR-52 logic, shown in Figure 16-4, is similar to thatfor SGR-51 but has reclose-initiate and reclose-blockfunctions. Also, the closing contact CR is notpreclosed. To initiate reclosing, CR must be energizedvia the reclose-initiate circuit. This input is under thecontrol of logic that identifies pilot tripping hasoccurred or some other consideration that assures thesuccess of high-speed unsupervised reclosing. Thereclose-initiate circuit includes a 100/0-msec timerthat allows time for the 52b contact to operate, evenif only a momentary closure of the reclose-initiatecontact occurs.

When a trip for which reclosing is desired (such as apilot trip) occurs, the reclose-initiate auxiliary contactcloses to provide a 1 input to OR-1. The outputdeenergizes the 100/0 timer and places a 0 on the inputof OR-2. This puts a 0 on the negated AND-1 input.As the breaker opens and 52b closes, the middlenegated input to the AND-1 goes to 0. If a lockout hasnot occurred, the three-input AND-1 is satisfied toinitiate the reclosing timer (0 to 2 or 2 to 20 sec). Theoutput of the timer satisfies AND-2 to operate CR andinitiate reclosing.

The 52b switch may close some time after thebreaker actually interrupts the flow of fault current. Ifthe reclose timer is set on 0, the CR relay will beenergized at the same time as the 52b switch closure.This produces only a slight delay in reclosing relativeto the SGR-51 with its preclosed CR contacts. For

EHV applications where intentional delay is required,the delay is provided by the 0- to 2-sec timer. Whenlonger delays are required, 2- to 20- or 6- to 60-sectimers are available.

Reclose blocking is necessary for functions such asout-of-step tripping and breaker-failure transfer-trip-ping. The reclose block input to OR-3 operates a timerto input OR-2. This output of OR-2 overrides theinitiate input and blocks reclosing by removing one ofthe AND-1 inputs, even though the pilot system maybe calling for a reclosure. To avoid critical resettingbetween initiate and block inputs, the 0.2/60-msectimer provides a continuing block signal for 60msec,after removal of the blocking input.

The lockout and reset operation occurs in the samemanner as described above for the SGR-51 relay.

5.3 Multishot Reclosing Relays

5.3.1 Electromechanical RC Relay

This relay provides multishot (up to 6) reclosuresthrough the use of drum-operated switches that allowvarious reclosing strategies to be elected. Timing iscontrolled by a small synchronous motor that drivesthe drum. Instantaneous-trip lockout, multishot select-able time interval reclosing, intermediate lockout, andinitiated first-shot instantaneous or time-delayedreclosures are possible using this device.

Since resetting following successful reclosure isdependent on full drum travel, a minute or more of

Figure 16-3 Logic diagram of the SGR-51 reclosing relay.


time may be required to accomplish this. Faultsoccurring during this resetting period will producelockout. Fast resetting of solid-state and microproces-sor relays overcomes this problem.

5.3.2 Solid-State 79M Relay Operation

The 79M circuit-shield reclosing relay provides up tothree shots. Each shot has its own reclose time andreset time settings and the first shot can be set toinstantaneous. The 79M close contact remains closeduntil 52b opens. The instantaneous cutout contactopens on a selected trip count and closes on reset orlockout.

A tap changer cutout contact is provided that openson first trip and closes on reset or lockout.

5.3.3 Microprocessor-Based Reclosing Relays

Several varieties of microprocessor-based relays areavailable. Functionally, they are similar to theirelectromechanical and solid-state counterparts, but inaddition, possess the qualities one expects frommicroprocessor relays: self-testing, ‘‘watchdog’’ timing,failure alarm, etc.

One example the, MRC-2, provides the following:

1. Instantaneous-trip enable (from selected statesof the sequence)

2. Drive to lockout (from external contact)3. Load tap changer lockout (to avoid stepping

during reclosing)4. In-progress output (to identify remotely that the

reclosing relay is sequencing)

Figure 16-4 Logic diagram of the SGR-52 reclosing relay.

372 Chapter 16

plus independent alarming of

1. Lockout2. Failed reclose3. Alarm (processing failure)

5.3.4 Reclosing in Feeder and DistanceProtection Terminals

The feeder protection DPU2000R and distance protec-tions REL301/302, REL512, and REL521/531 all haveoptional reclosing functions. A summary of thereclosing relay implementation in these protections isgiven here.

Reclosing in Feeder Protection DPU2000R

DPU2000R provides three overcurrent steps that canbe coordinated with up to four-shot reclosing. Eachcurrent element can be enabled or disabled at eachstep. For instance, instantaneous-trip lockout can beemployed as described in Section 3.6.

79 CUTOUT TIMER. The 79 cutout timer (79-CO)function provides for the detection of low-levelintermittent faults prior to the resetting of the reclosesequence as shown in Figure 16-5. At the end of theselected cutout time period, all overcurrent functionsare reenabled based on the 79-1 settings. In fuse-savingapplications involving downstream fuses, the 50P and50N instantaneous functions are set below the fusecurve to detect faults on tapped laterals. Thesefunctions are blocked after the first trip in the reclosesequence. The 51P and 51N time overcurrent functionsare set above the fuse curve. This results in the

upstream protection being less sensitive to anintermittent or low-level fault during the subsequentreclose operations. If the reset time is too short, thereclosing relay may reset before the fault is detectedagain. If the reset time is too long, the intermittent orlow-level fault is not cleared fast enough by theupstream protective device. In schemes using discretereclosing relays, blocked instantaneous overcurrentfunctions are placed in service only after the reclosingrelay has reset. However, the 79-CO function inDPU2000R reenables the instantaneous functions atthe end of the selected cutout period. Set the time forthe 79-CO function according to how long it takes adownstream fuse or other protective relay to cleardownstream faults. The typical time setting is between10 and 15 sec. If an intermittent or low-level faultexists, it will be detected at the end of the 79-CO cutouttime period, and the DPU2000R will trip and continuethrough the reclose sequence until the fault ispermanently cleared or lockout is reached. The 79-CO function allows the reset time to be set beyond60 sec without jeopardizing sensitivity to intermittentor low-level faults.

79V VOLTAGE BLOCK FUNCTION. The DPU2000R79V voltage block function blocks reclosing when oneor more of the input voltages is below the 79V voltagesetting. When the input voltage is restored within the79V time delay setting, the recloser operation isunblocked and the open time will begin. If the voltageis not restored within the 79V timer setting, the recloserwill proceed to lockout. This function is useful inpreventing a feeder breaker reclosure when the busvoltage is low because an upstream device has tripped.

Figure 16-5 Cutout time.


When voltage is restored following successful reclosingof the upstreambreaker, the reclosing cyclewill proceed.

REL512 Reclosing Coordination for Ring BusApplications

When applying the REL 512 reclosing to a ring busconfiguration, it is important to understand that therecloser module performs the reclosing function usingits own I/O, independent of the protection function.The only exceptions are the reclose initiate and blocksignals developed in the protection which are passedinternally to the reclose module. All other inputs,including bus and line system ac voltages and binary(status) dc voltage, are external inputs connected viatermination blocks associated with the reclosingmodule. Thus, this system allows the protectionfunctions to trip two breakers and the reclosingfunction to control one breaker. This method ofprotection and reclosing control is shown in Figures16-6 and 16-7. For simplicity, all REL 512 protectionfunctions are identified by 21 and reclosing functionsby 79.

There are separate voltage and breaker auxiliarycontact connections made to the protection andreclosing modules of the relay. For ring bus applica-tions, both breakers common to the protected lineshould have their auxiliary contacts appropriatelyconnected to the protection module via programmableinputs (middle row of terminal blocks). When theprotection operates both breakers trip and providetheir change of state information to the protection.

For reclosing, one of the line’s breakers is desig-nated as the lead breaker and the other is the followbreaker. Only the lead breaker’s auxiliary contacts areconnected to the reclosing module. It is recommendedto use the 52b. Connecting the 52a is optional.

The protected line’s voltage is connected to theprotection. This voltage is also connected to reclosermodule as the line voltage. It will be used to control thelead breaker. Furthermore, this voltage is alsoconnected to the reclosing module of the adjacentrelay that shares control of the follow breaker. It willbe connected as the bus voltage there.

The REL 512 pilot tripping for a fault on itsprotected line will provide a reclose initiate signal

Figure 16-6 REL-512 protection control of ring bus circuit breakers.

374 Chapter 16

internally to the recloser module within its chassis. Italso will provide an Rl signal for the reclose module ofthe adjacent relay that controls the follow breaker.

If the lead breaker fails to successfully reclose it willadvance to ‘‘lockout.’’ When in this state the reclosingmodule will need to lock out the other breakerassociated with this line. This situation requires theuse of the reclose module lockout output to drive theadjacent recloser controlling the follow breaker to‘‘lockout’’ via its input ‘‘drive to lockout.’’

A failed reclose will be the result of either exceedingthe maximum wait time waiting for the appropriatevoltage supervision to be satisfied or by reclosing into afault. In the latter condition a ‘‘reclose-block’’ signalwill be generated. This signal will be passed internallyto the recloser in the same chassis if the reclose shotnumber has the reclose-block enabled. The reclose-block signal can also be mapped to a protectionprogrammable contact and used to lock out theadjacent reclosing function.

The number of reclosures is set to 2. The successfulreclose reset time delay should be set to 10 sec. Thistime is arbitrary, but should be coordinated with the

shot #2 dead time (reset timeþmargin> shot #2 deadtime).

The reclose shot #1 is set up for high-speed reclosingas required. Generally the reclose will be set up with novoltage supervision or HBDL supervision. If theremote end has closed first and voltage has beenrestored, then synchronism checking should also beenabled if HBDL is enabled. Otherwise reclose will notbe attempted if the line voltage has been restored.

The reclose shot #2 is required to provide the sync-check reclose. The reclose dead time is 12 sec. Themaximum wait time should be set longer than theshot #1 maximum wait time. Reclose block is enabled.This will allow immediate blocking and bypassing ofshot #2 of the lead breaker for CIFT and advance therecloser to ‘‘lockout.’’

A successful reclose sequence follows:

1. A fault occurs on line M-H.2. 21-H trips breakers 11 and 12 and sends

internal RI to 79-H and external RI to 79-F.3. 79-H starts shot #1 0.5 sec. dead time count.

79-F bypassed shot #1 (RI #2 input starts at

Figure 16-7 REL-512 reclosing control of ring bus circuit breakers.


shot #2) and starts shot #2 12.0 sec. dead timecount.

4. 79-H recloses breaker 11 after 0.5 seconds ofdead time on HBDL.Recloses on synchronism check if line voltagehas been restored from remote end.

5. 79-H resets 10 seconds after reclose (thisassumes HBDL was immediately satisfied).

6. 79-F recloses breaker 12.0 seconds on synchron-ism check.

7. 79-F resets 10 seconds later.

The following is a failed reclose sequence:

1. A fault occurs on line M-H.2. 21-H trips breakers 11 and 12 and sends

internal Rl to 79-H and external Rl to 79-F.3. 79-H starts shot #1 0.5-sec dead time count.

79-F bypassed shot #1 (RI #2 input starts atshot #2) and starts shot #2 12.0-sec dead timecount.

4. 79-H recloses breaker 11 after 0.5 sec of deadtime on HBDL into fault.

5. 21-H trips on CIFT and sends internal blockreclose to 79-H.

6. 79-H advances to lockout and sends externalsignal to 79-F to advance to lockout.

Breaker Terminal REB551 in Breaker-and-a-HalfApplications

The microprocessor breaker terminal REB551 is acomplete protection package for functions associatedwith a circuit breaker and may in addition to reclosingand synchronizing functions also include breakerfailure protection. The recloser function of thisterminal (which is also implemented in distance relayterminals REL521/531 and line differential terminalsREL551/561) is briefly described here.

In breaker-and-a-half applications, the breakercontrol functions, such as reclosing, are associatedwith the breaker rather than with the line protection.In breaker-and-a-half substations, there are twobreakers that trip for a line fault. The reclosing ofthe breaker should be made sequentially in order tominimize breaker wear. First one breaker (MASTER)recloses and if it does not trip again (the fault wastransient), the second breaker (SLAVE) is allowed toclose. In case the fault is permanent, the MASTERbreaker trips again, and closing of the SLAVE breakeris blocked.

Figure 16-8 shows the operation for a transientfault. The sequence is as follows:

1. Both circuit breakers, CB1 and CB2, are trippedfor the line fault.

Figure 16-8 Sequential reclosing in breaker-and-a-half application.

376 Chapter 16

2. The MASTER recloser starts its dead time andissues a blocking signal, WF MASTER (waitfor MASTER), to the SLAVE recloser.

3. The MASTER recloses its breaker, CB1, andremoves the blocking signal.

4. The SLAVE starts its dead time and recloses itsbreaker CB2.

For a permanent fault, CB1 will trip again and boththe MASTER and SLAVE recloser locks out, providedthat a single shot reclosing was set. For multiple-shotreclosing, CB1 will continue its cycle and CB2 will beallowed to reclose only when CB1 has been closedsuccessfully for 1 sec.

6 SYNCHRONISM CHECK

Synchronism-check relays verify that the voltages onthe two sides of a breaker are approximately the samein magnitude and phase in order to ensure that aminimum impact occurs to the power system when acircuit breaker is closed. The synchronism-check relaysupervises automatic or manual closing of a circuitbreaker. One type, the CVX, is shown in Figure 16-9and its connections are typical. The 52b contactensures that the CVX contact is open immediatelyafter tripping and that a conscious determination ofsynchronism is made prior to closure.

The driving force for current flow following breakercontact closure is the voltage appearing across theopen contacts just prior to closing. The difference

between the voltages on the two sides of the breaker iscalled the phasing voltage. The phasing voltage can becontrolled by several methods; the conventionalangular and the more precise phasing voltage methods.

6.1 Phasing Voltage Synchronism CheckCharacteristic

The phasing voltage method provided by CVX, andalso the circuit shield type 59S/V, form the character-istics shown in Figures 16-10 and 16-11. The line-to-ground voltage (bus voltage) will provide a reference,and a comparison is made with an input voltage whichis supplied to the relay from the other side of thebreaker that is to be controlled (line voltage).

The phasing voltage method verifies that the voltagedifference between the two voltages, bus voltage andline voltage, is less than the set limit determined by theangular adjustment. The normal angular adjustment is208, which may be increased to 608 if such a wideclosure angle will not disrupt the system. Figure 16-10describes the relationship required in order for closingof the circuit breaker to be permitted. The center of thecircle is established by the reference voltage phasor(bus voltage). The radius of the circle is determined bythe voltage difference setting. Operation results whenthe synchronizing voltage (line voltage) falls inside thecircle. A ‘‘close’’ output is produced in response to areclosing-directed synchronism check or a manualclose command. The 52b contact assures that syn-chronism which exists prior to tripping is not allowed

Figure 16-9 Synchronism check relay schematic diagram.


to falsely indicate that synchronism exists followingtripping.

Usually in conventional synchronism-check relay-ing, a relatively long time measurement is used toinsure that the voltages across the open breaker are insynchronism. However, this long time delay, whichmay be 10 sec or more, is undesirable if both ends ofthe line are being reclosed at high speed. If the timemeasurement is shortened, a faster synchronism-checkmeasurement can be made, but this may result inreclosing for a nonsynchronous condition with slipfrequencies that are higher than desired for properreclosing. A slip cut-off frequency function can provide

a high-speed synchronization determination whenvoltages are in synchronism without the risk ofreclosing if high slip frequencies are actually present.Figure 16-12 plots the maximum frequency across theopen breaker that will allow the CVX contact to close.For example, at a 208 setting of angle of closure and aslip frequency of 0.0167 (the value that gives asynchronscope pointer movement at the rate of aclock’s second hand) would require a time-dial settingon the CVX of 2 or less.

6.2 Angular Synchronism Check Characteristic

The angular method is the one most commonly used inmicroprocessor relays. The line-to-ground voltage(VBUS) will provide a reference, and a comparison ismade with an input voltage (VSYNC) which is suppliedto the relay from the other side of the breaker that is tobe controlled. It verifies that the angular difference(ySYNC) is within the acceptable range of set angle andverifies both voltage magnitudes are in the acceptableband between VMAX and VLIVE. The resulting syn-chronism region looks like an automobile windshield

Figure 16-10 Phasing voltage synchro-check relay closing

characteristic.

Figure 16-11 Typical voltage angle characteristics of CVX

for various angle settings. (Rated voltage on one circuit.)

Figure 16-12 Approximate maximum slip frequency for

which operation occurs. (Rated voltage on both sides.)

378 Chapter 16

wiper stroke with VBUS at the center as shown inFigure 16-13. Operation results when the synchroniz-ing voltage (VSYNC/LINE) falls inside the synchronizingregion and the slip frequency is below the set value ofSLIP FREQ.

The synchronism-check function is intended toassure that the two segments of the power systemthat are to be interconnected by the closure of abreaker are in synchronism with a standing angle. It isnot intended to provide an automatic synchronizingfunction for two systems that are operating at differentfrequencies. The relays contain no provision forenergizing the closing coil at the precise angle thatwould ensure zero voltage across the breaker contactsat the instant of closure.

7 DEAD-LINE OR DEAD-BUS RECLOSING

At some locations, reclosing is initiated when the bus ishot and the line dead, or vice versa. The synchronism-check function will not provide an output to permitreclosing if no voltage is present on one or both sidesof the open breaker. With the possibility that eitherside of the breaker, bus or line, may be deenergized(dead), hot-busdead-line, dead-bushot-line, or dead-busdead-line checks are necessary. In this case single-or three-phase voltages may be checked to establishclearly that the voltages are as they appear, and notpossibly erroneous because of such things as fusefailure. This arrangement may be used, for example, toenergize one end of a transmission line if the bus is hot.Reclosing at the other end of the line could besupervised by a synchro-check relay. The CVX relaycontains elements for the synchronism-check, LLDB,and LBDL controls. Figure 16-14 illustrates its basic

operation for controlling automatic breaker closure.Manual breaker closures by 101C can also besupervised as shown.

To avoid any possibility of pumping, synchronism-check closures should always be made under control ofa reclosing relay or similar device (see Fig. 16-14). Thesynchro-verifier relay does not include an antipumpprovision, nor will the antipump scheme in the breakerreliably prevent pumping when closing into a perma-nent fault.

8 AUTOMATIC SYNCHRONIZING

A synchronizing system can be used at unattended orattended locations for automatic synchronizing orsupervision of manual synchronizing. The heart ofthis system is the synchronizer, which compares thevoltage on two sides of an open breaker and energizesthe breaker close coil under the following conditions:

1. If the frequency difference is below a presetamount

Figure 16-13 Angular synchronism check characteristic.

Figure 16-14 Typcial synchronism check control of breaker

closing using CVX-1.


2. At such a phase angle that the breaker contactsclose when the systems are in phase

The automatic synchronizer Synchrotact 5 can beequipped with the following options:

1. An acceptor to restrict the range of voltages atwhich breaker closing in initiated

2. A governor control auxiliary to initiateangular adjustments needed for synchroniz-ing

3. A regulator control auxiliary to initiate voltage-level adjustments needed for synchronizing

4. A synchronism-check device to avoid possibleclosure because of component failure in theautomatic synchronizer

380 Chapter 16

17

Load-Shedding and Frequency Relaying


1 INTRODUCTION

When a power system is in stable operation at normalfrequency, the total mechanical power input from theprime movers to the generators is equal to the sum of allthe connected loads, plus all real power losses in thesystem. Any significant upset of this balance causes afrequency change. The huge rotating masses of turbine-generator rotors act as repositories of kinetic energy:When there is insufficient mechanical power input tothe system, the rotors slow down, supplying energy tothe system. Conversely, when excess mechanical poweris input, they speed up, absorbing energy. Any changein speed causes a proportional frequency variation.

Unit governors sense small changes in speedresulting from gradual load changes. These governorsadjust the mechanical input power to the generatingunits in order to maintain normal frequency operation.Sudden and large changes in generation capacitythrough the loss of a generator or key intertie canproduce a severe generation and load imbalance,resulting in a rapid frequency decline. If the governorsand boilers cannot respond quickly enough, the systemmay collapse. Rapid, selective, and temporary drop-ping of loads can make recovery possible, avoidprolonged system outage, and restore customer servicewith minimum delay.

2 RATE OF FREQUENCY DECLINE

Before designing a relay scheme for system overloadprotection, it is necessary to estimate variations in

frequency during disturbances. Figure 17-1 shows asystem S that consists of two interconnected subsys-tems, S1 and S2. For all of S, the following relationshipmust hold true for constant-frequency operation:

Generation ¼ loadsþ losses ð17-1ÞThere can, however, be more generation than load inS1 and more load than generation in S2, with thedifference being transferred by an intertie as shown. Ifthe total loads and losses are equal to the totalmechanical power input, there will be no change ingenerator speed or frequency with time.

If, however, the tie is suddenly lost as a result of apermanent fault, the kinetic energy in the S1 generatorsmust increase to absorb the excess power input; that is,the generators must speed up. Conversely, the S2generators must slow down. The classical expressionfor the initial rate of change of frequency is

df

dt¼ �DP

2Hð17-2Þ

where

df

dt¼ per unit initial rate of change of frequency

Dp¼ decelerating power in per unit of connectedkVA

H¼ inertia constant,MW� sec

MVAor

KW� sec

kVA

In this text, all H constants and DP values are relatedto a kVA rather than kW base.

The inertia constant (H) is defined as the ratio of themoment of inertia of a generator’s rotating compo-nents to the unit capacity. It is the kinetic energy in

381

these components at the rated speed. For example, aturbine-generator rated at 100MVA, with an inertiaconstant of 4, has a kinetic energy of 400MW-sec, or400MJ, in its rotor when spinning at rated speed. Ifboth power output and load were constant withdeclining frequency and speed, the generator couldsupply its full load (with p¼ 1) for 4 sec, with no powerinput to the turbine, before the rotor would come to acomplete halt. The inertia constant H for an individualunit is available from the manufacturer or may becalculated from

H ¼ 0:231WR2 RPM2 10�6

kVAð17-3Þ

For a system, a composite value is calculated asfollows:

Hsystem ¼ H1MVA1 þH2MVA2 þ . . .þHnMVAn

MVA1 þMVA2 þ . . .þMVAn

ð17-4Þwhere subscripts 1, 2, . . . , n refer to individualgenerating units.

The larger the inertia constant, the slower thefrequency decline for a given overload. Older water-wheel generators, with their massive rotors, haveinertia constants as large as 10. Newer turbine-generator units, however, may have inertia constantsof only 2 or 3, since the trend is toward larger outputswith smaller rotor masses. Power systems are becom-ing more prone to serious frequency disturbances forgiven amounts of sudden load change.

An example is shown in Figure 17-1. System S2 inthe figure has a net load of 1200MW and totalgeneration of 1000MW. The tie must carry 200MWfrom S1 to S2. The inertia constant for S2 is 4, and thepower-factor rating of the machines there is 0.85. If thetie is suddenly lost, the initial rate of frequency drop insystem S2 is calculated as follows:

DP ¼ tie load lost

kVA of S2¼ 200

1000=0:85

¼ 0:17 per unit

Then using Eq. (17-2), we obtain

df

dt¼ 0:17

2ð4Þ ¼ �0:0213 per unit

df

dt¼ �0:0213660 ¼ �1:275Hz=sec

The negative rate of change indicates frequency drop.As the frequency drops, experience has shown that

load power also decreases. A frequently used relation-ship is that 1% frequency drop produces 2% automaticload reduction. 1% frequency drop corresponds to0.6Hz. The resulting load change applies to the entireload in the system segment under consideration, not tothe DP, and in the example above would correspond to0.02 (1200)¼ 24 kW. This reduces the generationdeficiency to (200� 24)¼ 176 kW, slowing the decay.A frequency reduction of 8.35% would reduce the loadby 16.7% or 0.167 (1200)¼ 200 kW. However, at thisfrequency (55Hz), generating-plant station auxiliarieswould probably collapse and stable operation wouldnot be possible. Motor-driven auxiliaries will slowdown, reducing generator output. Safety margins ingenerator and motor cooling and bearing lubricationsystems may become dangerously small.

The simple diagram of Figure 17-2 may be used toillustrate a principle. Loss of generator 2 imposes anincreased load on generator 1. The load will continueto be supplied, but at the expense of decreasing speedof the rotating mass. The initial MW overload on the‘‘remaining’’ generation is exactly equal to the ‘‘lost’’MW generation. The load reduction which is influ-enced by the frequency reduction is related to theoriginal total load (1þDP). The new load followingfrequency reduction is 1.0, the original load on G1. Anew stable operating frequency is reached:

ff ¼ f0 1� DPdð1þ DPÞ

� �ð17-5Þ

where

ff ¼ new stable operating frequency in hertzf0 ¼ rated frequency in hertzDP ¼ per unit load reduction (per unit based on

remaining generation)d¼ per unit change in load per unit change in

frequency

The d factor may vary from 1/2 to 7, depending on themix of loads, although typically most utilities assumed¼ 2 (that is, a 2% decrease in load for each 1%decrease in frequency). An exact value of d can be

Figure 17-1 Interconnected system S.

382 Chapter 17

determined only by observing the variation of loadwith frequency on the system under consideration.

Figure 17-3 shows a typical frequency declineresulting from loss of generation. Figure 17-4 describesthe behavior of system frequency for several combina-tions of inertia constant and percent overload usingd¼ 2.0.

Most 60-Hz plants will operate down to 55Hz on atemporary basis, but under no circumstances shouldlong-blade turbines be permitted to operate loaded ona steady basis at a frequency below 59.5Hz or below58.5Hz for shorter-blade machines. The last rows oflow-pressure blades in steam and gas turbines aretuned to be free of resonance when operating at therated speed. Off-normal operation can produce failurein a matter of minutes.

Governor response will work to correct the defi-ciency (or excess) in system speed. However, time delayis involved in reestablishing a new stable relationshipin boilers, water flow, etc. To avoid operation atreduced frequency, it is necessary to shed load (tripcircuit breakers to disconnect load from its source ofpower).

3 LOAD-SHEDDING

For gradual increases in load, or sudden but mildoverloads, unit governors will sense speed change andincrease power input to the generator. Extra load ishandled by using spinning reserve, the unused capacityof all generators operating and synchronized to thesystem. If all generators are operating at maximumcapacity, the spinning reserve is 0, and the governorsmay be powerless to relieve overloads.

In any case, the rapid frequency plunges thataccompany severe overloads require impossibly fastgovernor and boiler response. To halt such a drop, it isnecessary to intentionally and automatically discon-nect a portion of the load equal to or greater than theoverload. After the decline has been arrested and the

Figure 17-2 Example of natural load-shedding.

Figure 17-3 Frequency change response to moderate over-

load (all per unit).

Figure 17-4 Behavior of frequency during overload.

Load-Shedding and Frequency Relaying 383

frequency returns to normal, the load may be restoredin small increments, allowing the spinning reserve tobecome active and any additional available generatorsto be brought online.

Frequency is a reliable indicator of an overloadcondition. Frequency-sensitive relays can therefore beused to disconnect load automatically. Such anarrangement is referred to as a load-shedding orload-saving scheme and is designed to reserve systemintegrity and minimize outages. Although utilitiesgenerally avoid intentionally interrupting service, it issometimes necessary to do so in order to avert a majorsystem collapse. In general, noncritical loads, usuallyresidential, can be interrupted for short periods,minimizing the impact of the disturbance on service.

Automatic load-shedding, based on underfre-quency, is necessary since sudden, moderate-to-severeoverloads can plunge a system into a hazardous statemuch faster than an operator can react. Under-frequency relays are usually installed at distributionsubstations, where selected loads can be disconnected.

The object of load-shedding is to balance load andgeneration. Since the amount of overload is not readilymeasured at the instant of a disturbance, the load isshed a block at a time until the frequency stabilizes.This is accomplished by using several groups offrequency relays, each controlling its own block ofload and each set to a successively lower frequency.The first line of frequency relays is set just below thenormal operating frequency range, usually 59.4 to59.7Hz. When the frequency drops below this level,these relays will drop a significant percentage of systemload. If this load drop is sufficient, the frequency willstabilize or actually increase again. If this first loaddrop is not sufficient, the frequency will continue todrop, but at a slower rate, until the frequency range ofthe second line of relays is reached. At this point, asecond block of load is shed. This process will continueuntil the overload is relieved or all the frequency relayshave operated. An alternative scheme is to set anumber of relays at the same frequency or closefrequencies and use different tripping time delays.

Techniques for developing schemes and calculatingsettings are described in Section 5.

4 FREQUENCY RELAYS

Many different types of frequency relays have beenused over the years. The induction-disk relay that wasthe forerunner of all frequency relays has faded intodisuse in favor of more accurate devices. Three general

classes of frequency relays are being applied: theinduction-cylinder relay, the digital relay, and themicroprocessor relay.

4.1 KF Induction-Cylinder UnderfrequencyRelay

This relay is fast and sufficiently accurate for mostapplications. The principle of operation of the relay isdescribed in Chapter 3 and is based on a circuit inwhich the phase angle changes as frequency changes.The phase relationship between the current in thiscircuit and a reference current produces torque thatchanges direction when the set-point frequency isreached. With this relay, as with all frequency relays,precaution must be taken in the design of the device tomake certain that the phase shift associated withphenomena such as faults, which appear as an extremechange of frequency due to the sudden change of phaseangle of the supply voltage, do not produce misopera-tion. In the case of the induction-cylinder relay, a timedelay of at least six cycles is required. The trippingcharacteristics for the KF induction-cylinder relay areshown in Figure 17-5. This type of plot is useful forpredicting the frequency at which tripping will occurduring frequency declines. It reflects the fact that thefrequency will continue to drop after the relay-settingfrequency is crossed, and during the time the relay isoperating. As a result, the actual contact closurefrequency will be somewhat below the set value.

The ‘‘cycles-of-delay’’ parameter associated witheach curve in the family is the intentional time-delaysetting after the cylinder unit closes its contacts. Theplot shown in Figure 17-5 includes the inherentcylinder operating time. An examination of the six-cycle delay curve with a 10-Hz/sec frequency declineindicates that trip contact will close when thefrequency is 2Hz below the setting. Total operatingtime following the crossing of the set-point frequencywill thus be 2/106 60 (base)¼ 12 cycles. In otherwords, the cylinder unit operates in six cycles and thetimer adds a six-cycle delay. Figure 17-6 shows anexample of the actual trip frequency that results fromthe necessary operating time of the induction cylinderand the required security delays. The relay in this casewas set for 58.9Hz and was unable to produce benefituntil the frequency had decayed to 57Hz. This is anexaggerated case. Frequency decay, in general, will notbe so severe and it will benefit from the automaticload-shedding resulting from frequency decay that isnot shown here.

384 Chapter 17

4.2 Digital Frequency Relays

Digital relays in general utilize a multimegahertzcounter. Zero crossing of voltage is detected, and acounter starts and continues counting until the nextvoltage zero or in some relays until the next positive-going zero crossing. The count accumulated isindicative of the period of the waveform and thus thefrequency is identified.

Accuracies of 0.005Hz are realizable utilizing thisconcept. Security is achieved by several expedients

such as requiring that an abnormal count occur inthree consecutive periods. One cycle of adequatefrequency will cause the relay to reset, requiring it tobegin the ‘‘count of 3’’ again. The sudden phase shiftcaused by a fault such as a ‘‘b to c’’ fault for a relaythat is sensing ‘‘a to b’’ voltage will be sensed as a shortperiod once but will not repeat.

4.3 Microprocessor-Based Frequency Relay

The principle applied in the microprocessor relay is thesame as that in a digital relay, but additionalsophistication is included. All the self-checking provi-sions and examination of various failure modes areconstantly achieved and alarm and lockout are aninherent part of these relays.

Multiple set points are common among digital andmicroprocessor frequency relays. Some may be usedfor over- or underfrequency applications and someinclude a ‘‘restore’’ function. The restore function maybe set at a frequency level to indicate that the powersystem has recovered and is now able to accommodatethe reapplication of the load that was shed. In general,a long time delay is required in the restore function toassure that pumping will not result. Often, an externaltimer is required.

5 FORMULATING A LOAD-SHEDDINGSCHEME

Several procedures and criteria must be consideredwhen designing load-shedding schemes for specificsystems. These include:

1. Maximum anticipated overload2. Number of load-shedding steps3. Size of the load shed at each step4. Frequency settings5. Time delay6. Location of the frequency relays

5.1 Maximum Anticipated Overload

Underfrequency relays should be able to shed a loadequal to the maximum anticipated overload. Logically,there is no reason to limit load-shedding to anypercentage of load. Indeed, it is preferable to shed100% of load, preserving interconnections and keepinggenerating units on line and synchronized, than toallow the system to collapse with customers still

Figure 17-5 Characteristics of 60-Hz induction cylinder

(KF) frequency relay.

Figure 17-6 Load shedding example.


connected. Even if 100% of the load is shed, service canbe restored rapidly; if the system collapses, a prolongedoutage would result. For this reason, it is necessary toevaluate the cost of the load-shedding scheme in lightof the probability that an overload of a given severitycan occur.

The system should be studied with respect to theoverload that would result from the unexpected loss ofkey generating units, transmission ties, and buses.Stability studies can help identify areas that, ifseparated or islanded from the rest of the system,would have a severe generation deficiency. These areaswill need more comprehensive load-shedding.

The load-reduction factor d should also be con-sidered, since it will reduce the overload once thefrequency has dropped. If spinning reserve, or addi-tional generation capacity equal to the overloadcompensated for by d, is not available shortly afterthe disturbance, it will be impossible to bring back thesystem to rated frequency. This will mean that anislanded system cannot be resynchronized, and inter-connections to neighboring utilities cannot be reclosed.(It should also be remembered that the turbine-generators must not be operated for extended periodsbelow rated speed.)

The load-reduction factor d is rarely known exactlyand may vary with time. To design a conservativescheme, which will tend to shed enough load for systemrecovery to normal frequency, it is safest to assumethat d equals 0.

5.2 Number of Load-Shedding Steps

The simplest load-shedding scheme is one in which thepredetermined percentage of the load is shed at oncewhen a group of relays senses a frequency drop.Although this scheme will arrest any anticipatedfrequency decline, it will often disconnect far morecustomers than necessary. A refinement then would beto use two groups of relays, one operating at a lowerfrequency than the other, and each shedding half thepredetermined load. The higher-set relays would tripfirst, halting the frequency decline as long as theoverload were half or less of the worst-case value. Formore severe overloads, the frequency would continueto drop, although at a slower rate, until the secondgroup of relays operated to shed the other half of theexpendable load.

The number of load-shedding steps can be increasedvirtually without limit. With a great many steps, thesystem can shed load in small increments until the

decline stops; almost no excess load need be shed. Sucha scheme may, however, inhibit system recovery. Asnoted below, it may also be difficult to coordinate somany steps.

Most utilities use between two and five load-shedding steps, with three being the most common.

5.3 Size of the Load Shed at Each Step

When possible, the size of the load-shedding stepsshould be related to expected percentage overloads.When a study of the system configuration, or astability study, reveals that there is a relatively highprobability of losing certain generating units ortransmission lines, the load-shedding blocks shouldbe sized accordingly. Sizing can be determined as in thefollowing example.

Assume that a power system has a generating plantA at a remote location, this plant is tied to the rest ofthe system by long lines, and the system also isconnected by a transmission tie B to a neighboringutility (Fig. 17-6). Assume in addition that A carries upto 20% of system load and B up to 12%. Stabilitystudies show that certain faults or disturbances mayresult in loss of synchronism between A and thesystem, so that its transmission ties must be opened.Furthermore, problems on the neighboring utilitysystem may necessitate the tripping of B. It isimportant for such a system to implement a load-shedding scheme that will preserve the remainingsystem if A and B are lost. It is logical, therefore, touse three load-shedding steps to handle overload

Figure 17-7 Frequency vs. time on 100% overload.

386 Chapter 17

resulting from (1) loss of A, (2) loss of B, or (3) loss ofboth A and B simultaneously. The overloads, in orderof increasing probability and seriousness, are listed inTable 17-1.

The following load-shedding steps are implementedto handle each situation in succession:

Step 1 Shed 12% of total load (12% of total).Step 2 Shed an additional 8% of remaining load

(20% of total).Step 3 Shed an additional 12% of remaining load

(32% of total).

% overload ¼ load� power input

power input6100 ð17-6Þ

Note that each step sheds only enough load to handlethe next, more serious contingency. Each step shouldbe evenly spread over the system by dropping loads atdiverse locations.

If the system under consideration is large, there maybe many possible combinations of events to consider,each causing only a small percentage overload in itself.In this case, a number of overload situations may belumped together and handled in one step. Conversely,it may be sufficient to shed a percentage of theoverload in a few equal steps. To implement theshedding evenly and at the distribution level, such asystem would require a large number of frequencyrelays distributed over the system. With so manyrelays, there is no cost penalty in using smaller steps(five, for example) to more closely balance generationand load, provided that all the steps can be coordi-nated.

5.4 Frequency Settings

The frequency at which each step will shed loaddepends on the system’s normal operating frequencyrange, the operating speed and accuracy of the

frequency relays, and the number of load-sheddingsteps.

The frequency of the first step should be just belowthe normal operating frequency band of the system,allowing for variation in the tripping frequency of therelay. The stable, solid-state type 81 relays or micro-processor relays may be set from 55 to 59.9Hz within0.01Hz of the lowest expected normal-frequencyexcursions to trip at the first indication of trouble.For induction-cylinder electromechanical relays, thehighest frequency setting should be approximately 0.1to 0.2Hz below the system’s lowest normal operatingfrequency. Whatever type of relay is used, thefrequency should be selected to avoid shedding forminor disturbances from which the system can recoveron its own.

The remaining load-shedding steps may be selectedas follows:

1. Based on the best estimate of DP, calculatedf/dt using Eq. (17-2). Employing relay trippingcurves, calculate the actual frequency at whichload will be shed by the first-step relays for themost severe expected overload. (See Fig. 17-6for guidance.)

2. Set the second-step relays just below thisfrequency, allowing a margin that will tolerateany expected frequency drift for both sets ofrelays.

3. Calculate the actual frequency at which thesecond load-shedding step will occur. The rateof frequency decline by the second-step relayscan be calculated as that resulting from themost severe expected overload minus the loadshed in the first step.

4. Again, allowing a margin for relay drift, set thethird-step relays below the lowest second-stepshedding frequency.

5. Repeat the calculations until settings areobtained for all steps. Determine the system’slowest frequency value before the final loadblock is interrupted for the worst-case overload.This value should not be below the system’slow-frequency operating limit.

Continuing with the example given in Section 5.3above, we may calculate frequency settings as follows.Assume the first load-shedding frequency is 59.5Hzand that H¼ 4 (based on remaining generator MVA).From Table 17-1 the worst expected overload is 47%(47% of remaining generator MVA). This will cause afrequency decline of approximately 3.5Hz/sec. The setpoint will be reached in 0.143 sec. By applying MDF

Table 17-1 Overloads Resulting From the Loss of A, B, or

Both

Event

Percent of

generation

lost

Percent

overload

from Eq. (17-6)

Loss of interconnection B 12 13.6

Loss of generator A 20 25

Loss of both A and B

simultaneously

32 47


relays with an effective 60-msec time delay anddistribution breakers with an interrupting time of fivecycles, the additional time to produce trip is 0.143 sec.The frequency at the time of interruption of loadwould be 59Hz.

With the first block of load chosen as 12%, there isat this point a reduction in the overload. Since DP mustbe based on the kVA of the remaining generation(68%) and the 12% is of the original total load, the loadis (100� 12)/68¼ 1.294. DP is then 29.4%. The rate offrequency decline becomes 606 0.294/(26 4)¼ 2.20Hz/sec.

A setting of 58.9 is usable for the next shedding step.This will be reached, with a 2.2Hz/sec decay rate, in(59� 58.9)/2.2¼ 0.045 sec. The relay delay and circuit-breaker opening time are again 0.143 sec. The totaltime for separating load is then 0.188 sec. Thefrequency would have decayed to 58.58Hz by thistime. A load corresponding to 8% of the original totalload is shed at this point.

The load now shed is 20% of the original total. Theload on the remaining generation is (100� 20)/68¼ 1.176 for a DP of 17.6%. The decay rate becomes60(0.176)/(26 4)¼ 1.32Hz/sec. A setting of 58.5Hz isrealizable for the next stage. The setting will be crossedin (58.58� 58.5)/1.32¼ 0.06 sec. As calculated before,the total time in this interval is 0.203 sec and thefrequency falls 0.268Hz for a shedding level of 58.3Hzfor dropping another 12% of the original load.

With 32% generation loss initially and 32% loadshed, the worst-case condition is handled with nofrequency excursion below 58.3Hz. Complete recoveryoccurs. With any lower level of generation loss,recovery will occur with less frequency drop in eachstage and fewer levels of underfrequency settings beingreached. Also with inherent load-shedding, the fre-quency decay would have been somewhat slower.

5.5 Time Delay

The above examples illustrate an important rule forload-shedding schemes: Use the minimum possibletime delay consistent with relay security. The less thedelay, the more easily the scheme can cope with severeoverloads. All unnecessary interposing auxiliarydevices should be avoided.

Naturally, there are exceptions when an extra timedelay may be needed. One such case is a frequencyrelay connected to a potential supply from a bus thatsupplies induction-motor loads (Fig. 17-8). If the linebreakers 1 and 2 trip and interrupt current to themotors, they will slow down rapidly. Because of

trapped, decaying flux, the motors will excite the buswith ac potential of decling frequency, possibly for0.5 sec or longer. As a result, the load-sheddingfrequency relay will trip unnecessarily and lock outthe feeder breakers 4, 5, and 6. When remote breakers1 and 2 reclose, service to the motors will not berestored and breakers 4, 5, and 6 must be reclosedmanually. This situation has caused frequency relayswith up to 30 cycles of delay to trip.

An intentional time delay long enough to ridethrough the residual voltage collapse is quite often toolong to be consistent with load-shedding requirements.A more effective method is to supervise the under-frequency relay using the overcurrent relay (50)connected to the source current transformer, as shownin Figure 17-8. The frequency relay (81) will tripbreakers 4, 5, and 6 and shed load only whensignificant load current is flowing into the bus.

Consideration should be given to the desirability ofallowing the motors to remain on the line as thereclosing takes place. This may produce a severephysical impact on any motor that has a back-voltagethat is 1808 out of phase with the reapplied systemvoltage. The main breaker (breaker 3 in Fig. 17-8) isoften tripped to avoid just such an unfavorablereenergization.

5.6 Location of the Frequency Relays

In large systems, the load-shedding relays should bespread throughout the system to avoid heavy powerflows and undesirable islanding. Load-shedding in one

Figure 17-8 Underfrequency relay with induction motor

load.

388 Chapter 17

concentrated area, for example, can cause heavy powerflow over transmission lines from the area where theload was shed to areas of excess load. Because of theoriginal disturbance, these lines may already beoperating at high emergency levels, and the unevenload-shedding may cause thermal overload or systeminstability.

Concentrated loss of generation in certain areas ofthe system will also result in frequency dispersion; thatis, the frequency in the overloaded areas will dropfaster than elsewhere. The difference in frequenciesnaturally produces rapidly increasing torque angles onthe transmission lines, which may cause the system togo out of step. Fortunately, load-shedding relays in thearea of greatest frequency decline will trip first. Thisaction alleviates the uneven loading, helps to bringback the system to uniform frequency, and avoids theimpending loss of synchronism. It is clearly important,however, to install some extra load-shedding capabilityin any portion of the system that is prone toconcentrated overload.

Finally, load-shedding priorities must be estab-lished. The nature of the loads shed can usually becontrolled only by tripping feeders at the distributionlevel. The implication is that frequency relays will beinstalled in many distribution substations and willcontrol relatively small blocks of load.

6 SPECIAL CONSIDERATIONS FORINDUSTRIAL SYSTEMS

Load-shedding programs are recommended for indus-trial power systems. Frequency relaying is highlydesirable for those systems in which loads are suppliedeither exclusively by local generation or a combinationof local generators and utility ties. Power must often bemaintained to certain essential processes to avoiddanger of personal injury, equipment damage, productloss, or process disruption.

For local generators, the same type of single- ormultiple-step frequency-based load-shedding programcan be applied as that described for utility systems.Special precautions may be necessary, however, toaccommodate the relatively small number of powersources, each of which can supply a considerable partof the total load. This type of scheme can producedifferent, more serious disturbances.

For example, the scheme shown in Figure 17-9a isfor a plant that generates about half its own powerrequirements, the balance being supplied through autility tie. If this tie is lost, the local generators will be

100% overloaded, and the rate of frequency drop willbe 7.5Hz/sec (if we assume that H¼ 4). Clearly, thisplant needs an exceptionally fast load-shedding schemethat can drop a large percentage of low-priority loadwith minimum delay. There may be no time formultiple steps. At the same time, the need for speedand sensitivity is often incompatible with securityrequirements for milder disturbances.

One solution, shown in Figure 17-9a, is to set thefrequency relay at a high, near-normal frequency andwith minimum delay. Its tripping circuit is supervisedby an undercurrent relay, whose contact closes if feederservice from the utility is lost. The large, sudden

Figure 17-9 Control of industrial plant load-shedding.


overload on the local generators is relieved by trippinglow-priority plant loads.

An example of an interruptible load would be for apulp mill where the chipper operation is not a part ofthe continuous papermaking process and, therefore, itstemporary interruption would not interfere with (andindeed might save) the process. Temporary interrup-tion of nonvital lighting in an industrial plant may alsobe considered for overload relief.

If the tie is in service and the utility suffers afrequency disturbance, the undercurrent relay willprevent load-shedding in the plant, while the utilitysheds its overload in lower-priority areas.

The scheme of Figure 17-9b trips breaker 3 for anyoutfeed of power from the plant and then sheds load asrequired. A variation of this scheme uses the samecomplement of relays, but uses the ‘‘watt’’ relay tosupervise the 81 trip. This prevents plant load-sheddingfor cases where the plant is supporting the utilitysystem and the frequency is low.

Figure 17-10 describes a load-shedding schemewhich may be used in a plant in which there is nolocal generation. Simple time-delayed overcurrentrelays are used to drop expendable load when onetransformer is out of service and the other becomesoverloaded. It is assumed that, with proper planning,the maximum load will not exceed the capability ofboth transformers, and load shedding is required onlywhen one transformer is out of service.

7 RESTORING SERVICE

In general, the reclosing of feeders that have beentripped for load-shedding is left to the discretion ofsystem or station operators. Frequency relays can beused, however, either to supervise restoration orrestore loads automatically.

The following considerations apply to any restora-tion of service, whether manual or automatic:

1. Frequency should be allowed to return tonormal before any load is restored. Reclosing feederswhen the frequency is still recovering may plunge thesystem back into crisis and will certainly preventreunification of islands. Resetting of load-sheddingfrequency relays cannot be used for the supervision ofrestoration.

2. Once the frequency has returned to normal, allserviceable interconnections must be allowed toresynchronize and reclose. Unifying an islanded systemas much as possible generally facilitates servicerestoration.

3. Load should be restored in very small blocks.Reconnecting an entire shedding-step load at once,even at normal system frequency, can cause anoverload. Not only may its size exceed spinningreserve, but also high currents resulting from coldload pickup can temporarily cause a severe overload.Reconnecting small blocks of load will cause onlysmall frequency dips, which can be handled by thegovernors.

More small blocks may be reconnected until mostor all of spinning reserve is active. At this point, no

Figure 17-10 Load-shedding schemes for industrial plants

with no local generation.

390 Chapter 17

further load should be added until additional generat-ing capacity is available. Restoring excessive load maycause the frequency to settle below-normal systemfrequency, making further reclosing of interconnec-tions impossible.

4. If a significant loss of generation occurs in aconcentrated area of the system, transmission lines intothat area may be heavily loaded just to supply essentialloads. In this case, the imbalance should not beincreased by restoring expendable loads.

If frequency relays are used for automatic restora-tion, as they sometimes are at unattended installations,they should have a frequency setting of the normalsystem frequency. The load should be restored inblocks of 1 to 2% of system load, and restorationshould be sequenced by time delay. After the initialsystem recovery to normal system frequency, thereshould be a delay of 30 sec to several minutes,implemented automatically with a timer or manuallyvia supervisory control. This delay allows for resyn-chronizing of islands, reclosing of interconnections,and starting of peaking generators when available. Thefirst block of load may then be restored; the frequencywill dip and return to the normal system frequency.The next block should also incorporate several secondsof delay to permit frequency stabilization.

Each successive block should use a slightly longertime delay than the previous one. Thus, the second-block relays will time out before the third and reclosenext. The frequency will reestablish at the normalsystem frequency, and the third block will time out andreclose. This process will continue until all blocks arerestored or the spinning reserve is exhausted.

When restoring ‘‘cold’’ loads, it may be necessary totemporarily disable the instantaneous-overcurrentfault protection to prevent the initial current surgefrom retripping the feeder.

Restoring load for the example used in Sections 5.3and 5.4 can be described as follows. Assume that 10%of generation is lost, causing an 11% overload. Thefirst-step relays shed a block equal to 12% of systemload. The block shed consists of groups of distributionfeeders located at six different unattended substations,each equipped with an underfrequency relay. A secondset of six frequency relays (or the same set of relaysequipped with a ‘‘restore’’ function) used in theoverfrequency mode automatically restores service.All relays are set at the normal system frequency andreclose feeders, one substation at a time, using externaltimers for sequencing and delay. The initial delay is45 sec after the frequency returns to normal; subse-quent delays are as follows:

Substation no. External delay (sec)

1 10

2 12

3 14

4 16

5 18

6 20

Figure 17-11 shows the behavior of frequency overtime for this shedding and restoration action.

8 OTHER FREQUENCY RELAY APPLICATIONS

1. Overfrequency relays are often applied togenerators. These relays protect against over-speed during startup or when the unit issuddenly separated from the system with littleor no load. Relay contacts either sound analarm or remove power input to the turbine.

2. Underfrequency relays, with long external timedelays, may also be connected to generatingunits to protect against turbine-blade damageresulting from prolonged full-load underspeedoperation. If the overload exceeds the capabilityof the load-shedding scheme, these generatorrelays will isolate the unit (with some load, ifpossible) to keep it in operation and avoid bladefatigue. When load is small or absent andvibration is minimal, a supervisory overcurrentrelay or manual switch can be used to preventtripping during startup.

3. Underfrequency relays can also be used to sensedisturbances and intentionally split systems byopening ties. System splitting will not alleviateoverload, but it may allow more orderly anddiscriminating load-shedding action followingthe split. Neighboring utilities should agree touniform load-shedding programs; otherwise,the utility that sheds the most load may finditself relieving overloads for disturbances onother systems. In such cases, the shedding utilitycould use underfrequency relays to trip theinterconnections (possibly in conjunction withreverse-power relays). This scheme would elim-inate the possibility of a utility aggravating itsown loading problems by adding those of aneighboring utility without adequate load-shed-ding.


4. Another application of underfrequency relayscauses the load-tap changer on a transformer tostep in the direction to lower the load voltage.With a predominately resistive load, this willreduce the load approximately as the square ofthe voltage. A load of 20MW, for example, willbe reduced, theoretically, to 16.2MW if thevoltage is reduced by 10%. By integrating thisprocedure with other load-shedding strategies,the seriousness of reduced-frequency operationcan be reduced. Reduction of voltage by thisexpedient on induction motor loads may havelittle effect because an induction motor willadjust its speed to a lower level and maintainessentially a constant kVA. In fact, the powerconsumed by a motor may actually increase inresponse to a small reduction in voltage.However, in the case we are considering here,this will be compensated to a degree by adecrease in power resulting from a decrease infrequency. The results of the combination offrequency and voltage reduction are not nearly

so predictable as for a predominately resistiveload.

5. Another application in which an underfre-quency relay finds use is for the supervision ofa generator’s offline status. Cross-compoundturbines have a high-pressure unit and a low-pressure unit, each of which has its owngenerator. The two generators are synchronizedto each other at reduced speed and are broughtup to rated speed together. They may operate atreduced speed for a substantial amount of time.An underfrequency relay is often used to sensethat the generators are not yet connected to thepower system. Any stator overcurrent duringthis period of reduced-frequency operation isindicative of the fact that the generators havebeen inadvertently connected to the powersystem, either through a generator breaker orthrough the station service bus. Immediatetripping is mandatory if damage is to beminimized or possibly avoided altogether as aresult of this incorrect action.

Figure 17-11 Behavior of frequency during automatic load shedding and restoration. (Example for a 60-Hz system.)

392 Chapter 17

6. Protection can be inserted by a frequency relay atthe appropriate level. Some relays are responsiveonly at rated frequency and are frustrated intheir assignment if prolonged operation ispossible at odd frequencies. Some generatorground relays fall into this category. Additionalinstantaneous relays are applied to provideoffline fault protection. They should be removedfrom service once the machines are synchronized

to the system. An overfrequency relay (set forperhaps 55Hz) may be used to do this. Anotherproblem exists for cross-compound machinesthat are not yet synchronized to the powersystem. Reactive interchange between unitsoperating at reduced frequency can undesirablyproduce operation of a loss-of-field relay. Toavoid this, an overfrequency relay should be usedto activate this protection at rated frequency.


Bibliography

BASIC FUNDAMENTALS

1. Anderson, P. M. Analysis of Faulted Power Systems. The

Iowa State University Press, 1981.

2. Bewley, L. V. Traveling Waves on Transmission Systems.

Dover Publications, New York, 1963.

3. Blackburn, J. L. Protective Relaying Principles and

Applications. Marcel Dekker, New York, 1987.

4. Blackburn, J. L. Symmetrical Components for Power

Systems Engineering. Marcel Dekker, New York, 1993.

5. Greenwood, A. Electrical Transients in Power Systems.

John Wiley & Sons, New York, 1971.

6. IEEE/PES Power System Relaying Committee. ‘‘IEEE

Standard Relays and Relay Systems Associated with

Electric Power Apparatus,’’ ANSI/IEEE C37.90-1989.

7. Peterson, H. A. Transients in Power Systems. Dover

Publications, New York, 1966.

8. Wagner, C. F., and Evans, R. D. Symmetrical Compo-

nents. Robert E. Krieger Publishing, Malabar, FL, 1982.

INSTRUMENT TRANSFORMERS

1. IEEE/PES Power System Relaying Committee. ‘‘Transi-

ent Response of Current Transformers,’’ IEEE New

York, 1976, 76-CH, 1130-4 PWR.

2. IEEE/PES Power System Relaying Committee. ‘‘Guide

for the Grounding of Instrument Transformer Secondary

Circuits and Cases,’’ ANSI/IEEE C57.13.3-1983.

3. IEEE/PES Power System Relaying Committee. ‘‘Potential

Transformer Application on Unit Connected Genera-

tors,’’ IEEE Transactions on Power Apparatus and

Systems, Vol. PAS-91, Jan/Feb, 1972 pp. 24–28.


Guide for the Application of Current Transformers Used

for Protective Relaying Purposes,’’ Standard No.:

C37.110-1996.

5. Wright, A. Current Transformers, Their Transient and

Steady State Performance. Chapman and Hall, 1968.

MICROPROCESSOR RELAYING

1. Phadke, A. G., and Thorp, J. S. Computer Relaying for

Power Systems. John Wiley & Sons, New York, 1988.

2. Sachdev, M. S. (Coordinator). ‘‘IEEE PSRC Tutorial,

Advancements in Microprocessor Based Protection and

Communication,’’ 97TP120-0.

3. Morrison, I. F. ‘‘Prospects of On-Line Computer Control

in Transmission Systems and Substations,’’ IE Australia,

Vol. EE3, No. 2, Sept. 1967, pp. 234–236.

4. Mann, B. J. ‘‘Real Time Computer Calculation of the

Impedance of a Faulted Single-Phase Line,’’ IE Australia,

Vol. EE4, March 1969, pp. 26–28.

5. Rockefeller, G. D. ‘‘Fault Protection with a Digital

Computer,’’ IEEE Transactions, Vol. PAS-88, No. 4,

April 1969, pp. 438–464.

6. Rockefeller, G. D., and Udren, E. A. ‘‘High-Speed

Distance Relaying Using a Digital Computer, Part II:

Test Results,’’ IEEE Transactions, Vol. 91, No. 3, May/

June, 1972, pp. 1244–1258.

7. Girgis, A. A., and Brown, R. G., ‘‘Application of Kalman

Filtering in Computer Relaying,’’ IEEE Transactions on

Power Apparatus and Systems, Vol. 100, No. 7, July 1981,

pp. 3387–3397.

8. Schweitzer, E. O. ‘‘Unified Protection, Monitoring and

Control of Overhead Transmission Lines Achieves Per-

formance and Economy,’’ 35th Relay Conference, Geor-

gia Tech, Atlanta, 1985.

395

GROUNDING SYSTEM

1. Electrical Transmission and Distribution Book. Westing-

house Electric Corp., Trafford, Pennsylvania, 1964.

2. IEEE. ‘‘Recommended Practice for Grounding of

Industrial and Commercial Power Systems,’’ IEEE Std.

142-1982

GENERATOR PROTECTION


Guide for Abnormal Frequency Protection for Power

Generating Plants,’’ ANSI/IEEE C37.106-1987.


Guide for AC Generator Protection,’’ ANSI/IEEE C37,

1995.


for Generator ground Protection,’’ ANSI/IEEE

C37.101-1985.

4. IEEE/PES Power System Relaying Committee. ‘‘Inad-

vertent Energizing Protection of Synchronous Genera-

tors,’’ IEEE Transactions on Power Delivery, April 1989,

p. 965.

MOTOR PROTECTION

1. ‘‘AC Motor Protection Guide, Industrial and Commer-

cial Power System Application Series,’’ PRSC-2E Feb.

1992, ABB Allentown.

2. Elmore, W. A., and Kramer, C. A. ‘‘Complete Motor

Protection by Microprocessor Relay,’’ RPL 87-1, May

1987, ABB Coral Springs, FL.


Guide for AC Motor Protection,’’ ANSI/IEEE C37.96-

2000.

4. ‘‘Some Thoughts on Large Motor Protection,’’ RP1 76-

2A Sept. 1978, ABB Coral Springs, FL.

5. Zocholl, S. E. ‘‘Motor Analysis and Thermal Protec-

tion,’’ IEEE/PES Winter Meeting, Feb. 1990, IEEE

Paper 90 WM 247-7 PWRD.

TRANSFORMER PROTECTION

1. Elmore, W. A. ‘‘Ways to Assure Improper Operation of

Transformer Differential Relays,’’ Texas A&M Relay

Conference, College Station, TX, April 15, 1991.


Guide for Protective Relay Applications to Power

Transformers,’’ ANSI/IEEE C37.91-2000.

3. Sonnemann, W. K., Wagner, C. L., and Rockefeller, G.

D. ‘‘Magnetizing Inrush Phenomenon in Transformer

Banks,’’ AIEE Transactions Part III, Power Apparatus

and Systems, pp. 884–892.

4. ‘‘Transformer Protection Guide, Industrial and Com-

mercial Power System Applications Series,’’ PRSC-3E,

June 1991, ABB Coral Springs, FL.

BUS PROTECTION

1. ‘‘Bus Protection Guide Industrial and Commercial

Power System Applications Series,’’ PRSC-9 Nov.

1989, ABB Coral Springs, FL.

2. Forford, T., and Linders, J. R., ‘‘A Half-Cycle Bus

Differential Relay and Its Application,’’ IEEE Transac-

tions on PAS, Vol. PAS-93, July/Aug. 1974, pp. 1110–

1120.


Guide for Protective Relay Application to Power System

Buses,’’ ANSI/IEEE C37.97-1979.

TRANSMISSION LINE PROTECTION

1. Elmore, W. A. Pilot Protective Relaying. Marcel

Dekker, 2000.

2. Andersson, Finn, and Elmore, W. A., ‘‘Overview of

Series-Compensated Line Protection Philosophies,’’

Western Protective Relay Conference 1990, Washing-

ton State University, Pullman, WA.

3. Cook, V. Analysis of Distance Protection. John Wiley

& Sons, New York, 1985.

4. Crockett, J. M., and Elmore, W. A., ‘‘Performance of

Phase to Phase Distance Units,’’ Western Protective

Relaying Conference 1986, Washington State Univer-

sity, Pullman, WA.

5. Elmore, W. A. ‘‘Modern Transmission Line Relaying

Variations,’’ Georgia Tech Relaying Conference,

Atlanta, May 3–5, 1989.

6. Elmore, W. A. ‘‘Zero Sequence Mutual Effects on

Ground Distance Relays and Fault Locations,’’ Texas

A&M Relay Conference, College Station, TX, April

1992.

7. IEEE/PES Power System Relaying Committee. ‘‘Pro-

tection Aspects of Multi-Terminal Lines,’’ Special

Publication 79TH0056-2-PWR.

8. Lewis, W. A., and Tippett, L. S., ‘‘Fundamental Basis

for Distance Relaying on a 3-Phase System,’’ AIEE

Transaction, Vol. PAS-66, 1947.

9. Martilla, R. J. ‘‘Directional Characteristics of Distance

Relay Mho Elements Part I,’’ IEEE Transactions on

PAS, Vol. PAS-100, Jan. 1981, pp. 96–102.

10. Sonnemann, W. K., and Lensner, H. W. ‘‘Compensator

Distance Relaying I—General Principles of Operation,’’

AIEE Transactions on PAS, Vol. 77, Part III, pp. 372–

382.

396 Bibliography

11. Sun, S. C., and Ray, R. E., ‘‘A Current Differential

Relay System Using Fiber Optics Communications,’’

IEEE Transactions on Power Apparatus and Systems,

Vol. PAS-102, No. 2, Feb. 1983, pp. 410–419.

12. Udren, E. A., and Cease, T. W. ‘‘Transmission Line

Protection with Magneto-Optic Current Transducers and

Microprocessor-Based Relays,’’ Texas A&M Relay

Conference, College Station, TX, April 13–15, 1992.

13. Udren, E. A., and Li, H. J. ‘‘Transmission Line Relaying

Conference,’’ 1987, Washington State University, Pull-

man, WA.

14. Crockett, J. M, and Elmore, W. A. ‘‘Performance of

Phase-to-Phase Distance Units,’’ Conference for Protec-

tive Relay Engineers, Texas A&M, College Station, TX,

April 13–15, 1987.

15. Gilchrist, G. B., Rockefeller, G. D., and Udren, E. A.

‘‘High Speed Distance Relaying Using a Digital Com-

puter,’’ IEEE Transactions, Vol. PAS 91, May/June

1972, pp. 1235–1258.

16. Calero F. ‘‘Development of a Numerical Comparator for

Protective Relaying: Part I,’’ IEEE Transactions on

Power Delivery, Vol. 11, No. 3, July 1996, pp. 1266–

1273.

17. Elmore, W. A., Calero, F., and Yang, L., ‘‘Evolution of

Distance Relaying Principles,’’ Conference for Protective

Relay Engineers, Texas A&M, College Station, TX,

April 3–5, 1995.

18. Wang, L., and Price, E. ‘‘High Speed Microprocessor

Distance Relaying for Transmission Lines,’’ Western

Protective Relaying Conference, Spoken, WA, October

20–22, 1998.

19. Elmore, W. A., and Price, E. ‘‘Ground Relaying

Fundamentals,’’ 53rd Texas A&M Conference for

Protective Relay Engineers, College Station, TX, April

11–13, 2000.


for Protective Relay Applications to Transmission

Lines,’’ IEEE Std. C37.113-1999.

21. Novosel, D., Phadke, A., Saha, M. M., and Lindahl, S.

‘‘Problems and Solutions for Microprocessor Protection

of Series Compensated Lines,’’ Sixth International

Conference on Developments in Power System Protec-

tion, The University of Nottingham, UK, March 25–27,

1997.

STABILITY AND OUT-OF-STEP RELAYING

1. ‘‘Fundamentals of Out-of-Step Relaying,’’ RPL 79-1C,

Nov. 1991, ABB Coral Springs, FL.

2. IEEE/PES Power System Relaying Committee. ‘‘Out-of-

Step Relaying for Generators—Working Group Report,’’

IEEE Transactions on Power Apparatus and Systems,

Vol. PAS-96, No. 5, Sept./Oct. 1977, pp. 1556–1564.

3. Taylor, C. W., et al. ‘‘A New Out-of-Step Relay with Rate

of Change of Apparent Resistant Augmentation,’’ IEEE

Transactions on Power Apparatus and Systems, Vol.

PAS-102, No. 3, March 1983 pp. 631–639.

4. Haner, J. M., Laughlin, T. D., and Taylor, C. W.,

‘‘Experience with the R-Dot Out-of-Step Relay,’’ IEEE

Transactions on Power Systems, Vol. 1, No. 2, April 1986

pp. 35–39.

5. Ohura, Y., et al. ‘‘A Predictive Out-of-Step Protection

System Based on Observation of the Phase Difference

Between Substations,’’ IEEE Transactions on Power

Delivery, Vol. 5, No. 4, November 1990 pp. 1695–

1704.

6. Morioka, Y., et al. ‘‘System Separation Equipment to

Minimize Power System Instability Using Generator’s

Angular-Velocity Measurements,’’ IEEE Transactions on

Power Delivery, Vol. 8, No. 3, July 1993 pp. 941–947.

7. Roemish, W. R., and Wall, E. T. ‘‘A New Synchronous

Generator Out-of-Step Relay Scheme, Part I: Abbreviated

Version,’’ IEEE Transactions on Power Apparatus and

Systems, Vol. PAS-104, No. 3, March 1985 pp. 563–

571.

VOLTAGE STABILITY

1. IEEE Special Publication 90TH0358-2-PWR, ‘‘Voltage

Stability of Power Systems: Concepts, Analytical Tools,

and Industry Experience,’’ Prepared by the IEEE Work-

ing Group on Voltage Stability, 1990.

2. IEEE/PES Power System Relaying Committee, ‘‘System

Protection and Voltage Stability,’’ 93THO 596-7 PWR,

June 1993.

3. Taylor, C. W., ‘‘Power System Voltage Stability,’’

McGraw-Hill, New York, 1994.

4. IEEE Special Publication 93TH0620-5-PWR, ‘‘Suggested

Techniques for Voltage Stability Analysis,’’ Prepared by

the IEEE Working Group on Voltage Stability, 1993.

5. IEEE/PES Power System Relaying Committee, ‘‘Voltage

Collapse Mitigation,’’ 24th Annual Western Protective

Relay Conference, Spokane, WA, October 21–23, 1997.

6. Wang, L., and Girgis, A. A., ‘‘On-Line Detection of

Power System Small Disturbance Voltage Instability,’’

IEEE Trans. on Power Systems, Aug. 1996, pp. 1304–

1313.

7. Vu, K., Begovic, M. M., Novosel, D., and Saha, M. M.,

‘‘Use of Local Measurements to Estimate Voltage-

Stability Margin,’’ IEEE Trans. on Power Systems, Aug.

1999.

8. Shuh, H. M., and Cowan, J. R. ‘‘Undervoltage Load

Shedding an Ultimate Application for Voltage Collapse,’’

Georgia Tech Protective Relaying Conference, April 29–

May 1, 1992.

9. Karlsson, D. ‘‘System Protection Schemes in Power

Network Based on New Principles,’’ 27th Annual Western

Protective Relay Conference, Spokane, WA, October 21–

23, 2000.

Bibliography 397

RECLOSING

1. IEEE/PES T&D Committee Working Group Report.

‘‘Arc Deionization Times on High Speed Three Pole

Reclosing,’’ IEEE Transactions on Power Apparatus and

Systems, Vol. S82, 1963, pp. 236–252.

2. Sturton, A. B. ‘‘One, Two and Three-Phase Automatic

Reclosing of 230 kV and 345 kV Lines,’’ IEEE Transac-

tions on Power Apparatus and Systems, Vol. S82, 1963,

pp. 304–318.

LOAD-SHEDDING AND FREQUENCYRELAYING

1. Dalziel, C. F., and Steinback, E. W. ‘‘Underfrequency

Protection of Power Systems for System Relief,’’ AIEE

Transactions on Power Apparatus and Systems, Part III-

B, Vol. 78, pp. 1227–1238.

2. Lokay, H. E., and Burtynk, V. ‘‘Application of Under-

frequency Relays for Automatic Load Shedding,’’ IEEE

Transactions on Power Apparatus and Systems, Vol.

PAS-87, No. 3, March 1968, pp. 776–783.

3. Smaha, D. W., Roland, C. R., and Pope, J. W. ‘‘Load

Shedding Coordination With Turbine-Generator Under-

frequency Protection on the Southern Electric System,’’

Georgia Tech Relaying Conference, Atlanta, 1980.

398 Bibliography

Index

ABB relays, 281

Active filters, 67–68

A/D converters (analog/digital), 95

Adders, 65, 69

Air-gap current transformers, 91, 167–

168

Aliasing, 97–98

overcoming, 98–99

Amplification units, 59

Amplifiers:

inverted, 66–67

non-inverting, 65, 67

See also Operational amplifiers

Analog/digital converters, 95

Analog relays, 2

Analog test, 103

AND logic, 61

AND unit, 52–53

Angle, of fault current, 20

Angle instability:

large disturbance, 355–356

Q angle, 358–360

small disturbance, 353–354

Angular synchronism check, 378–379

Annunciator circuits, 59–60

ANSI standard:

for relay accuracy, 77–79

for transformers, 85–87, 173, 186

for turbine generators, 123

Antialiasing filters, 98

Apparent impedance, 305–306

Application, 4–7

degree of protection, 5

existing system protection, 5

fault study, 5

impedances, 6

[Application]

maximum load, 6

rack mounting, 7–8

switchboard relays, 6–7

system configuration, 5

transformer data, 6

Arc/fault resistance, 272

Arcing grounds, 108–109

Arc resistance, 20–21, 265

Armature-winding resistance, 24

ARS out-of-step tripping, 349–350

Automatic reclosing, 196, 370

Automatic restoration, 391

Automatic synchronizer, 379–380

Autotransformer banks, 24–26

Autotransformers, 240–241

Auxiliary relays, 1, 331

multitrip and lockout, 331

Auxiliary switches, 327

Auxiliary units, 59–63

annunciator circuits, 59–60

coordinating and loop logic timers,

60–61

isolator and buffer circuits, 62–63

toggle or latching circuits, 61

Back-to-back capacitor switching,

73

Backup protection, 323–352

breaker-and-a-half and ring buses,

328–329, 336

breaker-failure relaying applications,

327–336

for generators, 123–125

improved breaker-failure scheme,

333–336

[Backup protection]

open conductor and breaker pole

disagreement, 336–337

remote vs. local, 323–327

single-line/single-breaker buses, 327–

328

single-pole trip-system, 337

traditional scheme, 329–332

problems with, 332–333

type SBF-1 breaker-failure relay,

334–336

Batteries, for tripping, 7

Battery circuit, grounding of, 75

BC phase-to-phase fault, 298

Blinders, 293–294, 348

double, 351

with lens, 350–351

in line protection, 249

Block-block comparison logic circuit,

55–59

BL-1 relays, 153, 154–155

Breaker-and-a-half, 326, 328–329, 336

reclosing, 376–377

Breaker-failure relaying, 327–336

breaker-and-a-half and ring buses,

328–329, 336

improved breaker-failure scheme,

333–336

recommendations, 327

single-line/single-breaker buses, 327–

328

for single-pole trip-system, 337

traditional scheme, 329–332


Breaker-failure relays, 325–326, 334–

336

399

Breaker poles, 336–337

Breakers, current-limiting, 280

Breaker-trip coil initiator, 59

Bridge-type relays, 154

Brushless machines, 136

Brush-type machines, 136

Buffer and isolator circuits

Buffer circuits, 62–63

Buffers, 78

Bus deenergization, 73–74

Bus differential relaying:

dc saturation effect, 89

directional comparison, 227

fault bus, 227

high-impedance, 219–222

with overcurrent relays, 216–217

partial, 226–227

Buses:

double, single-breaker, 216, 223

main-and-transfer, 215

multiple, 359–360

single-line/single-breaker, 327–328

single-machine infinite, 353–356

See also Ring buses

Bushing voltage transformers, 92, 93

Bus protection, 213–227

buses with transformer bank, 223–

224

current transformer saturation

problem, 213–215

differential comparator relays, 222–

223

differential relaying

with overcurrent relays, 216–217

double-bus single breaker, 216, 223

with bus tie, 224–226

high-impedance differential system,

219–222

information requirements, 215

KAB bus protection example, 222

multibus systems, 359–360

multirestraint differential system,

217–219

normal practices, 215–216

other schemes

directional comparison relaying,

227

fault bus (ground-fault protection

only), 227

partial differential relaying, 226–

227

against overloads, 226–227

Bus transfer systems, 139–143

fast transfer, 139–142

slow transfer, 142–143

BYZ current transformer, 148–149

Capacitors:

in line protection, 273

for suppression at source, 77

Capacitor switching, 73

CA relays:

generator protection, 166–167

shunt reactor application, 206, 207

transformer protection, 166–167,

193

CA, 175

CA-16, 206, 215, 217–219, 224,

226

CA-26, 167, 171, 224, 226

checks, 175, 178–179

C class, 86

CCVT. See Coupling capacitance

voltage transformers

Check-sum test, 103

Circuit-breaker auxiliary switches,

327

Circuit-breaker control, 8–9

Circuit breakers, 204, 344

See also Breaker-failure relaying

Circuit diagrams, voltage in, 12

Circuits:

battery, grounding of, 75

induction motor equivalent, 146

integrated, 63–70

logic, 54–63

open, 18

sample-and-hold (S/H), 95–97

secondary, 71

typical, 7

Clapper units, 44

Classification:

of electric power lines, 229

of performance, 4

of relays, 1–3

Closed transfer, 142

Close fail time, 368

CM relay, 156–157

Coincident time comparison, 55

Cold load, 236, 238

restoring, 390–391

Commercial and industrial power

transformers, 193–197

Common-mode surges, 72

Comparators, 281–285

distance relaying applications, 285–

294

multiple, 305

Compensation factor, 180–182

Complexity, 3, 4

Composite sequence current networks,

26–32, 48–49

Concentric circle scheme, 348

Coordinating and loop logic timers,

60–61

Coordinating time interval, 233

COQ relays, 157

CO-2 relay, 147–148

Cost effectiveness, 3–4

Coupling capacitance voltage

transformers (CCVT), 91, 92–93

short-line application, 271–272

switching, 74

CP relays, 155

Cross-blocking, 171

Cross-compound generators, 136–138

Current(s):

and buses, 213

distribution factors, 302–305

false residual, 148

and load flow, 302–305

magnetizing inrush, 163–166

maximum motor starting, 147

mismatched, 172

in multiple loop system, 308

natural laws, 101

notation for, 11

parallel-line, 265–267

phase-fault, in motors, 147

pickup, 119, 169

reverse flow, 27, 274

in two transformers, 118

Current polarization, 240–243

Current transformer ratio, 81

Current transformers, 81–88

direct current saturation, 82,

88–89

equivalent circuit, 82

and line protection, 233

performance estimation, 82–91

ANSI classes, 85–87

European classification, 87–88

excitation curve method, 83–85

formula method, 83

manufacturer data, 85–87

MOCT, 91

residual flux, 89–91

saturation, 75, 81–82

for short-line application, 271

CVD electromechanical relay, 107–108

CVQ relays, 155–156, 160

CV-8 relay, 120, 208

CV relays, 155

CVX relays, 377–378

CWC relays, 196, 207

CWP-1 directional ground relay, 149

CW watt-type relay, 158–159

Cylinder relay units, 47, 330

directional, 16

400 Index

D’Arsonval units, 47

See also DT-3 relay

Dc offset compensation, 101

Dead-line or dead-bus reclosing relay,

379

Dead-man timer, 103

Dead tank breakers, 4

Deionizing times, 366

Delta systems, 207–209

Dependability, 3, 4

Design, of protective relay system, 2–4

design criteria, 3–4

design factors, 2

economics vs. performance, 3–4

performance factors, 4

zones of protection, 3, 4

Device 46, 160, 161

Device 47/27, 159, 161

Device 49, 161

Device 49/50, 159

Device 50, 161

Device 50/51, 159, 193, 206

Device 51, 161

Device 67, 196

Device 87, 118–119, 161

Device 87E, 202

Device 50G, 159, 161, 195

Device 51G, 196

Device 51N, 231

Device 51N1, 202

Device 51N2, 202

Device 59N, 208

Device 67N, 231

Device 87N, 208

Device 87N3, 121–122

Device 50N/51N, 208, 209, 231

Device 51N/50N, 159, 161, 195

Device 67N/50N, 231

Device 87S, 199–200

Device 151G, 196

Diagrams:

circuits, 12

notation, 9, 11–15

phasors, 11, 12–15

Diesel engines, 131

Differential-mode surges, 72

Differential relaying:

for bus protection, 213

high impedance, 219–222

multirestraint scheme, 217–219

overcurrent, 216–217

partial, 226–227

for transformer protection, 166–173

differential relays, 166–171

guidelines, 171–173

sample checks, 173–182

[Differential relaying]

for multiwinding banks

for two-winding banks, 173–178

Differential relays:

comparator, 222–223

for generator protection, 118–119

for shunt reactor applications, 206–

207


Digital filters, 100, 101, 102

Digital relays, 2

frequency relays, 385

Diodes, 50

Direct current circuit energization, 75

Direct current coil interruption, 74–75

Direct current (dc) offset, 101

Direct current offset, 101

Direct current saturation, 82, 88–89,

215

Directional comparison schemes:

in bus protection, 227

and system stability, 344

Directional ground-relay polarization,

239–243

Directional overcurrent relays, 267–

268, 269–270

inverse-time, example of, 308–321

Directional relays:

instantaneous-trip, 309

overcurrent, 267–268, 269–270

inverse-time, 308–321

polarity, 15–16

and three-phase power systems,

17–18

types, 16

Directional sensing, 243–244

Distance relaying:

comparator applications, 285–294

for line protection, 247–250

ground-distance, 254–257

in- and outfeed effects, 260–262

length factor, 235, 257–259

equation, 281

phase-distance, 250–254

and series compensation, 275

for short-line, 270–272

and transformer banks, 261–265

zone application, 257–259

for motors, 150–151

for transformers, 192

Distance relays:

and current transformers, 271

in generator protection, 124–125

inverse-time, 268, 270

in line and circuit protection, 247–

250

[Distance relays]

in loop systems, 268–169

mho-type, 250, 270

in motor protection, 150

and out-of-step conditions, 344

Distribution factors, 35–36, 39

and impedance units, 302–305

Distribution feeder protection, 276–280

Distribution lines, 229

Double-bus single breaker, 216, 223

with bus tie, 224–225

Double phase-to-ground faults, 301–

302

Doughnut current transformer, 148

DPU2000R feeder protection, 365,

373–374

DT-3 relay, 154

Dual polarization, 243, 247

Earth, 105

Economics vs. performance, 3–4

EEPROM (electrically erasable

programmable read-only

memory), 95–96

Electrical separation, 76

Electromagnetic induction, 72

avoidance, 78

Electromagnetic transient programs

(EMTP), 3

Electromechanical out-of-step

relaying, 348–349

Electromechanical relays:

for line protection, 244, 256–257

for out-of-step systems, 348–349

for phase-unbalance protection,

156–157

for reclosing, 371–372

at reduced frequencies, 138(table)

for switchgear applications, 155


Electromechanical units, 43–47

D’Arsonval units, 47

ground faults, in ungrounded

systems, 47, 107–108

magnetic attraction units, 43–45

magnetic induction units, 45–47

thermal units, 47

Electrostatic induction, 71–72

Energization, inadvertent, of

generators, 132–134

EOVT (voltage-sensing device), 93

EPROM (erasable programmable

read-only memory), 95

‘‘Essentially in phase,’’ 15

Excitation, loss of:

in generators, 127–129, 133

Index 401

[Excitation, loss of]

in motors, 158–159

Existing protection, 5

Failures, causes of, 4, 18–20

Faraday effect, 91, 93

Fast transfer, 139–142

Fault angles, 20

Fault current:

angle of, 20

calculation, 31–37

maximum dc component, 101

sustained, 24

Fault-current detectors, 330


overcoming, 333–336

Fault detectors, 102–103

Fault resistance, 305–306

Faults, 18–21

balanced, 124–125

calculation of, in loop system, 32–37

characteristics, 20–21

large-magnitude, close-in, 81–82

phase-to-ground, 39, 298, 300–301,

305

phase-to-phase, 298, 300–301, 305

simultaneous, 18–20

three-phase, 298, 299–300, 345

turn-to-turn, 209–211

types and causes, 18–20

unbalanced, 123–124

Fault-sensing data processing units

magnitude comparison, 54–55

phase-angle comparison, 55–59

Fault study, 5

Fault voltage, calculation of, 31–32

Feeder protection relays, 365, 373–374

Feeders:

instantaneous reclosing, 369

and load-shedding, 390

with no-fault-power back-feed, 369

Ferroresonance, 107–108

Fiberoptics, 93

Field, loss of, 127–129, 133

Field ground detection, 134–136

brushless machine, 136

brush-type machine, 135–136

injection scheme, 136

50D relay, 148

Filters:

antialiasing, 98

digital, 100, 101

Fourier-notch, 100–101

Fixed-reference circuit, 54–55

Fixed shunt reactors, 203

Flip-flops. See Toggle circuits

Flux:

and frequency, 184

notation of, 11

saturation, 81–82, 213–215

and voltage, 184

Follow-breaker function, 368

46Q relay, 124

Forward, as concept, 306

Fourier-notch filter, 100–101

Four-wire systems:

multigrounded, 115

unigrounded, 114–115

Frequency:

decline rate, 381–383

and flux, 184

and overloads, 381–384, 387–388,

390–392

rating deviations, 138–139

reduced, 136–138

and turbine blades, 138–139

Frequency relays, 384–385

for automatic restoration, 391

digital frequency relays, 385

and generators, 391

KF induction-cylinder

underfrequency relay, 384

in load-shedding, 388–389

microprocessor-based, 385

non-load-shedding uses, 391–393

Fused breaker, 280

Fuses, coordination with, 236–237,

277–280

Gas turbines, 131

Generator motoring, 130–134

Generator protection, 117–143

ac overvoltage, 136

backup protection, 123–125

bus transfer systems, 139–143

field ground detection, 134–136

and frequency relays, 391, 392–393

against generator motoring, 130–134

for hydroelectric generators, 136

inadvertent energization, 132–134

loss-of-excitation protection, 127–

129

machine connections, 119

microprocessor-based, 7–8, 117, 143

off-frequency operation, 138–139

out-of-step protection, 139, 345–346

overload protection, 126

overspeed protection, 126–127

phase fault detection, 117–119

recommendation, 139

at reduced frequencies, 136–138

for station auxiliary loads, 139–143

[Generator protection]

stator ground fault protection, 120–

123

neutral third harmonic

undervoltage, 120

neutral-to-ground fault detection,

121–122

95% ground relays, 120–121

100% winding protection, 120,

122–123

unit-connected schemes, 120

technology trend, 117

in ungrounded systems, 105–108

volts per hertz protection, 126

Generators:

cross-compound, 136–138

hydroelectric, 136

off-line status, 392–393

with split-phase windings, 119–120

wye-connected windings, 119

Generator-transformer unit:

differential protection of, 183

overexcitation protection of, 184–

185

GIX-104, 122

GPU2000R relay, 137–138

Ground:

terminology, 14

virtual, 64

Ground directional relay applications,

239

Ground directional unit, 16

Ground-distance relays, 254–257

infeed effect, 306–308

and zero-sequence mutual

impedance, 265–267

Ground-fault protection:

for motor, 147–149

for power lines, 231 (table)

directional sensing, 243–244

high-voltage side, 241–242

low-voltage side, 242–243

in ungrounded systems, 105–108

Grounding:

of battery circuit, 75

classes of, 20

high resistance, 111–112, 149

low impedance, 110–112

low-resistance, 148

of shield, 77

in symmetrical system, 20

Ground overcurrent relays, 238

conventional transformers, 112–

113

zero sequence transformers, 114

Ground product relays, 113–114

402 Index

Ground-relay directional sensing, 243–

244

Ground relaying, sensitive, 112–114

Ground relays:

directional sensing, 243–244

negative sequence, 244

for line protection, 244–247

performance estimation, 85

setting, 233

Ground-Shield TM series, 148

Ground units, 289–293

quadrature-polarized, 290–291, 296–

297

and distribution factors, 303–304

and phase-to-ground faults, 299

and phase-to-phase faults, 300

and three phase faults, 299–300

reactance, 292–293

reverse characteristics, 295–297

self-polarized, 291–292, 297

and distribution factors, 304

and phase-to-ground faults, 300–

301

double, 301–302

and phase-to-phase faults, 300–

301

Ground wires, overhead, 265

Harmonic restraint unit (HRU), 169,

192–193

Harmonic voltage:

and generators, 120, 122

and transformers, 167–171, 180–182

Heat, generated within motor, 149

High-frequency distortion, 98

High impedance differential relays, 119

High impedance differential scheme,

219–222

High-pass filter, 66–67

High-reactance grounding, 108–109

High-resistance grounding, 111–112,

149

High-voltage problem, 221–222

Hi-Lo CO relays, 245

HU relays, 167–168, 171, 193

checks, 176–177

HU-1, 167–168, 171, 198–199

checks, 180

HU-4, 169, 171, 206, 224

modified, 169

Hydraulic turbines, 131–132

Hydroelectric generators, 136

Hysteresis, 67

Impedance:

apparent, 305–306

[Impedance]

of capacitance voltage transformers,

93

different, and motor ground faults,

148

and ground-distance relays, 265–267

leakage, 24

in locked-rotor situation, 150

and metal oxide, 273, 275

mutual, in power lines, 243

in shunt reactors, shorted turn, 211

source impedance ratio (SIR), 270,

272

in synchronous machinery, 24

three-winding banks, 24–26

in transformers, 24–26

in transmission lines, 26

zero sequence mutual, 265–267

Impedance unit characteristics, 281–306

apparent impedance, 305–306

comparators

basic concept, 281

distance relaying applications,

285–294

magnitude comparators, 282–283,

284, 285

phase comparators, 281–284

current distribution and load flow,

302–305

derived characteristics, 305

different fault types, 298–302

distance relaying applications, 285–

294

practical applications, distance

relaying


IMPRS relays, 153, 154–155, 157

Inadvertent energization, 132–134

Induction disc units, 45–47

Induction motor equivalent circuit, 146

Induction motor loads, 392

Industrial plants, 369

Industrial power systems:

industrial transformers, 193–197

load-shedding programs, 389–390

Infeed effect, 260–261, 306–308

Initial inrush, 163–164

Inputs, of relays, 1

Inrush currents, 163–166

cold load, 236, 238

monitoring, 171

Instantaneous overcurrent protection,

237–238

Instantaneous overcurrent unit, 68

Instantaneous reclosing, 368–369

Instantaneous-trip lockout, 367

Instantaneous-trip relays, 308–309, 325

Instrument transformers, 81–94

current transformers, 81–82

fiberoptic voltage sensing devices, 93

neutral inversion, 93–94

voltage transformers, 91–93

See also Current transformers

Integrated circuits, 63–70

basic operational amplifier units, 65–

68

operational amplifier, 63–64

relay applications, 68

Intermediate lockout, 367

Inverse relays, 268, 270, 280

example, 308–321

Inverse time, 234–235, 238, 268–269, 270

example, 308–321

Inverted amplifiers, 65

with capacitor, 66

Isolation, 75–76

Isolator circuits, 62–63

IT current unit, 222

Jacobean matrix, 358

Jam protection, 157

KAB relays:

in bus protection, 215, 219–222

in shunt reactor applications, 207

in transformer protection, 198

KC-4 relays, 330

KDAR relays, 250–252

KDTG unit, 257

KD-type relays, 125

KDXG relays, 256–257, 292–293, 306–

307

KF induction-cylinder underfrequency

relay, 384

Kirchoff’s law, 223

KLF curves, 128–129

KLF relays, 137, 158

Knee voltage, current transformers,

220

KRT unit, 257

KS-3 out-of-step blocking, 348–349

KST out-of-step tripping, 349

Large-disturbance (LD) voltage

instability, 355–356

Latching circuits, 61

LDAR relays, 306–307

LDG ground units, 289–290

LDG relays, 307

Leading phase identification, 102

Lead runs, 215

Leakage impedance, 24

Index 403

Lens scheme, 350–351

Level detectors, 67

Limiters, 280

Line and circuit protection, 229–321

directional overcurrent phase- and

ground-fault protection, 239–

247

distance phase and ground

protection, 247–267

distance relay basics, 247–250

fault resistance and ground-

distance relays, 265

ground-distance relays, 254–257

infeed and outfeed effects, 260–

261

length factors, 235, 257–259, 270–

272

equation, 281

phase-distance relays, 250–254

and transformer banks, 261–265

zero sequence mutual impedance

and ground-distance

relays, 265–267

distribution feeder protection, 276–

280

electric power line classes, 229, 257

impedance unit characteristics

infeed effect, on distance relaying,

260–261

infeed effect, on ground-distance

relays, 306–308

line protection techniques, 229

loop-system protection, 267–270

single-source, 267–269

multiple loop systems, 269–270, 308–

321

relay selection, 308

settings, 309–321

multiterminal, tapped lines, and

weak feed, 230–231

out-of-step relaying, 346

overcurrent phase-and ground-fault


power line classification, 229

relay selection, 244–247

relays for phase- and ground-fault

protection

series-capacitor issues, 273–276

short-lines, 270–272

Line-connected reactors, 203–204

Live-line/dead bus, live-bus/dead-line

(LLDB/LBDL), 367, 369

Live tank breakers, 4

Load flow, 302–305

and voltage instability, 358–359

Load loss protection, 157–158

Load-reduction factor, 386

Load-shedding:

formulation, 385–391

anticipated overload, 385–386

frequency settings, 387–388

location of frequency relays, 388–

389

relays, number of, 386

size of loadshed, at each step, 386–

387

steps, number of, 386

time delay, 388

and frequency relaying, 381–393 (see

also Frequency relays)

frequency decline rate, 381–383

in industrial systems, 389–340

load-shedding, 383–384

restoring service, 390–391

handling, 383–384

and system splitting, 391–392

and voltage stability, 362–363

Load tap changers, 362

Local backup:

and breaker failure, 324–326

vs. remote, 323–327

Local generation, 369

Locked-rotor, 149–152

Lockout, 367

Logic circuits, 54–63

amplification units, 59

auxiliary units, 59–63

fault-sensing data processing units,

54–59

and out-of-step tripping, 351

Logic units:

principal, 52–53

solid-state, 52

Loop-type system:

fault calculation example, 32–37

in power line protection, 267–270

multiple loops, 269–270, 308–

321

relay selection, 308

settings, 309–321

Loss-of-excitation protection

for generators, 127–129, 133

for synchronous motors, 158–159

Loss-of-field relays, 345–346

Low-pass filter, 66

Low-reactance grounding, 109–110

Low-resistance grounding, 110–111

Low-voltage breakers, 277–280

Low-voltage protection, 155

Magnetic attraction units, 43

Magnetic induction units, 45–47

Magnetizing inrush, 163–166

recovery inrush, 165

sympathetic inrush, 165–166

Magneto-optic current transducer

(MOCT), 91

Magnitude comparators, 282–283, 284,

285

Magnitude comparison logic units, 54–

55

Manual close, 368

Maximum load, 6

MCO relay, 234, 235

MDAR relays, 7, 252–254, 306–307

Measurement, phase directional, 18

Memory test, 103

Memory voltage, 299

Metal oxide, 273, 275

Mho-type relays, 250, 275, 291–292

Microprocessor architecture, 70

Microprocessor-based generator

protection, 7–8, 117, 143

Microprocessor-based relays:

for breaker-failure protection, 331–

332

for bus protection, 222–223

feeder-protection, 365, 373–374

frequency-sensitive, 385

for line protection, 236, 244

phase-distance relays, 252–254

MDAR, 7, 252–254, 306–307

for motor protection, 154–155

operation times, 235

rack mounting, 7–8

for reclosing, 372–377

reclosing and synchronism check,

365

for transformer protection, 180

Microprocessor relaying, 95–104

aliasing, 97–98

overcoming, 98–99

architecture, 70, 95–97

basic concepts, 95–97

choice of measurement principle,

99–103

dc offset compensation, 101

digital filters, 100, 101, 102

fault detectors, 102–103

Fourier-notch filter, 100–101

leading phase identification, 102

rms calculation, 100

symmetrical component filter, 102

and generators, 117

sampling, 95–97

nonsynchronous, 98–99

sampling problems, 97

self-testing, 103

404 Index

Microprocessors:

in generator protection, 7–8

and transformer protection, 171

Mismatched currents, 172

MMCO relay, 234

MMCO relays, 234–235

MOCT (magneto-optic current

transducer), 91

Monitoring relays, 1

Motoring, 130–134

Motor protection, 145–161

application combinations, 159–161

excitation loss, 158–159

and frequency relays, 392

general requirements, 145

ground-fault protection, 147–149

hazard types, 143–147

induction motor equivalent circuit,

146

jam protection, 157

KVA criteria, 147

load loss protection, 157–158

locked-rotor protection, 149–152

low-voltage protection, 155

motor thermal capability curves,

146–147

negative sequence current relays, 157

negative sequence voltage


out-of-step protection, 158

overload protection, 153

phase-fault protection, 147

phase-rotation protection, 155

phase-unbalance protection, 156–

157

thermal relays, 153–155

RTD-input type, 154

thermal replica, 154–155

Motor thermal capability curves, 146–

147

MOVT (voltage-sensing devices), 93

MPR relays, 153, 157

MRC-2 relays, 372–373

MSPC relays, 291–292

Multibus systems, 359–360

Multigrounded four-wire systems, 115

Multiplexors, 95

Multi-shot reclosing relays, 366, 371–

377

Multiterminal transmission lines, 230

Mutuals, 305–306

MVA margin, 360–361

National Electric Code (NEC), 153

Negative logic, 52

Negative sequence current relays, 157

Negative sequence directional units,

244

Negative sequence impedance, 24, 26

Negative sequence networks, 27, 35–

36, 50

Negative sequence polarization, 243,

245–247

Negative sequence voltage protection,

155

Neutral:

in symmetrical components, 23

vs. ground, 14–15

Neutral inversion, 93–94, 108

No-fault-power back-feed, 369

Noninverting amplifiers, 65, 67

Nonvolatile memory test, 103

Notation, 9, 11–15

for comparator applications, 287–

288

for logic units, 52

for operational amplifier, 63

for three-phase systems, 14

NOT unit, 53

NOVRAM (nonvolatile RAM), 95–96

testing, 103

Numerical relays, 2, 222–223

Nyquist criterion, 98

Off-frequency operation, of generators

One-shot reclosing relays, 366

Open circuits, 18

Open conductor, 336–337

Open transfer, 140

Operating principle, of relays, 2

Operational amplifiers, 63–64

relay application, 68–70

types of, 65–68

Operation times, 234–235

Optical isolators, 63, 78

OR unit, 53

Outages, extended, 236

Outfeed, 27, 261–262, 274

Out-of-step relaying, 345–352

blinders, 293–294, 348

concentric circle scheme, 347–348

generator out-of-step relaying, 139,

345–346

for motor protection, 158

philosophies of, 346–347

relays for, 348–351

strategy selection, 351–352

blocking vs. tripping, 352

for transmission lines, 346

Overcurrent directional relays, 267–

268, 269–270

inverse-time, example of, 308–321

Overcurrent phase- and ground-fault


Overcurrent protection:

in buses, 216–217

ground-fault, 238–239

instantaneous, 237–238

for shunt reactors, 205–206

in transformers, 185–192

Overcurrent relays, 150, 216, 344

Overcurrent units, instantaneous, 68

Overexcitation, of generator-

transformer unit, 184–185

Overfrequency relays, 138, 391, 393

Overlapping, 4

Overload protection:

for generators, 126

for motors, 153

Overspeed protection, 126–127

Parallel clamps, 77

Parallel-line current, 265–267

Percentage differential relays, 118–119

Performance, vs. economics, 3–4

Performance characteristics, of relays,

2

factors influencing, 4

Performance checks, of current

transformer, 177–178

Phase and ground relays, performance

estimation, 85

Phase-angle comparator logic circuitry,

55–59

Phase-angle regulators, protection of,

197–101

Phase comparators, 281–285, 299

Phase-comparison systems, 69–70, 344

Phase directional measurement, 18

Phase directional overcurrent

applications, 239

inverse, 308–321

Phase-distance relays, 250–252

Phase distortion, 21

Phase-fault protection:

for generators, 117–119

for motors, 147

for power lines, 230 (table)

Phase relays, 84

Phase rotation, 15, 155

Phase sequence, 6, 15

Phase-shift units, 67, 69

Phase-to-ground faults, 39, 298, 300–

301

apparent impedance, 305

double, 301–302

Phase-to-ground units, 276, 289–290

See also Ground units

Index 405

Phase-to-phase faults, 298, 305

Phase-to-phase units

directionality, 276


and magnitude comparator concept,

288



reverse characteristics, 294

and three-phase faults, 298

Phase-unbalance protection, 156–157

Phasing checks, for transformer banks,

173–178

Phasing voltage, 377–378

Phasors:

diagram notation, 11–15

in directional relays, 15–18

leading (time-coincident), 102

multiplication law, 13

notation, 12–13

phase rotation vs. phasor rotation, 15

sum of, and buses, 213

and symmetrical components, 22

Pickup current, 119, 169

Pilot-wire systems, 69–70, 344

Plunger units, 43

Pockel cell, 93

Polarity, 15–18

directional relays, 15–18

connections to three-phase

systems, 17–18

cylinder-type, 16

ground-type, 16

watt-type, 16

of protective relays, 15–16

Polarization:

dual, 243, 247

negative sequence, 243, 245–246

zero sequence, 243, 245–247

See also Quadrature-polarized

ground units; Self-polarized

ground units

Polar units, 45

Positive sequence impedance, 26

Positive sequence network, 26–27, 35

Potential polarization, 240

Potential transformers. See Voltage

transformers

Power, reactive, 362

Power factor, 20

Power lines:

classification, 229

cold load, 236, 391

length of, 235, 257–259, 270–272

equation, 281

protection selection, 229–230

[Power lines]

protection techniques, 229

series-capacitor compensated, 273

See also Line and circuit protection;

Transmission lines

Power systems:

faults, 18–21

loop type, 32–37

Problem, statement of, 4–5

PRO*STAR relay, 151–152, 153, 157,

160

Protective relays:

application, 4–6

definition, 1

in motor protection, 145

performance factors, 4

in systems, 2–4

Q angle, 358–360

Quadrature-polarized ground-distance

unit, 298–299

Quadrature-polarized ground units,

290–291

and distribution factors, 303–304




and three-phase faults, 299–300

Rack-mounted relays, 7–8

Radial systems, 39

RADSB relays, 169–171

checks, 175–176, 178–179

RADSS relays, 222

RAICB relay, 185

RAM (random access memory), 95

testing, 103

Rate-of-rise-pressure, 205, 209–211

Ratio checks, for transformer banks,

175–179

RC reclosing relay, 371–372

Reactance, for synchronous

machinery, 24

Reactance grounding, 108–110

low-, 109–110

Reactance ground units, 292–293

Reactive power control, 362

Reactor protection. See Shunt reactor

protection

Read-only memory (ROM), 95

check-sum test, 103

REB551 package, 365, 376

REB-103 relays, 222–223

Reclosers, 277

Reclosing, 334–335, 365–377

automatic, 196, 370

[Reclosing]

dead-line or dead-bus, 379

industrial plants, with local

generation, 369


of lines with sources at both ends,

369

precautions, 365–366

selective, 366

system considerations, 366–369

compatibility with supervisory

control, 367–368

deionizing times for three-pole

reclosing, 366

factors governing application of

reclosing, 368

inhibit control, 368

instantaneous-trip lockout, 367

intermediate lockout, 367

live-line/dead-bus, live-bus/dead-

line control (LBDL), 367,

369

one-shot vs. multiple-shot, 366

selective reclosing, 366

synchronism check, 366–367, 377–

378

voltage block function, 373–374

Reclosing relays:

definition, 1

multi-shot, 366, 371–377

operation, 369

single-shot, 366, 369–371

Recovery inrush, 165

RED-521 relays, 215, 222–223, 226

Redundancy, 143, 325

REG-100 relay, 124, 139, 143

REG-216 relay, 124, 139, 143

Regulating relays, 1

REL 521/531, 376

REL 551/561, 376

Relays:

auxiliary, 1, 331

breaker-failure, 325–326, 334–336

bridge-type, 154

classification, 1–3

coordination of, 232–237

definition, 1

directional overcurrent, 267–268,

269–270

inverse-time, 308–321

distance (see Distance relays)

fault-current detectors, 330

feeder protection, 365, 373–374

ground-distance, 254–257, 265–267

ground overcurrent, 112–113, 114,

238

406 Index

[Relays]

input types, 1

instantaneous, 308–309, 325

inverse, 268, 270, 280

for line protection, 230 (table), 244–

245

loss-of-field, 345–346

low excitation detecting, 128

monitoring, 1

for motoring detection, 131

for multiple-loop systems, 308–321

negative sequence, 157

operating principle, 2

operation time, 235

overcurrent, 150, 216, 344

overfrequency, 138, 391, 393

performance characteristics, 2

phase, 84

phase-distance, 250–252

for phase faults, 230 (table)

polarity, 15–16

protective, 1–6

reclosing, 1, 366, 369–373, 369–377

regulating, 1

synchronism-check, 1, 366–367, 377–

380

thermal, 126, 153–155

underfrequency, 138, 391–392

See also Devices; Distance relays;

Electromechanical relays;

Frequency relays; Solid-state

relays; specific relays

Relay taps:

in bus protection, 221–222

for overcurrent relay, in line

protection, 233

in transformer protection, 172, 201–

202

Reliability, 3, 4

REL-100 relays, 253

REL-300 relays, 252–253, 306

REL-350 relays, 276

REL-512 relays, 253–254, 374–375

Remote backup, 323–324

with breaker-failure protection, 326–

327

Remote ground, 105

Remote tripping, 197

REM543 relays, 153, 157

Requirements

degree of protection, 5

Reset times:

and breaker-failure protection, 332,

333

and reclosing relays, 368

Residual current, false, 148

Residual flux, 89–91

Resistance:

arc, 20–21, 265

arc/fault, 272

tower footing, 265

Resistance grounding, 110–112

Resistance temperature detectors

(RTDs), 126, 153–154

Resistive loads, 392

Resistor switching, 77

Resonant grounding, 109

Reverse, as concept, 306

Reverse active power, 130–134

Reverse characteristics, of impedance

units, 294–297

Ring buses:

backup protection, 326, 328–329, 336

normal protection, 215


ROM (read-only memory), 95

check-sum test, 103

Root-mean-square (rms) values, 100

RTD (resistance temperature

detector)-input type thermal

relays, 126, 154

RTQTB-061 units, 171

Sample-and-hold (S/H) circuit, 95–97

Sampling, 95–97

non-synchronous, 98–99

SA-1 relays, for shunt reactor, 206, 207

Saturation, 75, 81–82, 213–215

SBF-1 breaker-failure relay, 334–336

SC relay, 137

SDBU out-of-step tripping, 349–350

SDG ground units, 289–290

SDG relays, 307–308

SDGU ground units, 289–290


SD voltage, 354–355

Sectionalizers, 277

Security, 3

Selective backup protection, 226–227,

323–324

Selective reclosing, 366

Self-polarized ground units, 291–292


and phase-to-ground faults, 300–301

double, 301–302


reverse characteristics, 297

Self-testing microprocessor hardware,

103

analog text, 103

check-sum, 103

dead-man timer, 103

[Self-testing microprocessor hardware]

non-volatile memory test, 103

RAM test, 103

Semiconductors, 50–51

Sensitive ground relaying, 112–114

ground overcurrent with

conventional current


ground overcurrent with zero

sequence current

transformers, 114

ground product relay with



Sensitivity:

maximum, 18

of remote backup, 324

Separation, 75–76

Sequence filters. See Sequence

networks

Sequence networks, 47–50

composite sequence current, 26–32,

48–49

connections and voltages, 27–28

connections for faults and

unbalances, 28

and operational amplifiers, 69

reduction, 29–32

sequence voltage networks, 49–50

zero sequence networks, 27, 47–48

Series-capacitor compensated-line, 273

SGR relays, 371–371

Shielding, 77, 78

Short lines, 270–272

Shunt reactor protection, 203–211

differential protection, 206–207

overcurrent protection, 205–206

rate-of-rise-of-pressure protection,

205, 209–211

reactors on delta system, 207–209

shunt reactor applications, 203–205

turn-to-turn faults, 209–211

Silicon-controlled rectifiers (SCR). See

Thyristors

Simple high-pass filter, 66–67

Simple low-pass filter, 66

Simultaneous transfer, 142

Single-bank capacitor switching

Single-machine infinite-bus, 353–354,

355–356

Single-pole trip-system, 337

Single-shot reclosing relays, 369–371

Slow transfer, 142–143

Small disturbance (SD) angle, 353–354

Small disturbance (SD) voltage

instability, 354–355

Index 407

Small stations, 7

Solid-state out-of-step relaying, 349–

350

Solid-state relays:

for bus protection, 222–223

classification, 2

for line protection, 244

for negative sequence voltage

protection, 156

for out-of-step protection, 349–350

rack-mounting, 7–8

single shot reclosing, 370–371, 372

for switchgear applications, 155

Solid-state single-shot reclosing relay

Solid-state units, 50–53

principal logic units, 52–53

semiconductor components, 50–51

solid-state logic units, 52

SOQ relay, 124

Source impedance ratio, 270–272

Spark-gaps, 273

Speed, 3, 276

of remote backup, 324

Spinning reserve, 383–384

Split-phase winding, generators with,

119

Square wave detectors, 69–70

Stability. See System stability; Voltage

stability

Standardization, 2–3

Station auxiliaries, bus transfer system

for

Station-bus protection. See Bus

protection

Stator ground fault protection, 120–

123

neutral third harmonic

undervoltage, 120

neutral-to-ground fault detection,

121–122

95% ground relays, 120–121

100% winding protection, 120, 122–

123

unit-connected schemes, 120

Steady-state system stability, 339–340

Steam turbines, 131

Subtractors, 66

Subtransmission lines, 229

Sudden-pressure relays (SPR), 185, 202

Supervisory control, 367–368

Surges, 71, 72, 79

See also Transients and surges

Sustained fault current, 24

SV relays, 137

Swings, 341–343, 344

Switchboard relays, 6–7

Switchgear applications, 155

Symbols, 9, 11–15

for ANSI accuracy class, 85–86

for comparator applications, 287–288

for logic units, 52

for operational amplifier, 63

Symmetrical component filter, 102

Symmetrical components, 21–39

basic concepts, 21–23

fault calculation in loop-type power

system, 32–37

fault evaluations, 39

phase shifts through transformer

banks, 37–39

sequence impedances, 24–26

synchronous machinery, 24


transmission lines, 26

sequence networks, 26–32

connections and voltages, 27–28

faults and imbalances, 28

reduction, 29–32

sequences, in three-phase system,

23–24

system neutral, 23

Sympathetic inrush, 165–166

Synchronism:

automatic, 379–380

loss of, 128, 158

and reclosing, 369

Synchronism check:

angular method, 378–379

and bus tie breaker, 142

and generator protection, 142

phasing voltage method, 377–378

purpose, 379


Synchronizing relays, 1, 366–367, 377–

378

in automatic system, 379–380

Synchronous machinery, 24

System configuration, 5

System grounding, 20, 105–115

ground fault protection for three-

phase, four-wire system, 114–

115

multigrounded, 115


reactance grounding, 108–110

resistance grounding, 110–112

sensitive ground relaying, 112–114

ground overcurrent relay



zero sequence current

transformers, 114

[System grounding]

ground product relay with



ungrounded systems, 105–108

System stability, 339–352

out-of-step conditions, 343–345

circuit breakers, 344

directional comparison systems,

344

distance relays, 343

overcurrent relays, 344

phase comparison (pilot-wire)

schemes, 344


underreaching transfer-trip

schemes, 344

out-of-step relaying, 345–352

relay quantities during swings, 341–

343

steady-state, 339–340

transient, 340–341

Tapped transmission lines, 230

Taps:


load tap changers, 362

and loop circuits, 269

multiples, 311, 313

overcurrent relay, in line protection,

233

in transformer protection, 172, 201–

202

T class, 86

Temperature, 50

Termination, suppression by, 77

Testing, 327

Thermal capability curves, 146–147

Thermal relays, 153–154, 153–155

RTD-input type, 126, 154

Thermal replica relays, 126, 154–155

Thermal units, 47

Thermistors, 50

Three-line connections, 6

Three-phase, four-wire systems:

ground fault protection for, 114–115

multigrounded, 115


Three-phase faults, 298, 299–300, 345

Three-phase systems:

connected to directional relays, 17

notation, 14

sequences, 23–28

symmetrical components method,

21–23

Three-phase units, 288–289, 295

408 Index

Three-pole reclosing, 366

Three-winding banks, 24–26, 179

Threshold squarers, 69–70

Thyristors, 51

Time:

in breaker-failure schemes, 329–336

coordinating interval, 233

deionizing, 366

interval coordination, 233

inverse, 234–235, 238, 268–269, 270

example, 308–321

and load-shedding, 388, 390–391

and microprocessor relaying, 95–97,

236

nonsynchronous sampling, 98–

99

to reach saturation flux density, 89

relay operation, 235

Time-current characteristics, 234–237

Time-delay, 53–54, 60–61

and power line protection, 232–237

Time overcurrent protection, 232–237

Time-overcurrent relays, 308–309

Timers, 60–61

adjustability, 61

for breaker-failure protection, 331,

332, 333–336

and distance relay, 125

and locked-rotors, 150, 152

and reclosing, 373

and turbine blades, 138

Toggle circuits, 61

Torque, 47, 155, 196

Tower footing resistance, 265

TPU relays, 171, 172, 180, 193, 198, 199

checks, 177, 180

TPX transformers, 88

TPY transformers, 88

TPZ transformers, 88

Transfer bus, 215

Transfer-trip schemes, 344

Transformer banks:

and bus protection, 223–224

initial inrush, 163–165

and line protection, 261–265

maximum inrush, 165

phase shifts, 37–39

remote tripping, 197


Transformer protection, 163–203

differential relaying for, 166–173

differential relays, 166–171

general guidelines, 171–173

and frequency relays, 392

magnetizing inrush, 163–166

initial inrush, 163–164

[Transformer protection]

recovery inrush, 165


phase-angle regulators and voltage

regulators, 197–202

remote tripping of transformer

bank, 197

sample checks for differential relay

application, 173–182

checks for multiwinding banks,

178–180

checks for two-winding banks,

173–178

modern microprocessor relay, 180

typical applications, 180–193

differential protection generator-

transformer unit, 183

differential scheme with harmonic

restraint relay supervision,

180–182

distance relaying for backup, 192

ground source on delta side, 182–

183

industrial and commercial, 193–

197

overcurrent and backup


overcurrent relay with HRU

supplement, 192–193

overexcitation protection of

generator-transformer

unit, 183

sudden-pressure relay (SPR), 185

three-phase banks of single-phase

units, 183

zig-zag protection, 202–203

See also Shunt reactor protection

Transformers:

air-gap current, 91

burden limit calculation, 177–178

in bus protection, 215, 216

bus with banks of, 223–224

BYZ, 148–149

commercial, 193–197

doughnut current, 148

with ground product relay, 113–114

overexcitation limits, 184–185

phase-angle regulating, 197–202

phase shifts through banks of, 37–39

polarity of, 15

quality, 81

for short-line application, 271–272

small, 196

system impedance, 24–26

two current, 118

voltage, 91–93, 271–272

[Transformers]

voltage-regulating, 197–202

Wye-connected, 107, 195, 199

zero sequence type (BYZ), 114

See also Current transformers;

Instrument transformers

Transient overreach, 272

Transient reactance, 24

Transients:

definition, 71

and line protection, 275

Transients and surges, 71–79

differential- and common-mode, 72

electromagnetic induction, 72

electrostatic induction, 71–72

high-voltage system origin, 73–74

bus deenergization, 73–74

capacitor switching, 73

coupling capacitor voltage

transformer (CCVT)

switching, 74

transmission line switching, 74

low-voltage system origin, 74–75

battery circuit grounding, 75

current transformer saturation, 75

direct current circuit energization,

75

direct current coil interruption,

74–75

other sources, 74

protective measures, 75–79

buffers, 78

energy requirement increase, 79

optical isolators, 78

radial routing, of control cables, 78

separation, 75–76

suppression at source, 77–78

by shielding, 77, 78

by termination, 77

by twisting, 77–78

Transient stability, 340–341

Transistors, 50–51

Transmission lines:

as classification, 229

instantaneous reclosing, 369

length of, 235, 257–259, 270–272

equation, 281

mutual induction, 243–244

out-of-step relaying for, 346

switching, 74

system impedance, 26

See also Power lines

Tripping


Tripping:

energy for, 7

Index 409

[Tripping]

incorrect, 4

True earth, 105

Turbine-blade damage, 391

Turbines:

blades, 138–139, 391

and frequency, 138–139

gas, 131

generator, indirect cooling, 123

hydraulic, 131–132

steam, 131

and timers, 138

Turn-to-turn faults, 209–211

Twisting, 77–78

251G relay, 196

Underfrequency relays, 138, 391–392

Underreaching transfer-trip schemes,

344

Ungrounded systems, 105–108

ground faults on, 105–107

detection of, 107–108

Uniflex relays, 289–290

Unigrounded four-wire systems, 101–

102

Unijunction transistors, 51

Variable-reference circuit, 55

Varistors, 50

Vectors. See Phasors

Virtual ground, 64

Voltage:


in circuit diagrams, 12

of faults, calculation of, 31–32, 36–37

[Voltage]

and flux, 184

harmonic, 120, 122

induced, in a coil, 126

knee, 220

low, and arc resistance, 20

memory, 299

notation for, 12

and phase distortion, 21

phasing method, 377–378

in reactance-grounding, 109

reversal of, 274

SD, 354–355

in sequence networks, 27–28

standard U.S. ratings, 91

and torque, 155

in ungrounded systems, 106–107,

109

See also Harmonic voltage; Voltage

stability

Voltage block function, 373–374

Voltage-controlled overcurrent relay,

125

Voltage polarization, 240

Voltage-regulators, protection of, 197–

202

Voltage sensing devices, 93

Voltage stability, 353–363

instability indices, 357–362

MVA margin, 360–362

large-disturbance instability, 355–

356

protection schemes, 362–363

load tap changers, 362

reactive power control, 362

[Voltage stability]

small-disturbance instability, 353–355

Voltage transformers, 91–93

bushing, 92, 93

equivalent circuit, 92

for short-line application, 271–272

Volts per hertz protection, 126

Watt-type directional unit, 16

Weak feed terminal, 231

Windings:

three or more, 172, 178–180

two, 173–178

zig-zag connected, 202–203

Wye-connected transformers, 107, 195,

199

Zener diodes, 50, 77

Zero sequence impedance, 24, 26, 39

mutual impedance, 265–267

Zero sequence networks, 27, 35, 47–48

Zero sequence polarization, 243, 245–

247

Zero sequence transformer, 114

BYZ, 148–149

Zero sequence voltage comparison, 336

Z2G relays, 291–292

Z3G relays, 291–292

Zig-zag transformer protection, 202–

203

Zone application, of distance relaying,

257–259

Zones:

in distance relaying, 257–259

of protection, 3, 4

410 Index

About the Editor

WALTER A. ELMORE is Consulting Engineer, Blue Ridge, Virginia, and retired from the Relay Division of theABB Power T & D Company, Inc., Coral Springs, Florida. The author or coauthor of more than 100 professionalpublications including Pilot Protective Relaying (Marcel Dekker, Inc.), Mr. Elmore is a Life Fellow of the Instituteof Electrical and Electronics Engineers (IEEE) and a member of the National Academy of Engineering, and holdssix patents. He is a recipient of the IEEE’s Gold Medal for Engineering Excellence (1989) and the Power SystemRelaying Committee Award for Distinguished Service (1989). A registered Professional Engineer in Florida, Mr.Elmore received the B.S. degree (1949) in electrical engineering from the University of Tennessee, Knoxville

Documents

Electrical Protective Relay Theory and Applications