Transient Testing of Protective Relays: Study of Benefits and Methodology

Transcription

1 PSERC Transien Tesing of Proecive Relays: Sudy of Benefis and Mehodology Final Projec Repor Power Sysems Engineering Research Cener A Naional Science Foundaion Indusry/Universiy Cooperaive Research Cener since 1996

2 Power Sysems Engineering Research Cener Transien Tesing of Proecive Relays: Sudy of Benefis and Mehodology Final Projec Repor Projec Team Faculy Mladen Kezunovic, Projec eader, Texas A&M Universiy Sakis Meliopoulos, Georgia Insiue of Technology Ward Jewell, Wichia Sae Universiy Graduae Sudens Jinfeng Ren, Texas A&M Universiy Q. Binh Dam, Georgia Insiue of Technology Piyasak Poonpun and Miaolei Shao, Wichia Sae Universiy PSERC Publicaion March 2008

3 Informaion abou his projec For informaion abou his projec conac: Mladen Kezunovic, Ph.D. Texas A&M Universiy Deparmen of Elecrical Engineering College Saion, TX Tel: Fax: Power Sysems Engineering Research Cener This is a projec repor from he Power Sysems Engineering Research Cener PSERC. PSERC is a muli-universiy Cener conducing research on challenges facing he elecric power indusry and educaing he nex generaion of power engineers. More informaion abou PSERC can be found a he Cener s websie: hp:// For addiional informaion, conac: Power Sysems Engineering Research Cener Arizona Sae Universiy Deparmen of Elecrical Engineering Ira A. Fulon School of Engineering Phone: Fax: Noice Concerning Copyrigh Maerial PSERC members are given permission o copy wihou fee all or par of his publicaion for inernal use if appropriae aribuion is given o his documen as he source maerial. This repor is available for downloading from he PSERC websie Texas A&M Universiy, Georgia Insiue of Technology and Wichia Sae Universiy. All righs reserved.

4 Acknowledgemens We wish o hank he indusry advisors ha acively paricipaed in he discussions during he course of his projec: Bajarang Agrawal Arizona Public Service, Ali A. Chowdhury California Independen Sysem Operaor, John Horwah Exelon, Joseph Hughes Elecric Power Research Insiue, Richard Hun NxPhase, Bill Middaugh Tri-Sae Generaion and Transmission, and Don Sevcik CenerPoin Energy. Graduae sudens and research saff who carried ou projec aciviies over he years were Jinfeng Ren and Chengzong Pang Texas A&M Universiy; Q. Binh Dam, George Sefopoulos, and Dr. George Cokkinides, Visiing Professor Georgia Insiue of Technology; and Miaolei Shao and Piyasak Poonpun Wichia Sae Universiy. i

5 Execuive Summary The operaional securiy of he power sysem depends upon he successful performance of he housands of relays ha proec he sysem from cascading failures, ha proec equipmen, and ha help balance load wih generaion when sysem frequency is oo low or oo high. The failure of a relay o operae as inended may jeopardize he sabiliy of he enire sysem and equipmen in i. In fac, major sysem failures afer a disurbance are more likely o be caused by uninended proecive relay operaion raher han by he failure of a relay o ake an acion a all. Appropriae relay esing provides one line of defense agains relay failures. Relay esing can help validae he design of relay logic, compare he performance of differen relays, verify selecion of relay seings, idenify sysem condiions ha migh cause uninended relay operaion, and carry ou pos-even analysis o undersand he causes of uninended or incorrec relay acions. Relay esing improvemens need o coninue because of he new demands placed on relays from power sysem condiions ha are more variable in he pas, because of high cusomer expecaions for power delivery reliabiliy, and because of changing relay echnologies. The research described in his repor describes new approaches for esing disance relays, generaor proecion relays, and underfrequency load shedding relays. Resuls are provided for acual relay esing. Par I: Disance Relay Tess Texas A&M Universiy Disance relay esing can evaluae relay performance, calibrae relay seings, and idenify sysem condiions ha could cause unexpeced relay operaion. Developing a relay esing mehodology requires consideraion of how o model he power sysem o simulae specific sysem disurbances, how o selec and generae es scenarios, and how o execue relay ess efficienly. The efficiency and effeciveness of relay esing can be enhanced wih a es case library conaining scenarios ha enable consisen ye robus esing. In his research, a laboraory was used o es hree differen disance relays using a proposed es mehodology wih associaed es ools and es case library. The esing focused on proecive relay operaion under ransiens. Conformance and compliance ess were conduced. Conformance Tes: The objecive is o es he basic funcionaliy of a relay, o verify is operaing characerisics, o calibrae he relay seings, and o implemen periodic mainenance esing. Saisical performance daa are colleced on relay operaing characerisics and ripping imes using wide-ranging disurbance condiions generaed hrough simulaion. Compliance Tes: The objecive is o es if acual relay performance maches expeced performance under aypical ye possible power sysem condiions. The rip/no rip responses and relay operaing ime performance are measured under specific scenarios. Compliance ess can be used in a pos-even analysis o analyze he causes of an unwaned relay operaion The IEEE Power Sysem Relaying Commiee PSRC reference model and IEEE 14- bus sysem were used o simulae disurbance scenarios. Sofware programs were developed for auomaed esing for creaing es cases, execuing bach ess, and ii

6 collecing relay even repors. The es case library included es scenarios, records from digial faul recorders DFRs and blackou scenarios of ineres. Tes resuls provided informaion ha was no documened in he relay manuals, and ha definiely could affec proper coordinaion and performance of he relaying schemes. The conformance es resuls indicaed ha relay operaing characerisics should be carefully seleced applied o improve he dependabiliy of he relaying scheme. The compliance es resuls indicaed ha he zone 3 relays operaed incorrecly in a few unusual power sysem operaing condiions. Thus, quadrilaeral operaing characerisic may be needed o assure correc relay responses. Par II: Generaor Relay Tess Georgia Insiue of Technology Proecive generaor relays are usually esed agains simplified generaor models or simplified es signals. Many facors may vary wih he locaion and generaor, including he impedances of he nework o where he generaor is conneced o, operaing poin, grounding arrangemens, ec. The esing also should ensure ha he seings of he relay are consisen wih he inended proecion scheme. Generaor relay ess using realisic models of he generaor and he elecric power sysem can verify consisen behavior of a relay regardless of he proeced generaor, and asser ha he inended proecion schemes are robus for a variey of faul condiions. A comprehensive esing plaform was buil o reproduce and simulae condiions in he sysem as close o realiy as possible. The plaform included a a power sysem simulaor o accuraely compue shor-circui condiions as seen in an acual sysem by he proecive relays; b a signal condiioning uni ha reproduces he simulaed volages and currens a relay insrumenaion volage and curren level, as if hey were delivered by acual poenial and curren ransformers; and c a se of procedures o conduc and validae he differen ess of he generaor relay, including relay connecions, sofware configuraion, and differen es scenarios. A comprehensive se of generaor ransien evens were creaed o exercise all he funcions of a modern generaor relay. For accurae esing, as many common characerisics of all generaors are needed o simulae generaor responses ha are as close o field observaions as possible. To achieve he highes accuracy possible, he sofware plaform included a full ime domain, ransien, wo-axis synchronous generaor model wih access o generaor windings for faul creaion in he windings. The simulaion sofware models he power sysem more accuraely han mos oher exising approaches. The simulaion sofware is based on full hree-phase models of power sysem componens ha are described by heir physical parameers. The simulaor accuraely simulaes he dynamics of he models by using he quadraic numerical inegraion mehod, which is more precise compared o oher mehods commonly used mehods in power sysem analysis. Using virual relay esing, configuraion and waveform daa were sen direcly o he inpus of he relay funcions, and he relay oupus were processed on he hos compuer wih he benefis of specialized analysis sofware. Virual esing faciliaes relay esing by eliminaing he consrains of a hardware seup, including waveform generaion, wiring, and communicaions. Comprehensive ransien esing was conduced on wo differen generaor proecive relays. The deailed resuls are given in he repor. iii

7 Par III: Underfrequency oad Shedding Relay Tess Wichia Sae Universiy Researchers If insufficien generaion is available on he sysem o mainain sabiliy, non-criical loads can be removed or shed from he sysem o resore a balanced condiion. Such mehods of auomaic load shedding are designed as a las resor o preven a major sysem ouage. Underfrequency load shedding UFS relays deec overload condiions by sensing low sysem frequency and shedding enough load o rebalance generaion and load, and reesablish he nominal frequency. UFS relays are able o auomaically resore load afer frequency recovery. UFS is an effecive and reliable mehod ha helps o preven blackous. A review of he Final Repor on he Augus 14, 2003 Blackou in he Unied Saes and Canada: Causes and Recommendaions, prepared by he U.S.-Canada Power Sysem Ouage Task Force, indicaes ha, during he cascading evens leading up o he widespread blackou: UFS relays operaed properly, according o heir seings Seings for some UFS relays may no have been appropriae for heir applicaions Regardless of seings, load shedding by UFS and undervolage relays would no be expeced o miigae he magniude of evens ha occurred during his disurbance. While UFS relays appeared o operae as se during he 2003 blackou, a number of issues regarding heir operaion were neverheless idenified. These issues, which are no addressed by convenional relay es mehods, include he effecs on relay operaion of: Rae of change of frequency Coninuous, raher han sep, changes in frequency Rapid flucuaions, including boh increases and decreases, in frequency Overfrequency evens Oher evens idenified by simulaion or recording of acual evens. To address hese issues, wo es proocols, which go beyond hose ess usually performed using commercial UFS relay es sysems, were developed. The conformance es proocol subjecs a relay o a series of ess whose values are deermined by he relay specificaions. The applicaion es proocol subjecs he relay o evens generaed hrough simulaions of a ypical sysem using elecromagneic ransiens sofware. A hird se of ess can also be performed if acual recorded even daa is available. Recorded evens can be played back in he laboraory o deermine relay response o acual evens. Boh conformance and applicaion ess were performed on wo commonly-used digial UFS relays. Relay response was ou of manufacurers specificaions for some of he ess. Indusry eam members indicaed, ha he magniude of he errors idenified were well wihin olerances expeced by indusry, and ha such errors had no pracical effec on he relays abiliies o shed load as expeced during underfrequency evens. Fuure Work: I has been recognized ha forming a library of es cases using records from blackous or common power sysem model would be quie beneficial. Developing mehodology and ools for boh laboraory and field esing aimed a evaluaing how GPS synchronized IEDs, including relays, will perform under various operaing condiions is also needed. iv

8 Table of Conens 1.0 INTRODUCTION PART I: DISTANCE REAY TEST TAMU Inroducion Tes Mehodology Tes Classificaion Tes Sysem Model Tes Scenarios Generaion Tes Case ibrary Tes Implemenaion Tes Procedure aboraory Seup Tes Resuls Power Sysem Daa for Conformance Tes Power Sysem Daa for Compliance Tes Disance Relay Seing Tes Resuls and Analysis Fuure Work PART II: GENERATOR REAY TEST GATECH Overview Descripion of Plaform Generaor Proecion Relays Overview of he Tesing Plaform Sofware Tes Bench Virual Relay Tesing Generaion Relay Tesing Seup Purpose Even Simulaion and Tesing Procedure Descripion of Tes Sysem # Descripion of Tes Sysem # Beckwih Relay Seup v

9 3.3.6 Simulaion of Power Sysem Evens Reporing Tess and Simulaed Evens Basic Even Triggering and Oscillographic Record Analysis Beckwih M3425-A SE 300-G Equaions for he Proecion Variables Noaion Seup 1 Single Curren Source a Neural Side Only Seup 2 Same Currens In and Ou Operaing Curren and Resrain Curren Individual Proecion Funcion Tess M-3425A Common Procedures Funcion 87 Phase Differenial Funcion 27 Phase Undervolage Expanded Tes Scenarios Mock Generaor Acceleraion Three-Phase Faul wih Unsable Swings afer Clearance Wide-Area Parial oad Shedding Inadveren Generaor Breaker Operaion Generaed Waveforms and Relay Response Disconneced Phase Three-Phase Faul followed by Generaor Breaker Operaion Tes on Boh Relays Fuure Work PART III: OAD SHEDDING REAY TEST WSU Inroducion Background Under-frequency oad Shedding UFS Relay Inroducion UFS Tess UFS Research Repor Organizaion Review of UFS Relay Operaion during he 2003 Norh American Blackou vi

10 4.2.1 Background UFS Relay Tes Sysem UFS Relay Tes Sysem Overview UFS Relay Tes Sysem Hardware Sofware Under-frequency oad Shedding Relays Under-frequency oad Shedding Relay Tes Scenarios Conformance Tes Tes Waveforms Applicaion Tes UFS Relay Tes Resuls Conformance Tess Applicaion Tess Inerpreaion of he Resuls Conformance Tess Applicaion Tess Error Analysis Fuure Work CONCUSION Disance relays Generaor Relays Underfrequency oad Shedding Relays PROJECT PUBICATIONS REFERENCES APPENDIX A: INE DISTANCE REAY TEST A.1 Relay Seings A.2 Tes Resuls APPENDIX B: GENERATOR REAY TEST B.1 Generaor Relay Proecion Scheme and Connecions B.2 is of Generaor Evens for Relay Tesing B.3 High-Fideliy Generaor Model for Even Simulaion B.3.1 Inroducion B.3.2 Synchronous Machine Full Transien -Domain Model vii

11 B.3.3 Exciaion Sysem Model B.3.4 Prime Mover Sysem Model B.4 Example Response Char for Generaor Relay Tesing Evens B.5 IEEE COMTRADE Sandard Informaion for Relay Tesing B.5.1 A Primer on he IEEE COMTRADE File Forma APPENDIX C: OAD SHEDDING REAY TEST C.1 Tes Resuls C.2 13-Bus Tes Sysem viii

12 is of Figures Figure 2.1: One-line diagram for IEEE PSRC sysem... 4 Figure 2.2: Deailed model for IEEE PSRC sysem... 4 Figure 2.3: ATPdraw model for IEEE PSRC sysem... 4 Figure 2.4: One line model for IEEE 14-bus sysem... 5 Figure 2.5: ATPdraw model for IEEE 14-bus sysem... 5 Figure 2.6: Bach simulaion program block diagram... 6 Figure 2.7: ATPdraw model for manual simulaion... 6 Figure 2.8: Tes case library Figure 2.9: Program inerface for convering MAT file o RA file Figure 2.10: Program inerface for convering ATP file o COMTRADE file Figure 2.11: Example for loading es cases Figure 2.12: Example for waveforms displayed by Relay Assisan sofware Figure 2.13: Example of es resul for inernal faul Figure 2.14: Example of even repor shown as oscillograph Figure 2.15: Sofware framework for relay esing Figure 2.16: aboraory seup for relay esing Figure 2.17: Block diagram for relay es environmen Figure 2.18: One-line diagram of he ransmission line model for Conformance Tes Figure 2.19: One-line diagram of he ransmission line model for Compliance Tes Figure 2.20: Example of comparaive es resuls Figure 2.21: Three-phase volage and curren waveforms from relay even Figure 2.22: Ou-of-sep funcion and parameers Figure 2.23: oad encroachmen funcion and parameers Figure 3.1: Overall projec approach Figure 3.2: The Beckwih M3425-A generaor proecion relay Figure 3.3: The SE 300-G generaor proecion relay Figure 3.4: Overview of he esing plaform Figure 3.5: Sample model definiions in he high-fideliy simulaor sofware: a ransmission line physical design parameers and b hree-phase subsaion bus connecions Figure 3.6: Comprehensive generaor model of he sofware plaform: a parameer definiion window and b visual represenaion of he connecion poins ix

13 Figure 3.7: Simulaion oupu in various forms sored in a COMPTRADE file Figure 3.8: Summary of he sofware porion of he esing plaform Figure 3.9: Tes bench layou Figure 3.10: Picure of he acual laboraory seup Figure 3.11: Tes bench layou for relay esing wih a scale model Figure 3.12: Three-phase auxiliary volage channels for relay and PMU esing Figure 3.13: Three-phase auxiliary volage channels for relay and PMU esing Figure 3.14: Virual relay esing principle Figure 3.15: Nework schemaic of es sysem # Figure 3.16: Generaor and sep-up ransformer grounding scheme Figure 3.17: Seings for he generaor grounding ransformer Figure 3.18: Seings for he generaor sep-up ransforme Figure 3.19: Seings for he ransmission line in he es sysem Figure 3.20: Seings for he equivalen source a he infinie source Figure 3.21: Nework schemaic of es sysem # Figure 3.22: Parameers for Generaor 1 Tes Sysem # Figure 3.23: Parameers for Generaor 2 Tes Sysem # Figure 3.24: Parameers for Generaor 3 Tes Sysem # Figure 3.25: General relay seings dialog box Figure 3.26: Manual oupu conac conrol wih IPSuiliy and visual feedback from he relay Figure 3.27: The oscillograph rerieval screen Figure 3.28: Graphical sample of he waveforms and sae of he oupu conac Figure 3.29: Records of he iniiaion of he volage supply o he relay Figure 3.30: Characerisic of Funcion 87 funcion wih 0.3 A pickup and 10% slope. 54 Figure 3.31: Seings for he differenial relay funcion Figure 3.32: Funcion 87, phase differenial, waveforms for he combined pickup and dropoff ess Figure 3.33: Funcion 27, phase undervolage, waveforms from pickup and dropoff es Figure 3.34: Funcion 27, phase undervolage, waveforms from ime delay es Figure 3.35: M-3425A relay rerieved waveforms for he proecion scenario Figure 3.36: Capure of he waveforms sen o he relays for Tes Sysem # x

14 Figure 3.37: Principle for wide area parial load shedding Figure 3.38: RMS values of elecric quaniies unil one second afer load drop Figure 3.39: Expanded simulaion shows coninuous increase of generaor frequency and roor slip Figure 3.40: Waveforms capured for inadveren breaker operaion Figure 3.41: Generaor 1 and load wih per-phase circui breaker Figure 3.42: Response of he sysem afer opening phase A of Generaor Figure 3.43: Response of he sysem afer opening phases B and C of Generaor Figure 3.44: Response of he sysem afer opening phase A of oad Figure 3.45: Response of he sysem afer opening phases B and C of oad Figure 3.46: Response on a hree-phase faul followed by opening of generaor breaker75 Figure 3.47: Superimposed relay measuremens of phase A volage afer display scaling ime shifing Figure 3.48: Phase undervolage oupu from a he 300-G and b he M-3425A relay 76 Figure 4.1: Configuraion of UFS relay es sysem Figure 4.2: UFS relay es sysem Figure 4.3: UFS relay es sysem sofware Figure 4.4: decay Figure 4.5: Volage wih 5 h, 7 h, 11 h, and 13 h harmonics Figure 4.6: Variable volage magniude Figure 4.7: Single line diagram of 13-bus equivalen sysem Figure 4.8: Generaor frequencies wihou UFS implemenaion Figure 4.9: Generaor frequencies wih UFS implemenaion 1 sep Figure 4.10: Generaor frequencies wih UFS implemenaion 2 sep Figure A.1: Phase disance proecion Figure A.2: Power swing proecion Figure A.3: oad encroachmen proecion Figure B.1: Insrumenaion connecions of he generaor proecion relays Figure B.2: M-3425A deailed connecions of measuremen channels o relay inpus for a ypical proecion scheme Figure B.3: M-3425 funcions available from ypical volage and curren wirings o he relay Figure B.4: Typical connecion diagram for he 300G relay xi

15 Figure B.5: 300G funcions available from ypical volage and curren wirings o he relay Figure B.6: Connecions beween signal amplifiers and he esed generaor relays Figure B.7: Elecrical model of a synchronous machine as a se of muually coupled windings Figure B.8: Mechanical model of synchronous machine as a roaing mass Figure B.9: Saor self-inducance as a funcion of θ Figure B.10: Muual inducance beween saor windings Figure B.11: Elemens of a generaor exciaion sysem Figure B.12: Volage source wih inernal impedance Figure B.13: Curren source circui Figure B.14: DC armaure circui wih inernal impedance Figure B.15: Elemens of a generaor prime mover sysem Figure C.1: Exciaion sysem model for synchronous machine Figure C.2: Branch daa of 13-bus equivalen sysem xii

16 is of Tables Table 2.1: Faul Scenarios for Conformance Tes... 7 Table 2.2: Non-faul Scenarios for Conformance Tes... 8 Table 2.3: Tes scenarios for Compliance Tess Table 2.4: Funcions and sofware for seleced disance relays Table 2.5: Power sysem daa for Conformance Tes Table 2.6: Secondary impedances for Conformance Tes Table 2.7: Power sysem daa for Compliance Tes Table 2.8: Funcions able applied for es Table 2.9: Example of saisical es resuls Table 2.10: Summary for relays performance Table 3.1: Funcion 87, rigger and arge imes and corresponding curren Table 3.2: Funcion 87, rigger and drop-ou imes and corresponding currens Table 3.3: Funcion 27, imes and volage levels for funcion rigger and release Table 3.4: Funcion 27, ime delay es resuls Table 3.5: Seings for seleced proecion funcions Table 4.1: Characerisic of power amplifiers Table 4.2: Relay 1 specificaions Table 4.3: Relay 2 specificaions Table 4.4: Seings of UFS scheme Table 4.5: oad shedding ime Table 4.6: Acual pickup frequency in Hz 100% Volage, 0% THD, Relay Table 4.7: Acual pickup frequency in Hz 100% Volage, 5% THD, Relay Table 4.8: Acual pickup frequency in Hz 100% Volage, 0% THD, Relay Table 4.9: Acual pickup frequency in Hz 100% Volage, 5% THD, Relay Table 4.10: Acual ime delay 100% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay Table 4.11: Acual ime delay 100% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay Table 4.12: Acual ime delay 100% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay Table 4.13: Acual ime delay 100% Volage, 5% THD, 0.9 Hz/sec Rae of Change, Relay xiii

17 Table 4.14: Acual ime delay 100% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay Table 4.15: Acual ime delay 100% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay Table 4.16: Acual ime delay 100% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay Table 4.17: Acual ime delay 100% Volage, 5% THD, 0.9Hz/sec Rae of Change, Relay Table 4.18: Applicaion es of relay 1 : 2 Cycles Table 4.19: Applicaion es of relay 2 : 3 Cycles Table 4.20: Daa for pickup frequency es 55 Hz Sepoin Table A.1: Seing able for SE-421 for Conformance Tes Table A.2: Seing able for SE-421 for Compliance Tes Table A.3: Tes resuls for condiion F1 for SE Table A.4: Tes resuls for condiion F2-1 for SE Table A.5: Tes resuls for condiion F2-2 for SE Table A.6: Tes resuls for condiion F3 for SE Table A.7: Tes resuls for condiion F4-1 for SE Table A.8: Tes resuls for condiion F4-2 for SE Table A.9: Tes resuls for condiion F5 for SE Table A.10: Tes resuls for condiion F6-1 for SE Table A.11: Tes resuls for condiion F6-2 for SE Table A.12: Saisical es resuls for inernal fauls for SE Table A.13: Tes resuls for no-faul scenarios for SE Table A.14: Compliance es resul for SE Table A.15: Tes resuls for condiion F1 for SE Table A.16: Tes resuls for condiion F2-1 for SE Table A.17: Tes resuls for condiion F2-2 for SE Table A.18: Tes resuls for condiion F3 for SE Table A.19: Tes resuls for condiion F4-1 for SE Table A.20: Tes resuls for condiion F4-2 for SE Table A.21: Tes resuls for condiion F5 for SE Table A.22: Tes resuls for condiion F6-1 for SE Table A.23: Tes resuls for condiion F6-2 for SE xiv

18 Table A.24: Saisical es resuls of inernal fauls for SE Table A.25: Tes resuls of no-faul scenarios for SE Table A.26: Compliance es resul for SE Table A.27: Tes resuls for condiion F1 for GE D Table A.28: Tes resuls for condiion F2-1 for GE D Table A.29: Tes resuls for condiion F2-2 for GE D Table A.30: Tes resuls for condiion F3 for GE D Table A.31: Tes resuls for condiion F4-1 for GE D Table A.32: Tes resuls for condiion F4-2 for GE D Table A.33: Tes resuls for condiion F5 for GE D Table A.34: Tes resuls for condiion F6-1 for GE D Table A.35: Tes resuls for condiion F6-2 for GE D Table A.36: Saisical es resuls for inernal fauls for GE D Table A.37: Tes resuls of no-faul scenarios for GE D Table A.38: Compliance es resul for GE D Table B.1: Type, exension, and purpose of he hree COMTRADE file ypes Table C.1: Acual pickup frequency in Hz 100% Volage, 0% THD, Relay Table C.2: Acual pickup frequency in Hz 85% Volage, 0% THD, Relay Table C.3: Acual pickup frequency in Hz 115% Volage, 0% THD, Relay Table C.4: Acual pickup frequency in Hz 100% Volage, 5% THD, Relay Table C.5: Acual pickup frequency in Hz 85% Volage, 5% THD, Relay Table C.6: Acual pickup frequency in Hz 115% Volage, 5% THD, Relay Table C.7: Acual pickup frequency in Hz 100% Volage, 0% THD, Relay Table C.8: Acual pickup frequency in Hz 85% Volage, 0% THD, Relay Table C.9: Acual pickup frequency in Hz 115% Volage, 0% THD, Relay Table C.10: Acual pickup frequency in Hz 100% Volage, 5% THD, Relay Table C.11: Acual pickup frequency in Hz 85% Volage, 5% THD, Relay Table C.12: Acual pickup frequency in Hz 115% Volage, 5% THD, Relay Table C.13: Acual ime delay Table C.14: Acual ime delay 100% Volage, 0% THD, 0.5 Hz/sec Rae of Change, Relay Table C.15: Acual ime delay 100% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay xv

19 Table C.16: Acual ime delay 85% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay Table C.17: Acual ime delay 85% Volage, 0% THD, 0.5 Hz/sec Rae of Change, Relay Table C.18: Acual ime delay 85% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay Table C.19: Acual ime delay 115% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay Table C.20: Acual ime delay 115% Volage, 0% THD, 0.5 Hz/sec Rae of Change, Relay Table C.21: Acual ime delay 115% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay Table C.22: Acual ime delay 100% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay Table C.23: Acual ime delay 100% Volage, 5% THD, 0.5 Hz/sec Rae of Change, Relay Table C.24: Acual ime delay 100% Volage, 5% THD, 0.9 Hz/sec Rae of Change, Relay Table C.25: Acual ime delay 85% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay Table C.26: Acual ime delay 85% Volage, 5% THD, 0.5 Hz/sec Rae of Change, Relay Table C.27: Acual ime delay 85% Volage, 5% THD, 0.9 Hz/sec Rae of Change, Relay Table C.28: Acual ime delay 115% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay Table C.29: Acual ime delay 115% Volage, 5% THD, 0.5 Hz/sec Rae of Change, Relay Table C.30: Acual ime delay 115% Volage, 5% THD, 0.9 Hz/sec Rae of Change, Relay Table C.31: Acual ime delay 100% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay Table C.32: Acual ime delay 100% Volage, 0% THD, 0.5 Hz/sec Rae of Change, Relay Table C.33: Acual ime delay 100% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay Table C.34: Acual ime delay 85% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay xvi

20 Table C.35: Acual ime delay 85% Volage, 0% THD, 0.5 Hz/sec Rae of Change, Relay Table C.36: Acual ime delay 85% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay Table C.37: Acual ime delay 115% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay Table C.38: Acual ime delay 115% Volage, 0% THD, 0.5 Hz/sec Rae of Change, Relay Table C.39: Acual ime delay 115% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay Table C.40: Acual ime delay 115% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay Table C.41: Acual ime delay 115% Volage, 5% THD, 0.5 Hz/sec Rae of Change, Relay Table C.42: Acual ime delay 115% Volage, 5% THD, 0.9 Hz/sec Rae of Change, Relay Table C.43: Acual ime delay 85% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay Table C.44: Acual ime delay 85% Volage, 5% THD, 0.5 Hz/sec Rae of Change, Relay Table C.45: Acual ime delay 85% Volage, 5% THD, 0.9 Hz/sec Rae of Change, Relay Table C.46: Acual ime delay 115% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay Table C.47: Acual ime delay 115% Volage, 5% THD, 0.5 Hz/sec Rae of Change, Relay Table C.48: Acual ime delay 115% Volage, 5% THD, 0.9 Hz/sec Rae of Change, Relay Table C.49: Applicaion es of relay 1 delay: 2 Cycles Table C.50: Applicaion es of relay 2 delay: 3 Cycles Table C.51: Excier daa Table C.52: Generaor daa xvii

21 1.0 Inroducion This repor summarizes resuls from PSERC projec T-30, Transien Tesing of Proecive Relays: Sudy of Benefis and Mehodology, which was a join projec conduced by Texas A&M Universiy TAMU, Georgia Insiue of Technology GaTech and Wichia Sae Universiy WSU. The TAMU eam focused on he sudy of disance relay behavior under ransien and dynamic condiions. The GaTech eam sudied generaor proecion es requiremens and developed es mehods. The WSU eam approached load shedding relay es requiremens and performed series of ess o evaluae heir performance. All aciviies were aimed a specifying es requiremens ha reflec some difficul real ime scenarios and implemening novel es mehodology and ools o perform such ess. 1

22 2.0 Par I: Disance Relay Tes TAMU 2.1 Inroducion The securiy and reliabiliy of power sysem highly depend on he performance of he housands of relays. The correc operaion of proecive relay is supposed o clear he faul, as well as reduce and/or eliminae he impac of disurbances on power sysem. On he conrary, uninended or incorrec operaion may furher deeriorae he sysem condiion and even jeopardize he sabiliy of he enire sysem. A review of major sysem disurbances, such as blackous, indicaes ha a faal consequence of a disurbance is more likely o be caused by an uninended operaion of a proecive relay raher han he non-acion [1]. Appropriae relay esing should help validae he design of he relay logic, compare he performance of differen relays, verify selecion of relay seings, idenify vulnerable condiions ap o causing uninended operaions, and carry ou pos-even analysis for he undersanding of uninended or incorrec relay behavior. Many scieniss and engineers have pu much effor in developing esing ools and mehodologies. The Power Sysem Relaying Commiee of he IEEE Power Engineering Sociey formed he working group for Relay Performance Tesing I-13 in 1989, which promoed developmen of new relay es approaches. This secion of he repor describes he es classificaion, mehodology, power sysem modeling, es scenario generaion, and creaion of library of es case for esing disance relays for ransmission line proecion. I also commens on implemenaion and execuion of ess on disance relays in TAMU s lab, as well as on he resuls obained by such ess. According o he difference in he inpu signals, he relay ess can be classified ino wo caegories: phasor-based and ransien-based. Phasor-based relay ess: Predefined phasors represening differen pre-faul and faul condiion are used. Tes waveforms can be derived by simulaion from a simple power sysem model. The ideal sinusoidal signals are hen replayed ino relay inpus. By adjusing he magniude and angle of he signals, he operaing characerisic of a relay is measured and compared o a generic one or he one given by he vendor. Phasor-based relay esing mehod is a radiional one, which was widely used in he field in he pas. This seady sae mehod canno represen an acual siuaion during a faul and may no be used o fully verify he securiy or dependabiliy of a relay [2]. Transien-based relay ess: Transien signals used during esing represen acual ransiens generaed during fauls. They are replayed ino a relay hrough a digial simulaor. The ransien signals can be obained from simulaed faul scenarios or recorded waveforms from subsaions. Resuls of ransien-based relay esing are more 2

23 accurae han hose of radiional phasor-based mehods because he waveforms are much closer o he real faul signals [3]. 2.2 Tes Mehodology The es mehodology including comprehensive power sysem modeling, generaing es scenarios, auomaing simulaion and forming es case library is given as follows: Selec sandard power sysem models suiable for creaing differen disurbance scenarios. Generae a se of es scenarios hrough simulaion, and/or collec disurbances of ineres from digial faul recorder DFR and blackou evens. Form a library of es cases for easy reuse and uilizaion as a reference. Auomae he simulaion o minimize he es ime. Implemen comparaive ess for a se of differen relays wih similar funcions. Collec relay response evens, analyze he resuls and summarize hem in a es repor wih comparaive resuls Tes Classificaion The ransien ess for disance relay were he focus of he TAMU group sudy. Two differen ypes of ess wih differen es objecives are defined: conformance es and compliance es. The ransien-based mehod is used o implemen he conformance and compliance es on disance relays a TAMU labs. Conformance Tes: The objecive is o es he basic funcionaliy of he relays, verify he operaing characerisics, calibrae relay seings and implemen periodic mainenance es. The concern of his es is he saisical performance relaed o he relay operaing characerisic and ripping ime. To fulfill his es, a bach of es cases wih a variey of disurbance condiions including fauls and non-fauls are generaed hrough simulaion. Compliance Tes: The objecive of compliance applicaion es is o verify wheher a relay can operae correcly under peculiar circumsances in power sysem paricularly during abnormal operaing condiions. Tha is o say, his ype of es is o invesigae he compliance feaure ha real performance of a proecive relay complies wih is expeced performance. The concern of his es is he rip/no rip response and relay operaing ime performance under specific scenarios. A ypical example is he use of he recorded daa o analyze causes of an unwaned relay operaion in a pos-even analysis Tes Sysem Model Power Sysem Model for Conformance Tes Two power sysem models are used o simulae disurbances for he conformance es and compliance es: IEEE PSRC sysem and IEEE 14-bus sysem respecively. A reference model creaed by IEEE Power Engineering Sociey Power Sysem Relaying Commiee PSRC used for conformance es is described in [4]. The one-line diagram is given in Figure 2.1. This sysem has hree sources, four buses and single and parallel muualcoupled overhead ransmission lines. The deailed model is shown as Figure

24 BUS 1 BUS 2 BUS 4 nz2 1-nZ2 S1 Faul ocaion ZM Z4 5 S2 mz1 1-mZ1 CT PT TR2 REAY Alernae Poenial ocaion SW Alernae Poenial ocaion REAY BUS 3 S3 Z3 Alernae Poenial ocaion REAY Figure 2.1: One-line diagram for IEEE PSRC sysem BUS1 2F1 2F3 2F5 2F7 BUS2 S1 SRC1 4F1 4F5 BUS4 TR2H S2 1BK 1RBK 4F3 4F7 TR2 1F1 1F3 1F5 1F7 REAY REAY 3RBK BUS3 SRC3 S3 3F1 3F3 3F5 3F7 B3BK REAY Figure 2.2: Deailed model for IEEE PSRC sysem The ATPdraw implemenaion is given in Figure 2.3. This model is used for manual simulaion by seing various individual scenarios. Figure 2.3: ATPdraw model for IEEE PSRC sysem 4

25 Power Sysem Model for Compliance Model The one-line diagram and ATPdraw model for IEEE 14-bus sysem are given as Figure 2.4 and Figure 2.5 [5]. I has 5 synchronous machines, 20 branches, 11 consan impedance loads, circui breakers, and volage and curren measuremens. Various power sysem disurbances can be simulaed on his model as well as specific operaing sae sudy o find vulnerable condiions ap o cause relay uninended operaions. Figure 2.4: One line model for IEEE 14-bus sysem Figure 2.5: ATPdraw model for IEEE 14-bus sysem 5

26 2.2.3 Tes Scenarios Generaion Tes Scenarios for Conformance Tes The IEEE PSRC reference sysem is used for conformance ess. Boh auomaic and manual simulaion can be implemened in ATP [6]. The Bach simulaion program is developed in MATAB for PSRC reference model sysem is developed based on he ex version of ap file [7]. The simulaion block diagram is given in Figure 2.6. This se up can auomaically simulae differen faul scenarios wih differen faul ypes, locaions, resisances and incepion angles. The oupu of he waveforms can be P4, MAT convered by P42MAT program and COMTRADE convered by P42COM program. The ATPdraw model is developed for manually generaing es cases as shown in Figure 2.7. For he bach simulaion, he faul poin should be wihin 10%-90% of he line lengh because an ATP basic model has limiaion for disribued line model. For he faul posiions beween 0% and 10% beween 90% and 100%, ATPdraw model is used. For each simulaion, faul ype, locaion, resisance and incepion angle in his case need o be se manually and one scenario is generaed a one ime. Figure 2.6: Bach simulaion program block diagram Figure 2.7: ATPdraw model for manual simulaion For he conformance es using he IEEE PSRC reference sysem, wo caegories of scenarios are chosen: faul scenarios and non-faul scenarios. The deailed es iems are given in Table 2.1 and Table

27 a Faul Scenarios dependabiliy F1 Inernal Faul: Verify wheher he relay has successful deeced inernal fauls. F2 Exernal Faul: Verify wheher he relay has no ripped for fauls ouside he proeced zone. F2-1: fauls on ine 2 F2-2: fauls on ine 4 F3 One-End-Open Inernal Faul: Verify he abiliy of he relay o deec a faul wih no infeed. F4 Swich ono Faul: Verify he abiliy of he relay o deec a faul immediaely afer closing he line. F4-1: Bus-side Poenial Transformer F4-2: ine-side Poenial Transformer F5 Faul during Power Swing: Verify he abiliy of he relay o rip properly when a faul occurs during a power swing. F6 Inernal Faul during Flucuaion: Verify he capabiliy of relay o rip properly when he sysem frequency flucuaes wihin a normal range. F6-1: increases o 60.5Hz F6-2: decreases o 59.5Hz Table 2.1: Faul Scenarios for Conformance Tes Condiion Type ocaion [%] Incepion Angle [deg] Resisance [Ω] F1 F2-1 F2-2 F3 F4-1 F4-2 F5 AG, BC, BCG 0, 5, 10, 25 0, 50, 70, 90 0, 45, 90 ABC 0 AG, BCG 0, 10, 25 10, 50, 90 0, 45, 90 BC, ABC 0 AG, BC, BCG, 0, 50, 90 0, 45, 90 0 ABC AG, BCG 0, 5 0, 50, 90 0, 45, 90 BC, ABC 0 AG, BCG 0, 25 0, 50, 90 0, 45, 90 BC, ABC 0 AG, BCG 0, 25 0, 50, 90 0, 45, 90 BC, ABC 0 AG, BCG, BC, ABC 0, 50, 90 0, 45, 90 0 F6-1 F6-2 AG, BCG, BC, ABC AG, BCG, BC, ABC 50, 90 0, 45, , 90 0, 45, 90 0 For he bach simulaion, he faul poin should be wihin 10%-90% of he line lengh ATP limiaion for disribued line model. For he posiions wihin 0% and 100%, please noe ha he ATPdraw model is used. 7

28 b No-faul Scenarios securiy N1 ine Closing: Verify he abiliy of he relay no o rip when line closing occurs. N1-1: Bus-side Poenial Transformer N1-2: ine-side Poenial Transformer N2 oss of Poenial: Verify he abiliy of he relay no o rip in case of loss of phase volage inpus. N3 oss of oad: Verify he abiliy of he relay no o rip in case of loss of load. N4 Resoring he Poenial: Verify he abiliy of he relay no o rip in case of resoring he volage inpus. N5 Power Swing: Verify he abiliy of he relay no o rip during a power swing condiion. N6 oad Encroachmen: Veriy he capabiliy of he relay no o rip during heavy sload. Condiion SW saus Table 2.2: Non-faul Scenarios for Conformance Tes Objec N1 Open ine 1 breakers N2 N3 N4 Close Source S1, S2, S3 Open Bus 4 breaker Bus 2 breaker Close SW Open SW afer 2 cycles Close Source S1, S2, S3 Operaion Three phases close afer 2 cycles Phase A close afer 2 cycles Phase B, C close afer 2 cycles Remove S1, S2, S3 respecively afer 2 cycles Remove S2, S3 afer 2 cycles Remove S1, S2, S3 simulaneously afer 2 cycles Remove S1, hen S2 afer 2 cycles, hen S3 afer 2 cycles Open Bus 4 breaker afer 2 cycles Open Bus 2 breaker afer 2 cycles Resore S1, S2, S3 respecively afer 2 cycles Resore S2, S3 afer 2 cycles Resore S1, S2, S3 simulaneously afer 2 cycles Resore S1, hen S2 afer 2 cycles, hen S3 afer 2 cycles N5 Open ine 1 Power swing occurs afer hree-phase faul on ine 1 N6 ine 1 Increase he load on Bus 2 from normal o maximum For load encroachmen scenarios, he use of differen impedances measured by relays a Bus 1 represens he various level of load increase. According o he sysem parameers, he secondary impedances under normal load and maximum load are 57.4Ω and 7.86Ω respecively. Four condiions whose corresponding secondary impedances are 31.88Ω, 22.34Ω, 12.74Ω and 7.90Ω were seleced as he scenarios o execue he load encroachmen logic esing. Tes Scenarios for Compliance Tes The firs ask for he applicaion of compliance es is o selec hose possible scenarios which may cause relay uninended operaion. Two approaches named seady sae approach and dynamic sae approach are proposed o achieve his ask. 8

29 Seady sae approach: This approach uses he seady sae mehods o find some ransmission lines ha are designaed as vulnerable lines due o sressed operaing condiions. Those imporan lines mus have high securiy of he proecion scheme. For a given sysem, opology processing mehod [8] will find he lines, such as ie-lines, or single-connecion lines whose ouage will disconnec he generaor, load or even par of an area, parallel lines, long lines, ec. Power flow mehod is used o idenify ransmission lines which may have overload condiions and whose conneced buses may have low volage problems. Under such condiions, he apparen impedance seen by disance relays may fall ino heir backup proecion zones. They may rip he lines and rigger he cascading ouage. Dynamic sae approach: This approach sudies he proecive relays in dynamic condiions such as he case when afer he faul and is clearing, he power swing occurs, which may confuse some disance relays as he apparen impedance may fall ino he proecion zones. The relays may operae as no inended and cause condiions ha may resul in furher ripping. The dynamic apparen impedance phasors can be rerieved from he ime domain ransien sabiliy analysis and such waveforms are evaluaed o selec he power sysem scenarios of ineres for he applicaion relay ess [9]. For a large sysem, he number of relays ha need o be carefully evaluaed in deail will be grealy reduced by using his approach. By having a digial relay model embedded in EMTP/ATP [10] one can selec a group of scenarios ha could cause relay unwaned operaion. The EMTP/ATP relay model can be conneced o he ransmission line models from he lis creaed by he seady sae and dynamic sae sudy selecions. We can obain each relay acions hrough a se of coningency scenarios. If incorrec relay operaion is found, ha scenario will be recorded and saved ino he es case library, which will be used o validae behavior of physical relays. The following condiions were used for ess using he IEEE 14-bus sysem for he compliance es: [A1]: Single 3-phase faul wih criical clearing ime CCT a he base load condiion: o verify relay operaion blocking during sable power swing. [A2]: Single 3-phase faul wih criical clearing ime CCT a he increased load condiion: o verify relay operaion blocking during sable power swing and overload condiions. [A3]: Two successive 3-phase fauls, firs faul wih fixed clearing ime, second wih CCT, a he base load condiion: o verify relay operaion blocking during sable power swing. [A4]: Two successive 3-phase fauls, firs faul wih fixed clearing ime, second wih CCT, a he increased load condiion: o verify relay operaion blocking during sable power swing and overload. To see relay performance during ou of sep condiion boh a he base load and overload condiions, he following ess are performed: [A5]: Ou of sep: single 3-phase faul wih clearing ime larger han CCT a he base load condiion. 9

30 [A6]: Ou of sep: single 3-phase faul wih clearing ime larger han CCT a he increased load condiion. [A7]: Ou of sep: wo successive 3-phase fauls, firs faul wih fixed clearing ime, second wih clearing ime larger han CCT, a he base load condiion. [A8]: Ou of sep: wo successive 3-phase fauls, firs faul wih fixed clearing ime, second wih clearing ime larger han CCT, a he increased load condiion. [A1] [A2] [A3] [A4] [A5] [A6] [A7] [A8] Table 2.3: Tes scenarios for Compliance Tess Purpose Tes Sequence Tes Variaions Verify he relay abiliy no o rip a sable power swing Verify he relay abiliy no o rip a sable power swing and overload condiions Verify he relay abiliy no o rip a sable power swing Verify he relay abiliy no o rip a sable power swing and overload condiions To see he relay performance a ou of sep condiion To see he relay performance a ou of sep condiion To see he relay performance a ou of sep condiion To see he relay performance a ou of sep condiion 3-phase faul, wih criical clearing ime CCT, base load condiion 3-phase faul, wih criical clearing ime CCT, increased load condiion Two successive 3-phase fauls, firs wih fixed clearing ime, second wih CCT, base load condiion Two successive 3-phase fauls, firs wih fixed clearing ime, second wih CCT, increased load condiion 3-phase faul, wih clearing ime larger han CCT, base load condiion 3-phase faul, wih clearing ime larger han CCT, increased load condiion Two successive 3-phase fauls, firs wih fixed clearing ime, second wih clearing ime larger han CCT, base load condiion Two successive 3-phase fauls, firs wih fixed clearing ime, second wih clearing ime larger han CCT, increased load condiion Faul locaion: 10%, 50%, 90% Same as above Same as above Same as above Same as above Same as above Same as above Same as above The purpose of hose es scenarios is o sudy he influence of sable power swing, ou of sep and overload condiions on disance relays Tes Case ibrary For each of he relay ypes considered, a library of power sysem models and disurbance scenarios was creaed. As shown in Figure 2.8, he es scenarios generaed for he applicaion of conformance es and compliance es are placed ino he library. The abnormal power sysem operaing condiions and vulnerable ransmission lines which may cause relay uninended operaions can also be described and sored ino he library. The scenarios of ineres from digial faul recorder DFR records and blackou evens can be added o he library as well. The es case library can be used widely as reference es cases for relay performance evaluaion and rouble shooing. 10

31 Figure 2.8: Tes case library 2.3 Tes Implemenaion Two auomaic conversion programs are developed in C++. One is o conver from MAT file generaed hrough MATAB/ATP o RA file Relay Assisan sofware [11] file, whose inerface is shown in Figure 2.9. Anoher is o conver from ATP file o COMTRADE file [12], as shown in Figure Figure 2.9: Program inerface for convering MAT file o RA file Figure 2.10: Program inerface for convering ATP file o COMTRADE file 11

32 2.3.1 Tes Procedure The procedure of performing relay es is described as follows: 1 Generae es scenarios. Tes cases ATP, MAT or COMTRADE file are generaed hrough bach simulaion program ATP and MATAB and/or cases of ineres are seleced from digial faul recorder DFR files or blackou even files. 2 Conver Daa forma. The program developed in C++ is used o auomaically conver various formas of es cases o he forma which can be recognized by Relay Assisan sofware, such as COMTRADE. 3 Creae es session. The es session is creaed by loading seleced es cases wih Relay Assisan sofware. Each es session conains specific scenarios sored by differen ypes of disurbances or power sysem operaing condiions. For example, he faul session can be sored by faul ype, locaion, incepion angle and resisance. Figure 2.11 and Figure 2.12 show he loading process and he loaded waveforms displayed in Relay Assisan sofware user inerface. Figure 2.11: Example for loading es cases Figure 2.12: Example for waveforms displayed by Relay Assisan sofware 12

33 4 Se proecive relays. The relay seings group corresponding o a given ransmission line and proecion scheme is acivaed from he relay fron-panel buons or hrough he relay seing program. 5 Execue simulaion. The signal waveforms volage and curren are sen o he digial simulaor o generae he real volage and curren signals for relays. 6 Collec relay response or even repor. The relay responds o he inpu signals for each case and generaes an even repor conaining he deailed operaion informaion. The rip signals are capured by simulaor as oupu signals and used o auomaically calculae an operaion feaure such as ripping ime. The even repors are colleced by he file rerieval program for furher sudy. Figure 2.13 and Figure 2.14 show he relay response capured by Relay Assisan sofware and even repor conaining oscillography daa recorded as COMTRADE file. Figure 2.13: Example of es resul for inernal faul Figure 2.14: Example of even repor shown as oscillograph 13

34 Figure 2.15 gives he framework for sofware implemenaion. Figure 2.15: Sofware framework for relay esing By execuing he simulaion, he signal waveforms volage and curren are sen from compuer o he digial simulaor I/O box o generae he real volage and curren for he relays. The disance relay will respond o he waveforms and send rip signal o he simulaor digial conac inpus if a faul is deeced. Also he digial inpus can be sen o relays hrough simulaor digial oupus for cerain purpose, such as he rip circui breaker signal, ec. Then he rip signal and even repor are colleced o analyze he relay performance. Field recordings can also be replayed in Relay Assisan sofware o es relays aboraory Seup The laboraory seup is shown in Figure The major componens include a PC used o run relaed sofware programs, a digial simulaor used o generae real volage and curren signals and he physical relay under es. A commercial sofware program called Relay Assisan residing on he PC communicaes wih digial simulaor. I is capable of sending ransien volage and curren daa and receiving conac saus daa [11]. The digial simulaor applies he volage and curren waveforms o he relay and records he relay rip conac saus. A relay seing sofware program residing on he PC communicaes wih he relay o configure relay seings and an auomaed relay file rerieval sofware program residing on he PC communicaes o he relay o auomaically rerieve relay even repors riggered by cerain pre-se condiions. The connecions beween compuer, digial simulaor amplifiers and D/A converers, and disance relay are marked in Figure

35 The es environmen including he sofware and hardware as shown in Figure 2.17 describes he flow of relay es execuion. Figure 2.16: aboraory seup for relay esing Figure 2.17: Block diagram for relay es environmen 2.4 Tes Resuls The SE 321, SE 421 and GE D60 disance relays were seleced for he sudy. Table 2.4 liss heir funcions and sofware used for inerfacing wih compuer. Comparaive sudy was carried ou o evaluae hese relays performance. Table 2.4: Funcions and sofware for seleced disance relays Type Funcion Manual Sofware SE-421 High-speed ine Proecion, auomaion, and conrol sysem [13][14] SE-5030 [15] SE-321 Phase and ground disance relay [16] SE-5010 [17] GE-D60 Transmission line disance relay [18] UR Seup [19] 15

36 2.4.1 Power Sysem Daa for Conformance Tes For he Conformance Tes, IEEE PSRC sysem is used. Relays are applied a Bus1 o proec ine 1. Figure 2.18 shows he one-line diagram of ransmission line model and relay s posiion. Table 2.5 liss he power sysem daa for his applicaion. Figure 2.18: One-line diagram of he ransmission line model for Conformance Tes Table 2.5: Power sysem daa for Conformance Tes Parameer Value Nominal sysem line-o-line volage 230kV Nominal relay curren 5A secondary Nominal frequency 60Hz ine 1 lengh 45miles ine 1 impedances: Z 11, Z primary, primary Zero-sequence muual coupling: Z 0M primary Source Bus1 impedances: Z 1B1, Z 0B primary, primary ine 4 impedances: Z 41, Z primary, primary PTR poenial ransformer raio 230kV: 100V = 2300 CTR curren ransformer raio 2kA: 5A = 400 Raio of CTR o PTR: k Phase roaion ABC In order o calculae relay seings, he power sysem impedance should be convered from primary o secondary using raio k. Table 2.6 liss he corresponding secondary impedances. Table 2.6: Secondary impedances for Conformance Tes Parameer Value ine 1 impedances: Z 11, Z secondary, secondary Zero-sequence muual coupling: Z 0M secondary Source Bus1 impedances: Z 1B1, Z 0B secondary, secondary ine 4 impedances: Z 41, Z secondary, secondary 16

37 2.4.2 Power Sysem Daa for Compliance Tes For he Compliance Tes, IEEE 14-Bus sysem is used. Relays are applied a Bus2 o proec he ransmission line beween Bus2 and Bus3. Figure 2.19 gives he one-line diagram of ransmission line model and relay s posiion. Table 2.7 liss he power sysem daa for his applicaion. Figure 2.19: One-line diagram of he ransmission line model for Compliance Tes Table 2.7: Power sysem daa for Compliance Tes Parameer Value Nominal sysem line-o-line volage 138kV Nominal relay curren 5A secondary Nominal frequency 60Hz ine lengh 33miles ine 1 impedances: Z 11, Z 01 PTR poenial ransformer raio 138kV: 100V = 1380 CTR curren ransformer raio 500A: 5A = 100 Raio of CTR o PTR: k Phase roaion ABC primary, primary secondary, secondary Disance Relay Seing To fully es he relay funcionaliy and operaing characerisic hree zone proecion schemes are applied in hree seleced relays. These applicaions are for a single circui breaker, hree-pole ripping cases. Some funcions, such as power swing, load encroachmen, ec, are applied as well o calibrae he relay performance during paricular power sysem operaing condiions. The applied funcions are described as follows: Three zones of phase mho and ground mho disance proecion Zone 1 forward-looking, insananeous under reaching proecion, covers 80% of he proeced line. Zone 2 forward-looking, ime-delayed rip covers 100% of he proeced line. Zone 3 forward-looking, ime-delayed rip covers 100% of he proeced line, backup proecion for he adjacen downsream line. 17

38 Swich Ono Faul SOTF proecion, fas ripping when he circui breaker closes This iem is no applicable for Compliance Tes. Power swing, Ou-of-sep logic, prevens uninended ripping when power swing occurs. oad encroachmen logic, prevens uninended ripping during increasing load condiions. Table 2.8 gives a brief summary of he funcions applied for conformance es and compliance es. The seing ables for esed relays are given in Appendix A.1. They provide he crucial seing values for boh conformance es and compliance es so ha he es can be repeaed. Table 2.8: Funcions able applied for es Funcion/Elemen Conformance Tes Compliance Tes Faul ocaion Enabled Enabled Phase Disance Mho 3-Zone Mho 3-Zone Ground Disance Mho 3-Zone Mho 3-Zone Swich Ono Faul Enabled Disabled Power Swing/ Ou of Sep Enabled Enabled oad Encroachmen Enabled Enabled Tes Resuls and Analysis According o he number of es repeiions for each es scenario, he es approach can be divided ino wo classes: random ess and saisical ess. For he random ess, each es case is applied only once, and he relay responses rip or no rip, rip zone and rip ime are recorded. For he saisical ess, some ineresing scenarios are seleced and repeaed 30 imes, and he relay responses of each es are recorded o calculae he rip or no rip rae, maximum rip ime, minimum rip ime, mean rip ime and deviaion of rip ime if he relays rip. The comparaive resuls among he hree relays are sudied as well. The complee es resuls for all he es iems lised in secion Tes Scenarios Generaion as shown in Table 2.1, 2.2 and 2.3 for he hree seleced disance relays can be found in Appendix A.2. Two examples o describe how he saisical es resuls would look like are given here. One example obained by execuing conformance es on SE-421 relay is given in Table 2.9. In his example, differen es cases were simulaed for differen ype of fauls, locaions, and incepion angles. Each es is repeaed 30 imes, and saisical mehods are used for deermining operaing ime for he esed relay. One can noice very ineresing resuls wih respec o differences in operaing imes for differen faul condiions as well as differences beween maximal and minimal values of operaing ime for he same faul condiion. 18

39 Table 2.9: Example of saisical es resuls Type oc [%] α [deg] Trip Zone No.T MeanT [ms] MaxT [ms] MinT [ms] Devn [ms] AG 50 0 I AG I AG II BC 50 0 I BC I BC II BCG 50 0 I BCG I BCG II ABC 50 0 I ABC I ABC II Anoher example obained by applying conformance es is given in Figure The figure depics a comparaive analysis of rip ime vs. faul incidence locaion for hree disance relays. Trip ime shown in his figure is obained saisically afer several ess cases are repeaed. Relays are se o operae in zone 1 coving 80% of he line. An ineresing oucome is ha he rip ime, for some relays, becomes much longer han expeced. Figure 2.20: Example of comparaive es resuls In general, he hree relays operaed correcly bu some excepions sill exis as shown in Table 2.10, which summarizes heir performance. Table 2.10: Summary for relays performance Tes Iems SE-421 SE-321 GE D60 F1 Wrong rip Zone Z1 Wrong rip Zone Z1 Wrong rip Zone Z1 Z2, delayed rip ime. Z2, delayed rip ime. Z2, delayed rip ime. F2 Zone 3 failed o rip Zone 3 failed o rip Zone 3 failed o rip F3-F6 Correc Correc Correc N1-N6 Correc Correc Correc A1-A7 Correc Correc Correc 19

40 For he unexpeced operaions ha occurred in scenarios F1, deails are given in Appendix A.2 as shown in Table A.3, A.15 and A.27. I appears ha he increased faul resisance caused he incorrec operaions. Even SE-321 failed o rip when he resisance increased o 10Ω. All hese incorrec operaions occurred because of he use of Mho characerisic disance elemen. The siuaion can be improved by applying he Quadrilaeral characerisic for he disance elemen. Zone 3 relay failed o rip as a backup proecion for he adjacen downsream line under he condiion of Phase A o ground faul a he 90% locaion of he adjacen line. Resuls are given in Table A.5, A.17 and A.29. Pracically zone 3 proecion elemens are se o rip breakers afer cerain ime delay from he ime he phase disance elemen or ground disance elemen picks up so ha coordinaion wih zone 1 and zone 2 relays is achieved. In his applicaion, he ime delay was se o commonly used 60 cycles. From he Figure 2.21 capured from SE-421 relay even, we can clearly see ha ground disance elemen Z3G picked up insananeously afer single phase-o-ground faul occurred. Then a power swing developed followed, which appeared as faul ype ransiion o he relay and caused phase disance elemen M3P o pick up. However, he apparen impedance seen by he relay changed due o he impac of developed power swing. Thus, he boh wo proecion elemens Z3G and M3P could no pickup and hold he pickup sae for he prese ime delay used o coordinae wih zone 1 and zone 2 relays, which finally resuled in failing o rip he breakers. Special proecion scheme should be applied o improve he relaying sysem for his condiion. This es case is added o he es case library as an ineresing condiion used o evaluae relay dependabiliy feaure. Figure 2.21: Three-phase volage and curren waveforms from relay even There are some cases ha relays operaed during non-fauled condiions, such as power swing condiions lised in conformance es scenarios N5 as shown in Table 2.2 and compliance es scenarios A1-A4 as shown in Table 2.3, and load encroachmen condiions lised in conformance es scenarios N6 as shown in Table 2.2. These 20

41 uninended relay operaions were improved by uilizing ou-of-sep funcion and load encroachmen funcion respecively, which are residing in relays as proecion elemens. Moreover, proper seings are essenial for relaying scheme o solve he corresponding siuaions. Figure 2.22 and Figure 2.23 presen hese wo funcions and heir parameers. Seings for he cases esed in he projec can be found in Appendix A.1. These cases discussed above are relaed o he relay securiy characerisics, which can also be added o he es case library. Figure 2.22: Ou-of-sep funcion and parameers 21

42 Figure 2.23: oad encroachmen funcion and parameers These resuls provide imporan informaion which was no documened in he relay manuals, and which definiely may affec proper coordinaion and applicaion performance of he relaying schemes. The conformance es resuls evaluae he operaing feaures and indicae ha oher funcion elemens should be applied o improve he dependabiliy of he relaying scheme. The compliance es resuls indicae ha he zone 3 relays operaed incorrecly during some unusual power sysem operaing condiions. Thus, some special schemes should be carried ou o improve hese siuaions. 2.5 Fuure Work The uninended operaion of proecive relays may cause cascades when power sysem operaes in abnormal condiions such as increasing heavy loads followed by muliple line rips. Appropriae relay es can help evaluae relay performance, calibrae relay seings and figure ou he vulnerable condiions ap o causing relay unwaned operaion. The proposed es mehodology includes he issues how o model power sysem used o generae disurbances and sudy specific condiions, how o selec and generae es scenarios, and how o execue relay es in efficien way. An idea of forming a common model o be used in indusry for simulaing ransmission line disurbances and fauls, including cascading evens, as well as forming a es case library for relay users so ha he es scenarios can be used repeaedly as a reference when evaluaing or purchasing relays should be pursued in he fuure. 22

43 3.0 Par II: Generaor Relay Tes GaTech 3.1 Overview This documen describes a comprehensive es plaform for ransien esing of proecive relays and is applicaion o generaor relay esing. Comprehensive ransien esing of generaor relays requires esing for a variey of evens ha should exercise all he funcions of a modern generaor relay. Mos of he work has been focused on defining he evens for which he generaor relays should be esed. We refer o hese as comprehensive se of generaor ransien evens. The implemenaion of esing of generaor relays agains hese evens requires an engine for creaing he ransien daa for hese evens and a plaform o feed he ransien daa ino a generaor relay. For his purpose a new, physically-based generaor model was developed wo axes model wih access o generaor windings for faul creaion in he windings. The model has been useful in creaing he ransien daa for he defined evens. In addiion, a plaform for feeding he ransien daa ino a sysem of amplifiers ha creae he acual volages and currens o be fed ino he generaor relay has been developed. The overall projec approach is illusraed in Figure 3.1. Training on relay seup Begin relay esing waveform playback Obain new laboraory space Ordering pars for scale model Assembly Assembly of scale model Seup Seup experimens permanenly Begin relay esing Tes scale model Tes he scale model Developmen of generaor model Figure 3.1: Overall projec approach This par of he repor is organized as follows: Secion 3.2 describes he plaform for esing generaor relays. Secion 3.3 describes he generaor relay esing procedure. Secion 3.4 presens esing resuls and observaions. Secion 3.5 provides a number of observaions and conclusions. There are several appendices supporing his repor. Appendix B.1 provides a ypical modern generaor relay connecion. Appendix B.2 provides a comprehensive lis of evens for esing generaor relays. Appendix B.3 provides he high fideliy generaor model ha is par of he high fideliy simulaor. The simulaor is used o creae he evens 23

44 lised in Appendix B.2. Appendix B.4 provides example generaor relay responses o specific evens. Finally, Appendix B.5 provides he srucure of he COMTRADE file. The COMTRADE sandard is used for exchange of esing cases. 3.2 Descripion of Plaform This secion describes he plaform for generaor relay esing. The plaform consiss of sofware and hardware ha can creae he appropriae signals a he correc volage levels for inpuing ino he relays. The secion provides an overview of boh he sofware and he hardware. The acual relays ha are esed are a he Beckwih M3425-A, and b he SE 300G generaor relays. These relays were donaed by Beckwih and SE. A descripion of he relays is provided followed by he esing hardware and sofware Generaor Proecion Relays Beckwih M3425-A One of he relays used in his sudy is he Beckwih M3425-A generaor proecion relay, illusraed in Figure 3.2. I was kindly donaed by is manufacurer, whose suppor is hence grealy appreciaed. The sofware o communicae wih he relay hrough a serial or Eherne connecion is provided by he manufacurer. The provided sofware can be used o define relay parameers and proecion scheme and rerieve recorded waveform daa. Many aspecs of he operaion of his relay are described in a comprehensive insrucion manual [20]. Figure 3.2: The Beckwih M3425-A generaor proecion relay SE 300-G The second relay used in his sudy is he SE 300-G illusraed in Figure 3.3. I has also been donaed by Schweizer Engineering aboraories SE, and heir suppor has been insrumenal in he seup of he laboraory. Figure 3.3: The SE 300-G generaor proecion relay 24

45 Overview of he Tesing Plaform A flexible esing plaform has been developed o perform proecive relay esing. The overall esing plaform is illusraed in Figure 3.4. The esing plaform can reproduce he volage and curren waveforms seen in he field by proecive relays. Such simulaed signals are hen sen o he differen inpus of he esed relay. Specifically, hese signals are generaed by simulaing he power sysem under sudy on a compuer. An even is simulaed in he sysem, and volage and curren waveforms a he locaion of he VT and CT ha are conneced o he relay are recorded. Wih proper scaling facors, hese waveforms replicae he oupus of volage ransformers VTs or curren ransformers CTs. The recorded compuer waveforms are hen ransformed ino elecrical signals using a D/A converer and an amplifier. Finally, each oupu of he D/A convereramplifier block is equivalen o he oupu of a VT or a CT, and is sen o he proper inpu of he relay under es. The consiuen pars of he esing plaform of Figure 3.4 are described in deail in secions and Hos Compuer High-Fideliy Simulaor Comprehensive Generaor Model PCI Board/Waveform Generaor Signal Amplifier Theaer Signal Amplifier Transformer Bank Sep-up Transformers Generaor Relay Figure 3.4: Overview of he esing plaform Sofware The firs componen of he esing plaform is he high-fideliy simulaion sofware program WinIGS-T. This compuer program models he power sysem more accuraely han mos of he oher exising approaches. Precise models of power sysem componens allow an accurae simulaion of a variey of evens encounered by proecive relays. The abiliy o simulae hese evens in a laboraory seup provides a benchmark for he robusness of he parameers enered in he considered generaor proecion schemes. In his secion a concise descripion of he sofware is provided. Elsewhere in he repor his sofware is uilized o develop ransien waveforms for esing generaor relays. The daa are sored in COMTRADE forma and herefore can be used by oher esing devices High-Fideliy Modeling and Simulaion The simulaion sofware is based on full hree-phase models of power sysem componens ha are described in erms of heir physical parameers. An example is provided in Figure 3.5. Models include ransmission lines, ransformers, circui breakers, and synchronous generaors. All connecions beween he differen models are explicily represened wih bus-bar links and swichgear. This approach uses acual device parameers insead of 25

46 equivalen sequence parameers. The simulaor accuraely simulaes he dynamics of he models by using he quadraic numerical inegraion mehod, which is more precise compared o oher mehods commonly used in power sysem analysis. The quadraic inegraion mehod has significan advanages in erms of numerical properies and can hence provide realisic resuls for a complee range of generaor, ransformer, and ransmission line evens ha can be esed in a laboraory seing. a b Figure 3.5: Sample model definiions in he high-fideliy simulaor sofware: a ransmission line physical design parameers and b hree-phase subsaion bus connecions Comprehensive Generaor Model The focus of his par of he projec is on esing generaor relays. I is herefore desirable o capure as many characerisics common o all of he generaors as possible o simulae a response of he generaor ha is as close o field observaions as possible. To achieve he highes accuracy possible, he sofware plaform includes a full ime domain, ransien, wo-axis synchronous generaor model. This model is described in greaer deail in Appendix B.3. Figure 3.6 illusraes he user inerface for his model. The generaor model is physically based, wih explici represenaion of he acual saor and roor physical windings. Fauls can herefore be simulaed anywhere in he coils of he generaor, wihou any model complicaions and herefore his model can be used o sudy and es he effeciveness of 100 % saor proecion schemes. Harmonics generaed by he acual, non-sinusoidal winding layou are also accuraely simulaed, resuling in a complee and accurae generaor represenaion. 26

47 a b Figure 3.6: Comprehensive generaor model of he sofware plaform: a parameer definiion window and b visual represenaion of he connecion poins Waveform Generaion from Scenario Analysis The simulaion sofware has he capabiliy o oupu he compued simulaed waveforms in he IEEE COMTRADE forma [21]. The COMTRADE forma is commonly used beween equipmen manufacurers o exchange ransien waveform daa. The main characerisics of COMTRADE files are compacness and ease of implemenaion. Once recorded, he waveforms can be ransferred o a compuer for offline analysis and furher processing. A descripion of he IEEE COMTRADE forma for he purpose of relay esing can be found in Appendix B.5. Wihin he simulaion sofware here is oupu processing sofware ha can presen he daa in various forms, for example, he rms value of a cerain waveform, he frequency of a cerain waveform, he phase angle of a cerain waveform, ec. In oher words, he COMTRADE file may conain a number of ime domain waveforms of physical quaniies, such as volages and currens, also and compued waveforms such as phasors, frequency, ec. Figure 3.7 illusraes hese capabiliies. Using he above procedures a number of evens have been simulaed and he resuls sored in COMTRADE forma. These evens will be described in subsequen secions. The acual COMTRADE files have been uploaded in a web sie for use by persons involved in his projec. The link o his web sie is: hp:// 27

48 Simulaed even WinIGS es case models Quasisaic ime-domain simulaion phasor info Creaion of full imedomain waveforms Generaed full ime-domain waveform To COMTRADE file To arbirary waveform generaor inerface Figure 3.7: Simulaion oupu in various forms sored in a COMPTRADE file Summary The implemenaion of he sofware porion of he relay esing plaform is summarized in Figure

49 Simulaed Even WinIGS Tes Case Models Simulaed Waveforms COMTRADE File Wih Simulaed Waveforms Hos Compuer Figure 3.8: Summary of he sofware porion of he esing plaform Tes Bench The second componen of he esing plaform is a es bench o reproduce condiions in an acual power sysem ha he esed relays have o proec Waveform Generaor Afer simulaing he power sysem, waveforms of ineres are recorded and played back o differen ypes of relays. A compuer-conrolled waveform generaor is uilized o generae volage and curren signals from he sored COMTRADE files. A D/A converer and signal generaor from Naional Insrumens ranslaes waveform daa o analog signals. The 10 V generaed signals are amplified using heaer sound equipmen. The 30 V oupus of he sound equipmen audio amplifiers are hen sepped up o sandard relay volages and currens 69 V or 115 V and 1 A or 5 A hrough a ransformer bank. The final signals are direcly sen o he relays, as hey replicae he oupus of he VTs and CTs in an acual power sysem. The layou of he es bench is presened in Figure 3.9. A picure of he acual seup is shown in Figure COMTRADE File Wih Simulaed Waveforms PCI Board Waveform Generaor Emulaed CT/PT oupu Transformer Bank Sep-up Transformers 10 V 115 V, 5A Trip! Hos Compuer Signal Amplifier 30 V, 5A Generaor Relay Figure 3.9: Tes bench layou 29

50 Figure 3.10: Picure of he acual laboraory seup Scale Model In addiion o he previously described esing plaform, an alernaive esing mehod has been developed using a scaled power sysem model. The scale model represens a simplified power sysem consising of hree subsaions, a generaing subsaion, and wo ransmission subsaions. The scaling facor is 1000:1. Despie he scaling, he model includes all major elemens of an acual power sysem, including ransmission lines, a source behind a sep-up ransformer, circui breakers, disconnec swiches, as well as poenial and curren ransformers for insrumenaion. The source uilizes he oupu of he waveform generaor menioned above. The scale model is used o es ransien relay response under acual raher han simulaed condiions. Such condiions may also include imbalances inheren imbalances from ransmission line consrucion and asymmerical condiions one or wo phases disconneced. The model iself does no replicae he effec of he res of he sysem o he volage source, since here is no feedback loop beween he wo. The scale model is inegraed wih he res of he esing plaform as i is illusraed in Figure Consrucion deails of he scaled power sysem are described in [33], [34]. 30

51 Generaor Model Simulaed Waveforms Scale Model Wih All Insrumenaion PCI Board Waveform Generaor Scale Model Measuremens Trip! Hos Compuer Signal Amplifier Transformer Bank Generaor Relay Figure 3.11: Tes bench layou for relay esing wih a scale model Auxiliary Independen Volage and Curren Channels Relay esing uilizing eiher of he wo sysems described above requires he use of independen volage and curren channels o generae he inpu signals o he relays a he appropriae level. Specifically, he oupu of he scaled model of he power sysem or he oupu of he COMTRADE daa convered ino analog are in general low level signals. The independen volage and curren channels simply amplify hese signals o levels appropriae for inpu o he relays under es. The independen volage and curren channels are designed o accep he oupu of a he PC-conrolled D/A converer and he exising heaer amplifier or b he oupu of he scaled power sysem model. As an example, he schemaic of he volage channel is illusraed in Figure 3.12, and is acual layou is shown in Figure

52 Figure 3.12: Three-phase auxiliary volage channels for relay and PMU esing schemaic, includes neural volage 32

53 Figure 3.13: Three-phase auxiliary volage channels for relay and PMU esing acual layou The independen curren channels have a similar design. For generaor relay esing, a minimum of four independen volage channels and hree independen curren channels are required Virual Relay Tesing For compleeness we discuss he process of virual relay esing. The basic idea/objecive of he virual relay esing is o es he manufacurers relay sofware direcly wihou he need o generae acual volage and curren inpus o he relay. Thus, he principle of virual esing is o perform he ess using only he binary code of he relay firmware. Configuraion and waveform daa can be direcly sen o he inpus of he relay funcions, and he relay oupus can be processed on he hos compuer wih he benefis of specialized analysis sofware. As a resul, virual esing eliminaes he consrains of a hardware seup, including waveform generaion, wiring, and communicaions, and faciliaes he esing of he relays. Therefore, all he relay esing is performed on a hos compuer as i is illusraed in Figure There is one major obsacle o virual relay esing, however. Specifically, manufacurers mus be willing o provide he binaries of heir relay funcions, and his is currenly no he case. This approach also requires ha manufacurers documen he inerface parameers of heir algorihms before he ess can be performed. Possible applicaions of virual relay esing are described in [22] and [23]. 33

54 Simulaed Even WinIGS Tes Case Models Simulaed Waveforms Relay Firmware Funcion/ogic Trip! Hos Compuer Figure 3.14: Virual relay esing principle 3.3 Generaion Relay Tesing Seup This secion presens he generaor relay esing procedure. A es sysem has been creaed for esing generaor relays. The es sysem has been so seleced as o be able o creae he evens required for a comprehensive ransien generaor esing, i.e., cable of creaing all he ransien evens ha are perinen o generaor relay operaion Purpose Proecive relays are usually esed agains simplified generaor models ha do no accoun for variaions in he elecrical properies of he generaor. Many facors such as soil properies may vary wih he locaion and generaor and affec he grounding impedance. Relays wih idenical seings and proecing he same ype of generaors are expeced o respond idenically o a given power sysem even since he proecion schemes are digially implemened. In realiy, he responses may vary because he relays operae in slighly differen environmens. Also, he waveforms seen by a relay may be affeced by facors inernal or exernal o he sysem. More generally, i is necessary o ensure he seings enered by proecion engineers are consisen wih he inended proecion scheme. The purpose of he ess is o verify a consisen behavior of he relays regardless of he proeced generaor and o check ha he inended proecion schemes are robus agains he acual parameers of he sysem Even Simulaion and Tesing Procedure The esing procedure for he available generaor relays is as follows, and i can also be applied o oher ypes of proecive relays. A se of evens, including fauls, imbalances, over- and under-exciaion is applied o a es sysem, and he behavior of he differen relay funcions are checked agains he inended proecion scheme. The ess concern all individual funcions of he relay for each even, as some scenarios call cerain funcions o arge or rip and oher funcions o remain passive. The robusness of he applied proecion scheme is esed using differen varians of he es sysem, where differences consis in minor changes in he generaor parameers. For boh varians of he sysem, a 34

55 response ha is idenical o he inended proecion scheme is expeced Descripion of Tes Sysem #1 The firs es sysem is illusraed wih he following simplified nework model in Figure Generaor Breaker 1 2 G BUSG BUS1 BUS2BUS3 FARSRC G SECBUS1 IM A R Figure 3.15: Nework schemaic of es sysem #1 The es sysem above includes mos aspecs of he generaor configuraion. Specifically, he nework above consiss of a comprehensive generaor model behind a sep-up ransformer, conneced o an infinie source via a ransmission line. An equivalen load is presen a he infinie source side o represen load condiions in he res of he sysem. The grounding connecions of he generaor and he ransformer secondary are an inegral par of he es sysem. I is exremely imporan for all grounding aspecs o be accuraely modeled. Indeed, he magniude of ground faul currens is ied wih he impedance of he ground connecion. As a resul, explici, comprehensive models of he generaor and ransformer grounding are provided. The grounding scheme includes a low-raed resisor a he secondary of a grounding ransformer placed beween he generaor neural and he remoe earh. Moreover, solid and resisive grounding can be modeled for boh he generaor and he ransformer. Wih such a comprehensive model of generaor grounding, i is possible o simulae a wide range of condiions and submi he resuling volage and curren waveforms o he generaor proecion relay for esing Generaor Model The full ime ransien generaor model developed for his projec is described in Appendix B.3. This is a 800 MVA, 15 kv generaor operaing a 60 Hz. ine-o-neural volage: 8.66 kv 800 Nominal base curren: I Base = = 30.8 ka

56 Generaor and Transformer Grounding The es sysem provides an explici model for generaor and ransformer grounding. The grounding scheme in he es case is equivalen o he one illusraed in Figure Figure 3.16: Generaor and sep-up ransformer grounding scheme The grounding resisor is 1 Ω behind a 8.66 kv/240 V single-phase grounding ransformer ha sis beween he generaor neural and he earh Figure The lowside of he grounding ransformer is conneced o he same remoe earh. The neural of he high-side of he sep-up ransformer is also conneced o he same remoe earh hrough a grounding pah. Noe ha he grounding pah for he sep-up ransformer has a resisance of 5 Ω. Single Phase Transformer/Cenerapped Secondary Tile Transformer wih Secondary Cenerap Single Phase Accep Cancel Transformer Raing kva 15.0 Side 1 kv Raing 8.66 Side 2 kv Raing 0.24 Series Resisance pu Series Reacance pu Side 1 Terminal Names Side 2 BUS1_N SECBUS1_1 SECBUS1_NN BUS1_G 1 Circui Number SECBUS1_2 Figure 3.17: Seings for he generaor grounding ransformer 36

57 Sep-up Transformer Model Seings The ransformer is a dela-wye sep-up 600 MVA ransformer. ow-side volage is 15 kv, high-side ransmission volage is 230 kv Figure Phase Transformer Cancel Accep Transformer Two-Winding, Three-Phase Side 1 Bus C c Side 2 Bus BUS1A BUS kv a Dela Wye B A b Dela Wye a Phase Connecion Sandard 30 0 Alernae A kv Transformer Raing MVA Winding Resisance pu eakage Reacance pu Nominal Core oss pu Nominal Magneizing Curren pu Tap Seing pu 1.0 Minimum pu 1.0 Maximum pu 1.0 Number of Taps 1 Circui Number 1 WinIGS-T - Form: IGS_M104 - Copyrigh A. P. Meliopoulos Figure 3.18: Seings for he generaor sep-up ransformer Transmission ine Parameers The ransmission line is a 10-mile secion wih hree phase overhead conducors and shield wires. The line operaes a 230 kv Figure

58 3-Phase Overhead Transmission ine Accep Three-Phase Overhead Transmission ine Cancel Phase Conducors Type Size ACSR JOREE N1 28.0' N1 Shields/Neurals Type Size HS 5/16HS B1 A1 C1 Tower/Pole Srucure Name Type 101A Circui Number 1 JellowJacke 67.8 fee Tower/Pole Ground Impedance Ohms R = X = Ge From GIS ine engh miles ine Span engh miles Soil Resisiviy Ohm-Meers Bus Name, Side 1 BUS Circui Number 1 GA. Power H-Frame WoodPole TOWER Bus Name, Side 2 FARSRC Insulaed Shields Transposed Phases Transposed Shields Operaing Volage kv Insulaion evel kv FOW Fron of Wave BI Basic Insulaion evel AC AC Wihsand WinIGS-T - Form: IGS_M102 - Copyrigh A. P. Meliopoulos Figure 3.19: Seings for he ransmission line in he es sysem 38

59 Equivalen Source and Infinie Bus The source a he infinie bus has parameers illusraed in Figure 3.20 below. Three Phase Source Behind Impedance Accep Source Volage Equivalen Infinie Source ine o Neural kv Updae -N Cancel Bus Name FARSRC ine o ine Phase Angle 0.0 Posiive Phase Sequence Negaive Zero Circui Number 1 kv Degrees Updae - C B A C A B Source Impedance Posiive Sequence Ohms Resisance Reacance PU Base MVA kv- Negaive Sequence Resisance Reacance ka Ohms Zero Sequence Resisance Reacance Updae Ohms Updae PU WinIGS-T - Form: IGS_M110 - Copyrigh A. P. Meliopoulos Figure 3.20: Seings for he equivalen source a he infinie source Insrumenaion Channels Insrumen channels are se as follows: Generaor high-side PT: 15,000:120 base line-o-neural volage is 8.66 kv and 69 V for he high and low side respecively, and he generaor is moderaely grounded no solidly grounded Generaor neural PT: 4:1 measuremen aken from low-side of grounding ransformer, raio 240: Generaor CT, low-side and high-side sandard raio: 35,000:5 base curren is 30.8 ka Generaor neural CT: 240:5 measuremen from secondary of grounding ransformer, max curren 240 A PT/CT correcion facor: we assume insrumenaion channels are perfec. Therefore, all addiive correcion facors are se o zero, and all muliplicaive correcion facors 39

60 are se o uniy Descripion of Tes Sysem #2 The second es sysem is depiced in Figure The sysem consiss of hree generaors operaing a 15, 18, and 20 kv Figure 3.22, Figure 3.23, Figure 3.24, hree-phase and single-phase loads aached a he generaor buses, and ransmission lines connecing he generaors ogeher. There is an addiional load beween generaors 1 and 3. All generaors are behind sep-up ransformers ha bring he volage o 115 kv nominal. ike he firs es sysem, he neural of each generaor is conneced o a grounding ransformer ha carries a small resisance a is secondary. Meers capure he volage and curren phasors ou of he phase erminals of he generaor and he neural. The meers also capure he frequency, real and reacive power, and roor angle. The sysem is equipped wih faul logic models o simulae a range of sysem evens and monior he response of he relay. The main focus is on Generaor 1 o he lef in he figure. Waveforms wih faul evens are recorded for his generaor and played back in he esed relays. This sysem is he saring poin for a number of he ess described in his documen, and he sysem is modified o accommodae some of he scenarios simulaed. GNDRES3 P, Q, f, roorpos P/Q P/Q P/Q P/Q GEN3N VA, VB, VC, VN P, Q, f, roorpos P/Q P/Q P/Q P/Q VA, VB, VC, VN OAD3 P/Q P/Q 1Ph 1 2 GEN3H GEN3 P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q IA, IB, IC, IN P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q IA, IB, IC, IN P/Q P/Q P/Q P/Q P, Q, f, roorpos P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q INETERM1 INETERM GEN1 GEN1H GEN2H GEN2 VA, VB, VC, VN P/Q P/Q P/Q P/Q 1Ph P/Q P/Q P/Q P/Q GEN1N 1Ph GEN2N IA, IB, IC, IN P/Q P/Q P/Q P/Q P/Q P/Q P/Q P/Q GNDRES1 GNDRES2 Figure 3.21: Nework schemaic of es sysem #2 40

61 Generaing Uni 1 Synchronous Generaor Model Accep Cancel Machine Idenifier N/A Circui Number 1 Machine Saus In Service Ou of Service Power Bus GEN1 T E Conrols Nominal Volage kv 15.0 PV Conrol PQ Conrol Volage Sepoin pu 1.0 Slack Volage Regulaed Bus GEN1 Oher Parameers Per Uni Ineria Consan 2.5 Oupu Power Minimum Power Real Reacive Maximum Power MW MVAr Reacive Power Allocaion Facor 1.0 Source Impedance Ohms PU Posiive Sequence Negaive Sequence Zero Sequence Resisance Reacance Resisance Reacance Resisance Reacance Updae Ohms Updae PU WinIGS-Q - Form: IGS_M149 - Copyrigh A. P. Meliopoulos Base MVA kv ka Ohms Figure 3.22: Parameers for Generaor 1 Tes Sysem #2 41

62 Generaing Uni 2 Synchronous Generaor Model Accep Cancel Machine Idenifier N/A Circui Number 1 Machine Saus In Service Ou of Service Power Bus GEN2 T E Conrols Nominal Volage kv 18.0 PV Conrol PQ Conrol Volage Sepoin pu 1.0 Slack Volage Regulaed Bus GEN2 Oher Parameers Per Uni Ineria Consan 3.0 Oupu Power Minimum Power Real Reacive Maximum Power MW MVAr Reacive Power Allocaion Facor 1.0 Source Impedance Ohms PU Posiive Sequence Negaive Sequence Zero Sequence Resisance Reacance Resisance Reacance Resisance Reacance Updae Ohms Updae PU WinIGS-Q - Form: IGS_M149 - Copyrigh A. P. Meliopoulos Base MVA kv ka Ohms Figure 3.23: Parameers for Generaor 2 Tes Sysem #2 42

63 Generaing Uni 3 Synchronous Generaor Model Accep Cancel Machine Idenifier N/A Circui Number 1 Machine Saus In Service Ou of Service Power Bus GEN3 T E Conrols Nominal Volage kv 20.0 PV Conrol PQ Conrol Volage Sepoin pu 1.0 Slack Volage Regulaed Bus GEN3 Oher Parameers Per Uni Ineria Consan 2.5 Oupu Power Minimum Power Real Reacive Maximum Power MW MVAr Reacive Power Allocaion Facor 1.0 Source Impedance Ohms PU Posiive Sequence Negaive Sequence Zero Sequence Resisance Reacance Resisance Reacance Resisance Reacance Updae Ohms Updae PU WinIGS-Q - Form: IGS_M149 - Copyrigh A. P. Meliopoulos Base MVA kv ka Ohms Figure 3.24: Parameers for Generaor 3 Tes Sysem # Beckwih Relay Seup Connecions and Wiring A schemaic of he connecions beween he relay and he generaor are shown in Figure B.1 in Appendix B.1. This schemaic is mosly derived from he connecions diagrams in he relay manual [20], and i would apply in he case of an acual generaor proeced by he Beckwih relay. In his sudy, however, he generaor insrumenaion is simulaed wih sofware and reproduced using a signal generaor and amplifier. The connecions and wiring beween he signal amplifier and he generaor relay are shown in Figure B.6 in Appendix B Communicaions The relay can be configured from a remoe compuer using serial communicaions hrough a null-modem cable. Alernaively, he relay can be configured using an Eherne connecion or even he fron panel. The focus of his documen is on serial communicaions o ake advanage of he manufacurer-provided configuraion sofware. The Beckwih M3425-A relay is buil o lisen o requess sen in a specific forma known 43

64 as he BECO communicaions proocol see [24] and [25] for more informaion. As a resul, he relay does no check he presence of a clien or provide a promp inerface wih common erminal programs. When he relay receives a well-formed reques, i responds by consrucing a formaed message ha he clien can rerieve. Response messages may include requesed daa from he clien such as se poins, oupu saus, and full oscillograph records. While i is possible o implemen his proocol independenly, he provided configuraion sofware is a beer choice for an iniial approach. The sofware makes he proocol requiremens ransparen o he user by ranslaing basic seup informaion ino a reques wih he appropriae forma Seup Sofware The provided sofware consiss of wo separae compuer programs: IPSCOM and IPSuiliy. IPSCOM is a general program o seup, monior, and rerieve relay configuraion and saus daa. I can ranslae and display he informaion rerieved from he relay in a number of exual and graphical ways. While mos parameers of he relay can be se via IPSCOM, he program does no provide manual conrol of he oupu conacs. IPSuiliy is a lighweigh program ha can perform a limied number of operaions. While very limied compared o IPSCOM, IPSuiliy can ake conrol of he oupu conacs. A procedure o seup he relay and rerieve informaion is described in he following subsecions General seings Esablishing communicaions wih he relay when i is conneced o a compuer hrough a null-modem cable is very simple using he provided sofware: - The communicaions por is se o COM1. - The baud rae can be se o he highes suppored value 9600 bps. - Oher fields are lef unouched. General seings perain o he calibraion of he relay and he availabiliy of a number of feaures. The IPSCOM program provides he user inerface o define he general seings Figure

65 Figure 3.25: General relay seings dialog box The user inerface for general relay seings also provides inpus for he following parameers: - nominal curren and volage, CT and PT configuraion and raios, according o he generaor nominal volage and curren, and he insrumenaion channel parameers defined in secion 3.3.3; - sae of he inpus: closed or open; - seal-in ime for each of he channels. Nominal CT secondary curren and base frequency canno be changed as hey are buil ino he relay a he ime of purchase Individual funcions Individual funcions are configured using a specific screen for each funcion. The inrinsics of each funcion is described in he insrucions book of he relay. The IPSCOM program also provides displays ha summarize he parameers and saus of all funcions 45

66 in he relay GPS Synchronizaion The esed relays can boh synchronize heir clocks o a signal provided from a GPS anenna. Boh relays have an IRIG-B inpu ha enables his capabiliy Simulaion of Power Sysem Evens The high-fideliy simulaion sofware allows fauls o be placed anywhere in he sysem. Fauls can be simulaed anywhere along ransmission lines, circui connecors, as well as inside he windings of generaor models and ransformers. A number of evens can be simulaed wih his model. Firs are fauls ouside of he generaor and ouside he proecion range of he generaor relay. These evens include fauls a he ransformer or along he ransmission line. For hese evens, he generaor relay should no perform any acion, unless a cerain amoun of ime has elapsed. Second are evens inside he proecion range of he relay and evens ha concern he generaor iself. These evens include ground fauls inside he generaor, urn-o-urn fauls along he saor of he generaor, roor fauls, and exciaion failures. The es simulaes hese generaor evens, and he response of he relays o such evens is noed. The simulaed evens can be reproduced using WinIGS and following he suggesed procedure provided in Appendix B Reporing Tess and Simulaed Evens The ess are repored in a relay response char ha conains a comprehensive se of evens used o es he response of each of he funcions of he relay. Two copies of he documen are needed for each series of ess: one for he es iself and one for he inended response of he relay for each of he evens lised. We provide a procedure o reproduce each even in he es sysem using he WinIGS sofware in Appendix B.2. An example response char of he relay for each of he evens simulaed is given in Appendix B Basic Even Triggering and Oscillographic Record Analysis Beckwih M3425-A The Beckwih relay has he capabiliy o record and sore waveforms for a ime window ha covers an even in he sysem. 16 oscillograph channels are available o record and sore various evens. The waveform daa can be downloaded a a laer ime o a remoe compuer using he IPSCOM program or any sofware ha is compaible wih he communicaion proocols of he relay Manual Even Triggering and Rerieval I is possible o rigger he oscillograph and record measuremens even if he relay does no deec any even in he sysem. The following was performed for his basic operaion: 46

67 - Generae hree-phase, balanced sinusoidal volages, RMS value 69 V, in posiive sequence 120 degrees apar using he waveform generaor. - Feed he volage measuremen inpus V A, V B, V C wih he generaed volages pins 38 o 43 a he back of he relay. - Close oupu conac number 8 using he IPSuiliy. Wih he defaul seings of he relay, closing an oupu conac using IPSuiliy riggers he oscillograph and records a arge hi. Upon conac closing, visual feedback is provided by he relay wih he arge ED, he oscillograph rigger ED, and he oupu conac ED illuminaed. In addiion, he relay flashes arge informaion on he fron panel display Figure Targe and oscillograph rigger ED are li. Screen flashes arge informaion. Oupu ED is li. Figure 3.26: Manual oupu conac conrol wih IPSuiliy and visual feedback from he relay The records generaed from his even can be viewed and/or erased using he IPSCOM program. Specifically, he oscillographic daa can be downloaded in wo formas: COMTRADE or he proprieary Beco forma Figure For compaibiliy, files are downloaded in he COMTRADE forma. Noe ha he relay can sore up o 16 oscillograph records. The maximum number of records available can be configured from he relay fron panel or he configuraion sofware. Figure 3.27: The oscillograph rerieval screen 47

68 Oscillographic Record Analysis The waveform records comply wih he 1999 version of he COMTRADE sandard. The record consiss of hree files: a ex configuraion file, a daa file in he binary forma specified in he sandard, and a header file. The configuraion file is shown in Appendix E. Twelve analog measuremen channels include he hree line volages, he six phase currens on each side of he coils, he neural volage and curren, and one muli-purpose volage. The file has provisions for 40 saus channels, bu only 14 channels are uilized o record he sae of he 8 oupus and 6 inpus. Having fed he relay wih generaed waveforms, he downloaded COMTRADE daa can be visualized o check consisency of he relay seup. The IGS-XFM program [22] and he Waveform Analyzers sofware [23], boh developed a Georgia Tech for he sudy of proecive relaying, can be used o display he daa. Figure 3.28 shows an excerp of he generaed volages seen by he relay. In addiion, he figure includes a plo of he saus of he oupu conac ha has been oggled. By defaul, he lengh of he rerieved oscillographic record is 4.6 seconds. There are 4480 daa poins sampled a 960 Hz 16 imes he base sysem frequency. Records may conain up o 472 cycles 7.8 seconds duraion a his fixed sampling rae. As expeced, he recorded values for he volage waveforms are abou 97 V peak, and a RMS value ha is sabilized beween and V. The phases follow he posiive sequence order. The plo for Oupu 8 shows is saus changing from zero open o one closed. Addiionally, if he volage inpus sar less han 4 seconds before he oupu conac is riggered, he oscillograph is able o record he connecion of he relay o he volage source Figure

69 Inpu Volages V 100 V A V B V C Oupu conac riggered VA + V VB V VC V OUTPUT µs Figure 3.28: Graphical sample of he waveforms and sae of he oupu conac Scaled up recorded by he Beckwih relay. For readabiliy, daa poins are shown for phase A volage and oupu conac saus only 49

70 Inpu Volages V V A V B V C VA V VB V VC V µs Figure 3.29: Records of he iniiaion of he volage supply o he relay SE 300-G The seup for he 300-G relay is performed hrough he SE Acseleraor QuickSe sofware [15]. The sysem seings applied for he M3425-A relay are enered in he 300- G relay as well. The 300-G relay has he capabiliy o record and sore waveforms for ime windows of differen lenghs o cover a variey of evens in he sysem. The number of oscillograph records available is only limied by memory. The waveform daa can be downloaded a a laer ime o a remoe compuer using he Acseleraor program. The records generaed from his even can be viewed and/or erased using he provided configuraion sofware. Oscillograph daa is downloaded in a proprieary even file CEV forma before i can be convered o COMTRADE by he viewing sofware Even Triggering, Rerieval, and Analysis I is possible o manually rigger he oscillograph and record measuremens even if he relay does no deec any even in he sysem. This can be done direcly using he configuraion sofware. The waveform records comply wih he 1991 version of he COMTRADE sandard. The record consiss of hree files: a ex configuraion file, a daa file in ex forma, and a header file. The file conains 12 analog measuremen channels: phase, neural, and 50

71 ground currens, phase and neural volages, power supply volage, one muli-purpose volage, and a record of he frequency. The file has provisions for over 400 saus channels. The sae of each funcion, oupu, and relaed variables is recorded in he file as well. The downloaded COMTRADE daa can be visualized using he IGS-XFM program and he Waveform Analyzers sofware. By defaul, he lengh of he rerieved oscillographic record is 256 daa samples a 960 Hz 16 imes he base sysem frequency, oal 266 ms. Analysis of even waveforms is described in he es case showing he same waveform sen o boh relays for comparison. 51

72 3.5 Equaions for he Proecion Variables This secion presens he basic proecion variables in generaing uni proecive relays. I defines he noaion and some basic ess o deermine he proper connecion o he inpus of he relay Noaion Variable Descripion Funcions Ĩ Nom Nominal CT secondary curren 5 A All Ĩ Pickup Pickup curren a CT secondary All Ĩ Op Operaing curren a CT secondary 87 Ĩ Resrain Resrain curren a CT secondary 87 Suffix _X Suffix _x Corresponding variable on phase X on highvolage side of generaor X is A, B, or C Corresponding variable on phase x on neural side of generaor x is a, b, or c All All Seup 1 Single Curren Source a Neural Side Only ~ ~ ~ ~ j120 ~ ~ j240 ~ jω Va = V ; Vb = Ve ; Vc = Ve ; V = 1.0e p.u. ~ ~ ~ ~ ~ ~ ~ j120 ~ ~ j240 ~ jω I A = I B = IC = 0; I a = I ; I b = Ie ; I c = Ie ; I = I e. This seup reproduces a urn-o-ground faul on all hree windings of he generaor simulaneously. In his case, he neural curren is equal o hree imes he zero sequence curren which is also zero wih all hree windings shored o he ground simulaneously. ~ ~ As a resul, he neural volage can be negleced, and I 0, V = 0. N = N ~ Va ~ I A Seup 2 Same Currens In and Ou ~ ~ ~ j120 ~ ~ j240 ~ jω = V ; Vb = Ve ; Vc = Ve ; V = 1.0e p.u. ~ ~ ~ ~ ~ j120 ~ ~ ~ j240 ~ = I = I ; I = I = Ie ; I = I = I e ; I = I e a B b C c No curren is los in any phase from he neural o he ransformer side of he generaor. This seup reproduces operaing condiions ha do no involve a faul or abnormal condiions wihin he generaor windings, and may also represen urn-o-urn faul condiions. jω. 52

73 3.5.4 Operaing Curren and Resrain Curren Operaing curren: ~ ~ ~ ~ ~ IOp _ X = CTC I X I x = I X I x assumes maching CTs and correcion facor CTC = 1 Resrain curren: ~ ~ ~ ~ ~ CTC I X + I x I X + I x I Resrain_X = = same assumpion Individual Proecion Funcion Tess M-3425A Common Procedures The M-3425A relay offers proecion engineers he possibiliy o use differen ses of parameers simulaneously for he same relay funcion. This is as if he same funcion was duplicaed wihin he relay. In paricular, each se of parameers includes a differen pickup level/zone and a differen ime delay. This is very useful for esing purposes as i is possible o compare differen seings in parallel or o compare differen seings wih he same reference. Applicable funcions, wih a leas wo available ses of parameers, are as follows: 21, 24 definie ime only, 27, 27TN, 32, 40, 49, 50, 50DT, 59, 59N, 59X, 64F, 81, 81A, 81R, 87, IPSogic Funcions no lised above have only one se of parameers and canno benefi from his common procedure: 24 inverse ime, 25, 46, 50N, 50BF, 50/27, 51N, 51V, 59D, 60F, 64B, 64S, 67N, 78, 87GD, BM, TC Tesing In one of he parameer ses used as a reference, he ime delay is se o he lowes value possible 1 cycle in mos cases, which is an almos insananeous oupu rigger, and a reference oupu channel is seleced. For each of he remaining parameer ses, he ime delay is se o values in incremens up o he maximum possible delay, and an oupu oher han he reference oupu is seleced. The pickup seing is he same in all he parameer ses. Because he differen parameer ses available can be enabled simulaneously, i is possible o compare he behavior of he funcion wih differen ime delays in a single run. Specifically, he oupu swiching imes deermine he accuracy of he ime delays in he relay Pickup Tesing In one of he parameer ses used as a reference, he pickup seing is se o he lowes value possible. For he remaining parameer ses, he pickup is se o values in incremens up o he maximum range possible. The ime delay in all he parameer ses is se o he minimum delay possible in mos cases 1 cycle for almos insananeous ripping. Because he differen parameer ses available can be enabled simulaneously, i is possible o compare he behavior of he funcion wih differen pickup seings in a single run. Specifically, he levels where he funcions rigger deermine he accuracy of 53

74 he pickup seing in he relay Funcion 87 Phase Differenial Descripion This funcion rips when he operaing curren I Op exceeds a value ha is a funcion of he resrain curren. For he Beckwih relay, we hink he equaion of he characerisic, in erm of RMS values, is max I Pickup, I Resrain slope if I Resrain < 2I Nom IOp = I Resrain 4 slope oherwise The funcion akes one slope parameer and one pickup parameer. The funcion acivaes designaed oupus afer a se delay. The characerisic is shown in Figure Seings for he IPSCOM program are shown in Figure Operaing curren A Trip Block Resrain curren A Figure 3.30: Characerisic of Funcion 87 funcion wih 0.3 A pickup and 10% slope Figure 3.31: Seings for he differenial relay funcion 54

75 Seup 1 Single Curren Source a Neural Side Only Assuming I < 2 I Nom, he relaionship beween operaing and resrain curren becomes ~ ~ ~ ~ ~ ~ ~ ~ I X + I x I x IOp _ X = I X I x = I x; I Resrain_X = = ; 2 2 ~ ~ Thus, I = 2I,and I = 2I. Op _ X Resrain_X Op _ X Resrain_X Minimum Pickup and Dropou evel Tes Seup 1 For he minimum pickup es, he slope coefficien mus be made passive, i.e. I Resrain slope < I Pickup for all I Resrain < 2I Nom The slope coefficien becomes passive does no affec he characerisic below 2 I Nom when I Pickup slope. 2I Nom Since he relay does no repor funcion pickup in any of is oupus, we reduce he delay from pickup o rigger o he minimum possible which is 1 cycle. The relay is se o rigger Oupu #2 afer one cycle. To es he funcion, we look in he recorded waveforms he ime when Oupu 2 changes from zero inacive o one acive. The funcion is esed firs wih each neural-side curren inpu energized individually while he ohers remain unenergized. Then, he funcion is esed wih all hree currens acive. The firs round of ess consis of applying a curren ramp from 0 o 1.0 p.u., increasing a 0.05 p.u./s, and o noe a which RMS value of he curren he relay picks up/riggers he funcion. For record purposes, he ime refers o he ime origin of he waveforms rerieved from he relay. Noe ha he relay does no provide oupus for funcion arges; herefore, arges are esimaed using he 1-cycle delay seing enered for his relay funcion. The resuls are shown in Table 3.1. Table 3.1: Funcion 87, rigger and arge imes and corresponding curren Run ID Phases Trigger ime µs Trigger RMS A Es. arge ime µs Es. arge RMS A for 1.00 A RMS 1011_ A _ B _ C _ All Remarks: The M3425-A has a curren accuracy of ±3 % or 0.1 A. The curren record is offse by approximaely 0.13 A, and his offse is observable when no curren is flowing 55

76 hrough he relay. Some errors may occur when rerieving oscillograph records corruped files, e.g. in _ In he second round of ess, a sinusoidal waveform in a riangular envelope is uilized o rigger he relay funcion and acivae Oupu #2. The objecive is o show ha he oupu conac here Oupu #2 is deacivaed when he RMS curren falls o a value below he pickup level defined for he relay. Seings are 0.80 A pickup, delay 1 cycle, slope 1 %, and waveform wih riangular envelope from 0 o 0.6 p.u. and 0.6 p.u o 0 a 0.1 p.u./s. Again, each phase is esed individually for differenial curren pickup. An example of recorded waveforms for he es on phase A is shown in Figure Trigger and dropou imes are shown in Table 3.2 and compared wih he imes compued RMS values reach he pickup seing. Ia A OUTPUT 2 Ia_RMS Figure 3.32: Funcion 87, phase differenial, waveforms for he combined pickup and dropoff ess Table 3.2: Funcion 87, rigger and drop-ou imes and corresponding currens Run ID Phases Trigger ime µs Trigger RMS A Drop-ou ime µs Dropou RMS A for 0.80 A RMS up for 0.80 A RMS down 1015_ A _ B _ C

77 Tes of Slope Seup 1 I is no possible o es he slope using seup 1 since IOp = 2 I Resrain = slope I Resrain, I Resrain < 2I Nom slope = 2, and he slope facor canno be greaer han Funcion 27 Phase Undervolage Descripion This funcion operaes afer a specified delay when he volage on cerain phases drops and remains below a specified value. General generaor seings: hese seings reproduce a volage sag a he bus of he generaor: Triangular signal, maximum = 64 V RMS, minimum = 57 V, Slope a 0.03 p.u./s. The enire even lass abou 5 seconds Pickup Tes Seup 1 The hree ses of parameers are defined wih volage hresholds: 62 V, 60 V, and 58 V. The ime delay is se o he minimum possible 1 cycle. The response of he relay is shown in Figure Resuls are shown in Table 3.3. VA V OUTPUT 1 OUTPUT 2 OUTPUT 3 Thres_62V Thres_60V Thres_58V VA_RMS Figure 3.33: Funcion 27, phase undervolage, waveforms from pickup and dropoff es 57

78 Table 3.3: Funcion 27, imes and volage levels for funcion rigger and release Se pickup level V Measured rigger level V Measured release level V V seles under pickup level Trigger ime V seles above pickup level Release ime Oupu 1 Oupu 2 Oupu Tes Seup 1 The hreshold for he funcion is se o 60 V for all parameer ses. The delays are 1 cycle, 10 cycles, and 20 cycles. The response of he relay is shown in Figure VA V OUTPUT 1 OUTPUT 2 OUTPUT 3 Thres_60V VA_RMS Figure 3.34: Funcion 27, phase undervolage, waveforms from ime delay es Pickup volage is V. Noe ha from daa, he relay wais unil he RMS volage on phase A drops below V for 4 sampling periods. The difference wih he previous experimen is ha all seings have he same dropoff delay of 0.23 s 13.8 cycles afer he RMS volage permanenly says above V. The dropoff volage is V. Resuls 58

79 are shown in Table 3.4. Table 3.4: Funcion 27, ime delay es resuls Se delay Measured delay w.r.. Operaed RMS volage V cycles Oupu 1 cycles Oupu 1 1 N/A Oupu Oupu To check for ime delay accuracy and check ha he relay does no operae below he ime delay, all delays are se a 4 seconds 240 cycles. Then, he riangular funcion is fed o he relay. As expeced, here was no operaion. Noe ha he general arge ED and he EDs in he arge pane are no illuminaed, unless an oupu conac is riggered. In paricular, he arge EDs remains unli volage drops below he hreshold and reurns above he hreshold before he imeou has elapsed. 3.7 Expanded Tes Scenarios This secion discusses a number of scenarios for esing generaor relays ha may involve he riggering of muliple proecive funcions of he relay. For each one of he composie scenarios, he appropriae daa files in COMTRADE forma have been generaed and briefly described in his secion. The acual daa can be found in he projec web sie Mock Generaor Acceleraion Tesed Funcions The following proecion funcions are included: 24 V/Hz, 27 undervolage, 50 overcurren, 51V inv. ime overcurren w/ volage resrain, 59 overvolage, 81 overfrequency. For each proecion funcion o be esed appropriae evens are generaed and sored in COMTRADE forma. These evens have been uploaded o he projec web sie Descripion This es scenario is mean o demonsrae he capabiliies of he laboraory seup described in secion 3.2. Using Tes Sysem #1, a faul wih an acceleraion of he roor of he generaor is simulaed by configuring he volage source o produce a volage and frequency ramp. For his experimen, he curren is kep in phase wih he volage by placing a symmeric resisive load beween he phases and he neural. Simulaed volages vary from 12.3 o 14.5 kv - 57 o 67 V a he relays. Currens in each phase change 59

80 from 10.6 o 12.3 ka 1.52 o 1.75 A a he relay. The frequency changes from 60 o 62 Hz. All he changes ake place in 3 seconds, while he waveforms are recorded hrough he relays. The objecive is o observe changes in he ransien response as he volages and currens gradually rise. Finally, he volage and curren supply is disconinued afer 3 seconds o simulae circui breaker operaion. Recording coninues beyond 3 seconds o capure lingering changes in he relay saus Example Resuls Waveforms were recorded for he M-3425A relay only. Oscillograph daa is rerieved from he relay in IEEE binary COMTRADE forma. The rerieved waveforms for he phase volages and oupus are shown in Figure In paricular, several changes in he differen oupu signals can be observed while he volage ramps up. V A Oupu 1 Oupu 2 Oupu 3 Oupu 4 Oupu 5 Figure 3.35: M-3425A relay rerieved waveforms for he proecion scenario The proecion seings and acual behavior are shown in Table 3.5 in relay merics. The proecion scheme responded as expeced despie simplificaions in he experimen. The frequency funcion failed o sar in his paricular rial, bu responded in anoher run wih slighly differen condiions. 60

81 Table 3.5: Seings for seleced proecion funcions Funcion Seing Oupu Expeced response Observed response 24 V/Hz cycles 3 No rip No rip 27 undervolage 58 V 1 cycle 1 Mus release Released VA = 58 V 50 overcurren 1.60 A 1 cycle 2 Mus rip Tripped IA = 1.68 A 51V inv. ime 1.60 A 2.0 dial 4 Check Tripped overcurren w/ volage resrain 59 overvolage 62 V 1 cycle 1 Mus rip Tripped 81 overfrequency 61 Hz 3 cycles 5 Mus rip No rip All oher funcions No rip No rip Three-Phase Faul wih Unsable Swings afer Clearance Tesed Funcions and Evens Funcions: 21 disance, 32 direcional power, 27 phase undervolage, 50 overcurren, 78 ou-of-sep, 81* frequency. Evens: 3-phase faul on ransmission line, generaor mooring, reacive power ransfer, sysem frequency increase/variaions, sysem insabiliy afer faul clearing Descripion Tes Sysem #2 is modified o include he faul logic o reproduce a hree-phase faul on one of he lines ha connecs Generaor 2 see Figure The faul lass 0.45 s and is cleared by opening he circui breakers a boh ends of he considered line. The ransien swings are observed afer he faul is cleared. The seings are seleced for he sysem o become unsable afer faul clearance. This example is derived from ransien sabiliy es sysems Generaed Waveforms and Relay Response Capures of volage and curren waveforms from Generaor 1 sen o he relays are shown in Figure Waveforms for Generaors 2 and 3 are similar. There are a few phenomena common o unsable generaor swings: volage swings increase in magniude, ransiions of he unis from generaing o mooring, and vice-versa, high faul currens, acceleraion of generaor frequency, and drifing of roor posiion. Considering he individual relay funcions, we expec he relays o rip for power reversal, frequency increases, roor posiion drifs, overcurrens, undervolages, and so on. I is also desirable 61

82 o obain a single response of he relays o his even, which is he riggering of he ouof-sep relay funcion 78 along wih an indicaor saing Unsable swing k Generaor1_VA_60Hz vols Generaor1_VB_60Hz vols Generaor1_VC_60Hz vols k m Generaor1_VN_60Hz vols u m k Generaor1_IA_60Hz vols Generaor1_IB_60Hz vols Generaor1_IC_60Hz vols m k m Generaor1_IN_60Hz vols u m M Generaor_1 Real_Power W Generaor_1 Reacive_Power VA M M Generaor_1 Roor_Posiion Deg m s Figure 3.36: Capure of he waveforms sen o he relays for Tes Sysem #2 62

83 3.7.3 Wide-Area Parial oad Shedding Funcions and Evens Tesed Funcions: 81* frequency, Evens: 10 % reducion of load P and Q, sudden parial loss of load, frequency increase/variaions Descripion Wide-area load shedding occur from he acion of uiliies or sysem operaors when load is in excess of he available power from he generaors in service. Noe ha differen eniies have called he populaion for volunary load cus such as urning off lighs, simulaneously, region or naion-wide, for a shor duraion five minues, from he order of 10 %. A simulaneous drop in load can lead o an undesired response from he sysem. Using Tes Sysem #2, 10 % of he hree-phase loads is shed afer 1 second of simulaion Figure The responses of he generaors are observed. oads o be shed circui breaker opens afer abou 1 second. These loads represen 10% of he remaining load. GNDRES3 GEN3N 1 2 OAD3 OAD13 1Ph GEN3H GEN GEN1 GEN1H OAD12 GEN2H GEN2 OAD11 GEN1N GEN2N GNDRES1 GNDRES2 Figure 3.37: Principle for wide area parial load shedding 63

84 Generaed Waveforms and Relay Response Averaged values are shown in Figure m u m u k k k k k k k k k k k k M M M M M M M M All neural volages All neural currens All generaor frequencies All roor generaor posiions Generaor_1 VA_Magniude V Generaor_2 VA_Magniude V Generaor_3 VA_Magniude V Generaor_1 IA_Magniude A Generaor_2 IA_Magniude A Generaor_3 IA_Magniude A All volage and curren phase angles Generaor_1 Real_Power W Generaor_2 Real_Power W Generaor_3 Real_Power W Generaor_3 Reacive_Power VA Generaor_1 Reacive_Power VA Generaor_2 Reacive_Power VA Figure 3.38: RMS values of elecric quaniies unil one second afer load drop Wih he load reduced by 10 %, a similar drop and sligh swing in curren, real and reacive power is observed. The volage a he generaor erminals increases by less han 64

85 1 % and remains a ha level. Noe ha he phase angles of all he elecric quaniies does no exhibi fas variaions compared o unsable swings, bu all angles do increase. Similarly, he frequency increases up o 60.5 Hz one second afer load has been cu. Roor angles seem o drif a a similar rae as well. In his siuaion, here is no overcurren, under/overvolage, or fas swings. Only he frequency increases, and he phenomenon may no be deeced unil afer a second. Afer a second, frequency relays are likely o pickup he frequency variaion and rip he generaors. Small swings of power and currens go undeeced. The expanded simulaion over 8 seconds is shown in Figure

86 8.710 k Generaor_1 VA_Magniude V k k k M Generaor_1 Real_Power W M M M Generaor_1 Roor_Posiion Deg Generaor_2 Roor_Posiion Deg Generaor_3 Roor_Posiion Deg Generaor_1 Hz Generaor_2 Hz Generaor_3 Hz Figure 3.39: Expanded simulaion shows coninuous increase of generaor frequency and roor slip 66

87 3.7.4 Inadveren Generaor Breaker Operaion Funcions and Evens Tesed Evens: Sudden loss of load Generaor 1, load exceeds generaing capaciy Generaors 2 and 3, frequency increases or decreases, saor overloads. Funcions: 81* frequency, 50 overcurren Descripion Using Tes Sysem #2, a circui breaker is placed on he high side of Transformer 1. The circui breaker is open afer one second of simulaion simulaion over 2 seconds Generaed Waveforms and Relay Response The responses of Generaors 1 o 3 are shown in Figure Noe ha RMS volages and currens are ploed. The firs hing o remark is ha Generaor 1 acceleraes dramaically o 70 Hz afer losing he load. Generaors 2 and 3 slow below 59 Hz in less han one second o compensae for he addiional burden. The evoluion of roor posiions and phase angles of all elecrical variables reflec he fas changes in frequency. Synchronism mus be achieved again before Generaor 1 can be reconneced. For generaor 1, he volage increases by 5 % upon loss of load. The volage decrease for Generaors 2 and 3 is abou 2 %, while heir curren rises by 45 % and 21 % respecively. The change in volage is no likely o be deeced by he relays; however, he relays will mos likely reac o he insananeous, significan increase in he saor curren. 67

88 Generaor_1 Hz Generaor_2 Hz Generaor_3 Hz Generaor_1 Roor_Posiion Deg Generaor_2 Roor_Posiion Deg Generaor_3 Roor_Posiion Deg k Generaor_1 VA_Magniude V k k Generaor_2 VA_Magniude V k k Generaor_3 VA_Magniude V k k Generaor_1 IA_Magniude A k Generaor_2 IA_Magniude A k k Generaor_3 IA_Magniude A k Figure 3.40: Waveforms capured for inadveren breaker operaion 68

89 3.7.5 Disconneced Phase Funcions and Evens Tesed Evens: disconneced phase, imbalances. Funcions: 49 negaive sequence, *N neural-relaed funcions, 27 undervolage, 46 phase curren balance, 47 phase volage balance, 50 overcurren, 59 overvolage Descripion Tes Sysem #2 is modified o include per-phase breakers, so phases can be disconneced independenly of each oher. The swiches/breakers are locaed a he high-side of he sep-up ransformer for Generaor 1 and a he load conneced o ha generaor see Figure The firs experimens involve opening one and wo phases of he generaor breaker. The second ype of experimens deals wih opening one and wo phases of he hree-phase load conneced o Generaor GEN1 GEN1CB GEN1H OAD1 GEN1N 1Ph GNDRES1 Figure 3.41: Generaor 1 and load wih per-phase circui breakers Generaed Waveforms and Relay Response The response of he sysem is observed in Figure 3.42, Figure 3.43, Figure 3.44 and Figure Relays should rigger on imbalance and frequency increase/swings. arge volage and curren swings should be deeced wih insananeous hreshold funcions. 69

90 Generaor_1 Hz Generaor_2 Hz Generaor_3 Hz Generaor_1 Roor_Posiion Deg Generaor_2 Roor_Posiion Deg Generaor_3 Roor_Posiion Deg k k k k k k k k k k k Generaor_1 VA_Magniude V Generaor_1 VB_Magniude V Generaor_1 VC_Magniude V Generaor_2 VA_Magniude V Generaor_2 VB_Magniude V Generaor_2 VC_Magniude V Generaor_3 VA_Magniude V Generaor_3 VB_Magniude V Generaor_3 VC_Magniude V Generaor_1 IA_Magniude A Generaor_1 IB_Magniude A Generaor_1 IC_Magniude A Generaor_2 IA_Magniude A Generaor_2 IB_Magniude A Generaor_2 IC_Magniude A Generaor_3 IA_Magniude A Generaor_3 IB_Magniude A Generaor_3 IC_Magniude A k Figure 3.42: Response of he sysem afer opening phase A of Generaor 1 70

91 Generaor_1 Hz Generaor_2 Hz Generaor_3 Hz Generaor_1 Roor_Posiion Deg Generaor_2 Roor_Posiion Deg Generaor_3 Roor_Posiion Deg k k k k k k k k k k Generaor_1 VA_Magniude V Generaor_1 VB_Magniude V Generaor_1 VC_Magniude V Generaor_2 VA_Magniude V Generaor_2 VB_Magniude V Generaor_2 VC_Magniude V Generaor_3 VA_Magniude V Generaor_3 VB_Magniude V Generaor_3 VC_Magniude V Generaor_1 IA_Magniude A Generaor_1 IB_Magniude A Generaor_1 IC_Magniude A Generaor_2 IA_Magniude A Generaor_2 IB_Magniude A Generaor_2 IC_Magniude A Generaor_3 IA_Magniude A Generaor_3 IB_Magniude A Generaor_3 IC_Magniude A k Figure 3.43: Response of he sysem afer opening phases B and C of Generaor 1 71

92 Generaor_1 Hz Generaor_2 Hz Generaor_3 Hz Generaor_1 Roor_Posiion Deg Generaor_2 Roor_Posiion Deg Generaor_3 Roor_Posiion Deg k k k k k k k k k k k Generaor_1 VA_Magniude V Generaor_1 VB_Magniude V Generaor_1 VC_Magniude V Generaor_2 VA_Magniude V Generaor_2 VB_Magniude V Generaor_2 VC_Magniude V Generaor_3 VA_Magniude V Generaor_3 VB_Magniude V Generaor_3 VC_Magniude V Generaor_1 IA_Magniude A Generaor_1 IB_Magniude A Generaor_1 IC_Magniude A Generaor_2 IA_Magniude A Generaor_2 IB_Magniude A Generaor_2 IC_Magniude A Generaor_3 IA_Magniude A Generaor_3 IB_Magniude A Generaor_3 IC_Magniude A k Figure 3.44: Response of he sysem afer opening phase A of oad 1 72

93 Generaor_1 Hz Generaor_2 Hz Generaor_3 Hz Generaor_1 Roor_Posiion Deg Generaor_2 Roor_Posiion Deg Generaor_3 Roor_Posiion Deg k k k k k k k k k k k Generaor_1 VA_Magniude V Generaor_1 VB_Magniude V Generaor_1 VC_Magniude V Generaor_2 VA_Magniude V Generaor_2 VB_Magniude V Generaor_2 VC_Magniude V Generaor_3 VA_Magniude V Generaor_3 VB_Magniude V Generaor_3 VC_Magniude V Generaor_1 IA_Magniude A Generaor_1 IB_Magniude A Generaor_1 IC_Magniude A Generaor_2 IA_Magniude A Generaor_2 IB_Magniude A Generaor_2 IC_Magniude A Generaor_3 IA_Magniude A Generaor_3 IB_Magniude A Generaor_3 IC_Magniude A k Figure 3.45: Response of he sysem afer opening phases B and C of oad 1 73

94 3.7.6 Three-Phase Faul followed by Generaor Breaker Operaion Tes on Boh Relays Funcions and Evens Tesed Evens: hree-phase faul, all phases disconneced. Funcions: 21 disance no implemened, 27 undervolage, 50 overcurren no implemened Descripion Tes Sysem #2 is modified o include he faul logic o reproduce a hree-phase faul a he sep-up ransformer of he generaor. The faul sars a 00:47: and lass 80 ms 5 cycles before i is cleared by opening he generaor circui breaker Comparison of Generaed Waveforms and Relay- Recorded Waveforms The volages seen a he generaor erminals as sen o he relays are shown in Figure The volage waveforms are sen o boh he M-3425A and he 300-G relays for comparison. Alhough curren is also simulaed in he es case, he curren waveform generaor converer is no complee, and funcions 21 and 50 in paricular canno be esed. Overall, his es case reflecs he closes seup of he desired laboraory seup for relay esing, where simulaed waveforms are fed ino muliple relays. The M-3425A relay repors he volages seen a is erminals whereas he 300-G relay repors he volages as hey should be a he PT based on inpu volages. As a resul, he rerieved volage records for boh relays are pu ogeher so ha hey fi on he same volage scale and visible wihin he same display window. The superimposed relay records for he volage on phase A are shown in Figure As expeced, he volage recordings of he relays and he simulaed volages are consisen. The sauses for he phase undervolage funcion 27 for boh relays are ploed in Figure The 300-G relay shows consisency in he riggering of he funcion. When applying repeaed loops of he volage waveforms shown, regular swiching can be heard. Surprisingly, i is no quie he case for he M-3425A relay. Firs, he relay seems o miss one ou of every hree fauls. Even wih 1 cycle rigger delay, he M-3425A does no rigger is assigned oupu Oupu 1 unil afer 7 cycles. In Figure 3.48, he dropoff from he previous faul is visible. 74

95 48.79 k V_Phase_A kv k k V_Phase_B kv k k V_Phase_C kv k I_Phase_A ka I_Phase_B ka I_Phase_C ka Figure 3.46: Response on a hree-phase faul followed by opening of generaor breaker 75

96 VA SimVA kv BwVA V SelVA kv µs Figure 3.47: Superimposed relay measuremens of phase A volage afer display scaling ime shifing VA kv VA V VAkV 27P1-50 VA V OUTPUT µs a b µs Figure 3.48: Phase undervolage oupu from a he 300-G and b he M-3425A relay 76

97 3.8 Fuure Work This projec resuled in a library of evens for ransien esing of several key relay funcions. Considering he fac ha many manufacurers of relays are moving in he direcion of incorporaing phasor measuremen capabiliies ino he relays, i is a naural exension o apply hese developed mehods for he esing of hese new relays and in paricular he funcions ha depend on GPS synchronizaion. For example, generaor relays wih GPS synchronizaion provide an improved proecion funcion agains unsable generaor swings. The performance of his funcion is dependen upon he GPS synchronizaion accuracy. New ransien esing procedures can be developed for hese ypes of relays as an exension of he mehodologies discussed in his secion. 77

98 4.0 Par III: oad Shedding Relay Tes WSU 4.1 Inroducion Background Generaion and demand mus be coninuously balanced in an ac sysem. During balanced condiions, frequency is consan, a is nominal value. The nominal frequency in Norh America is 60 Hz, while some counries mainain he frequency a 50 Hz. Deviaions from nominal frequency occur when generaion and load are unbalanced. The frequency increases when generaion is greaer han load and decreases when generaion is less han load. herefore can effecively indicae a balanced condiion of generaion and load. Wih coninuously changing load, generaing unis auomaically adjus heir oupu o follow load for small frequency deviaions. Auomaic generaion conrol AGC operaes o resore frequency back o he nominal value. A significan imbalance beween generaion and load, however, can exceed he AGC sysem s abiliy, causing he power sysem o fail, and hose failures may cascade across a large par of he inerconneced sysem Under-frequency oad Shedding UFS Relay Inroducion If insufficien generaion is available on he sysem o mainain sabiliy, non-criical loads can be removed shed from he sysem o resore a balanced condiion and preven sysem failure. Such mehods of auomaic load shedding are designed as a las resor o preven a major sysem ouage [1]. UFS relays are used o deec overload condiions by sensing low sysem frequency and shedding enough load o rebalance generaion and load, and reesablish he nominal frequency. UFS relays are able o auomaically resore load afer frequency recovery. UFS is an effecive and reliable mehod ha helps o preven blackous. Each UFS relay may uilize a differen mehod of frequency measuremen based on is manufacurer and echnology. The following hree ypes of UFS relays are employed in power sysem proecion [26]: Elecromechanical relays Solid-sae saic relays Microprocessor digial relays UFS relays play an imporan role in he curren resrucured power sysem. The inerconneced nework expands he influence of UFS relays o proec he whole elecric power sysem. The final repor on he Augus 14, 2003 blackou in he U.S. and Canada concludes ha one of he hree principle reasons for he widespread blackou is 78

99 he relay proecion seings for he ransmission lines, generaors and under-frequency load shedding in he norheas may no be enirely appropriae and are cerainly no coordinaed and inegraed o reduce he likelihood and consequences of a cascade-nor were hey inended o do so [1] [27]. The conribuions of UFS relays o he blackou are reviewed in secion UFS Tess The imporance of UFS relaying in prevening cascading ouages warrans furher esing beyond he sandard accepance ess specified by manufacurers. A new es proocol o mee hese needs was developed for his projec and is presened in secion 4.4 of his repor. The ess are specified in wo pars, conformance and applicaion ess. The objecives of conformance ess are, similar o accepance ess, o es he relay s funcion, verify is operaing characerisics and calibrae he relay s seings. Applicaion ess focus on how he relay performs during a specific even such as blackou or islanding. Daa for applicaion ess can be obained from simulaions or from recorders operaing during he even. Applicaion ess allow esing under realisic and relevan condiions UFS Research During planning for his projec, a number of issues regarding UFS relays were raised by PSERC indusrial members. These included: A ime delay, in addiion o he ime delay seing, has been observed in some UFS relays. This delay is invesigaed and quanified. delay seings are ofen given in cycles. Because frequency is changing during he operaion of a UFS relay, i is imporan o verify wheher his delay is based on nominal frequency or acual frequency. Validaion esing of UFS relays is usually done wih discree changes in frequency. Members waned ess performed wih coninuous frequency decay. Oher issues idenified during he course of he projec are: Because many UFS relays use zero-crossing as he mehod of calculaing frequency, disorion of he volage waveform may obscure he poin of he zero crossing and affec he relay s performance [26]. Volage magniude may also affec he operaion of UFS relays, and an undersanding of hese effecs is imporan o a relay user. Each relay s specificaions mus be referenced when specifying he volage levels for undervolage esing [26] [27]. The ess specified and he resuls presened in his repor go beyond hose usually performed using commercial UFS relay es sysems. Some commercial sysems are capable of presening waveforms conained in COMTRADE files o a relay under es, bu no sandard files exis for such esing [27]. Such ess are presened as a resul of his projec. Mos commercial UFS es sysems sill use pure sine waves for esing. A proocol for esing wih disored waveforms is presened here. 79

100 4.1.5 Repor Organizaion Secion 4.2 of his repor provides a summary of he 2003 Norh American blackou repor s findings on UFS operaion [1]. Secion 4.3 presens he UFS relay es sysem used a Wichia Sae, including sysem hardware and sofware. Secion 4.4 presens proocols for UFS conformance and applicaions ess. Tes resuls for wo commonlyused UFS relays are presened in Secion 4.5, wih complee daa shown in Appendix C.1. Tes inerpreaions are discussed in Secion 4.6, followed by conclusions and suggesions for furher work in Secion Review of UFS Relay Operaion during he 2003 Norh American Blackou Background This is a review of he blackou final repor [1] for references o load shedding relays. UFS and oher relay proecion seings are one of he hree principal reasons given for he blackou: Based on he invesigaion o dae, he invesigaion eam concludes ha he cascade spread beyond Ohio and caused such a wide spread blackou for hree principal reasons Third, he evidence colleced indicaes ha he relay proecion seings for he ransmission lines, generaors and under-frequency loadshedding in he norheas may no be enirely appropriae and are cerainly no coordinaed and inegraed o reduce he likelihood and consequences of a cascade nor were hey inended o do so. [1, p. 73] More specifically, regarding load shedding relays [1]: Auomaic load-shedding measures are designed ino he elecrical sysem o operae as a las resor, under he heory ha i is wise o shed some load in a conrolled fashion if i can foresall he loss of a grea deal of load o an unconrollable cause. Thus here are wo kinds of auomaic load-shedding insalled in Norh America under-volage load-shedding UVS, which sheds load o preven local area volage collapse, and under-frequency load shedding UFS, which is designed o rebalance load and generaion wihin an elecrical island once i has been creaed by a sysem disurbance. Auomaic under-volage load-shedding UVS responds direcly o volage condiions in a local area. UVS drops several hundred MW of load in pre- seleced blocks wihin urban load ceners, riggered in sages when local volage drops o a designaed level likely 89 o 92% or even higher wih a several second delay. The goal of a UVS scheme is o eliminae load in order o resore reacive power relaive o demand, o preven volage collapse and conain a volage problem wihin a local 80

101 area raher han allowing i o spread in geography and magniude. If he firs load-shed sep does no allow he sysem o rebalance, and volage coninues o deeriorae, hen he nex block of UVS is dropped. Use of UVS is no mandaory, bu is done a he opion of conrol area and/or reliabiliy council. UVS schemes and rigger poins should be designed o respec he local area s sysem vulnerabiliies, based on volage collapse sudies. As noed in Chaper 4, here is no UVS sysem in place wihin Cleveland and Akron; had such a scheme been implemened before Augus, 2003, shedding 1,500 MW of load in ha area before he loss of he Sammis-Sar line migh have prevened he cascade and blackou. Auomaic under-frequency load-shedding UFS is designed for use in exreme condiions o sabilize he balance beween generaion and load afer an elecrical island has been formed, dropping enough load o allow frequency o sabilize wihin he island. All synchronous generaors in Norh America are designed o operae a 60 cycles per second Herz and frequency reflecs how well load and generaion are balanced if here is more load han generaion a any momen, frequency drops below 60 Hz, and i rises above ha level if here is more generaion han load. By dropping load o mach available generaion wihin he island, UFS is a safey ne ha helps o preven he complee blackou of he island, which allows faser sysem resoraion aferward. UFS is no effecive if here is elecrical insabiliy or volage collapse wihin he island. The repor concludes ha UFS, bu no UVF, operaed during he cascading failures in aemps o sop he cascade. Bu he effecs of load shedding were no sufficien: I mus be emphasized ha he enire norheas sysem was experiencing large scale, dynamic oscillaions in his period. Even if he UFS and generaion had been perfecly balanced a any momen in ime, hese oscillaions would have made sabilizaion difficul and unlikely. [1. p. 92] The final repor divides he blackou ino seven phases. Mos of he UFS relays ha operaed did so during phases 6D and 7, he final phase. In phase 6D, Cleveland area load was disconneced by auomaic underfrequency load-shedding approximaely 1,300 MW, and anoher 434 MW of load was inerruped afer he generaion remaining wihin his ransmission island was ripped by under-frequency relays. This sudden load drop would conribue o he reverse power swing. [1, p.88] In phase 7 16:10:46 o 16:12 EDT, he large elecrical island in he norheas had less generaion han load, and was unsable wih large power surges and swings in frequency and volage. As a resul, many lines and generaors across he disurbance area ripped, breaking he area ino several elecrical islands. Generaion and load wihin hese smaller 81

102 islands was ofen unbalanced, leading o furher ripping of lines and generaing unis unil equilibrium was esablished in each island. [1, p.75] The repor s conclusion on UFS relay operaion was ha he relays operaed as se, bu he seings may no have been opimal for sysem proecion during cascading ouages: Proecive relay seings on ransmission lines operaed as hey were designed and se o behave on Augus 14. In some cases line relays did no rip in he pah of a power surge because he apparen impedance on he line was no low enough no because of he magniude of he curren, bu raher because volage on ha line was high enough ha he resuling impedance was adequae o avoid enering he relay s arge zone. Thus relaive volage levels across he norheas also affeced which areas blacked ou and which areas sayed on-line. Power swings and volage flucuaions caused by some iniial evens as seen on Augus 14 can cause oher lines o deec high currens and low volages ha appear o be fauls, even if fauls do no acually exis on hose oher lines. Proecive relay sysems work well o proec lines from damage and o isolae hem from he sysem under normal and abnormal sysem condiions. When power sysem operaing and design crieria are violaed because several ouages occur simulaneously, commonly used proecive relays ha measure low volage and high curren canno disinguish beween he currens and volages seen in a sysem cascade from hose caused by a faul. This leads o more and more lines being ripped, widening he blackou area. [1, p ] Auomaic load-shedding relay proecion mus avoid premaure ripping. I mus be coordinaed o reduce he likelihood of sysem breakup, and once break-up occurs, o maximize an island s chances for elecrical survival. [1, p. 92] The repor furher concludes ha UFS operaion while he sysem was sill experiencing dynamic condiions significanly reduced he beneficial effecs of UFS: Examinaion of he loads and generaion in he Easern New York island indicaes before 16:10:00 EDT, he area had been imporing elecriciy and had less generaion on-line han load. A 16:10:50 EDT, seconds afer he separaion along he Toal Eas inerface, he easern New York area had experienced significan load reducions due o under-frequency loadshedding Consolidaed Edison, which serves New York Ciy and surrounding areas, dropped over 40% of is load on auomaic UFS. Bu a his ime, he sysem was sill experiencing dynamic condiions as illusraed in Figure 6.26, frequency was falling, flows and volages were oscillaing, and power plans were ripping off-line. 82

103 Had here been a slow islanding siuaion and more generaion on-line, i migh have been possible for he Easern New York island o rebalance given is high level of UFS. Bu he available informaion indicaes ha evens happened so quickly and he power swings were so large ha rebalancing would have been unlikely, wih or wihou he norhern New Jersey and souhwes Connecicu loads hanging ono easern New York. This was furher complicaed because he high rae of change in volages a load buses reduced he acual levels of load shed by UFS relaive o he levels needed and expeced. [1, p. 98] The repor suggess ha fuure proecion sysems should allow more coordinaion among various ransmission and generaion relays: Proecive relays are designed o deec shor circuis and ac locally o isolae fauled power sysem equipmen from he sysem boh o proec he equipmen from damage and o proec he sysem from fauly equipmen. Relay sysems are applied wih redundancy in primary and backup modes. If one relay fails, anoher should deec he faul and rip appropriae circui breakers. Some backup relays have significan reach, such ha non-fauled line overloads or sable swings may be seen as fauls and cause he ripping of a line when i is no advanageous o do so. Proper coordinaion of he many relay devices in an inerconneced sysem is a significan challenge, requiring coninual review and revision. Some relays can preven resynchronizing, making resoraion more difficul. Sysem-wide conrols proec he inerconneced operaion raher han specific pieces of equipmen. Examples include conrolled islanding o miigae he severiy of an ineviable disurbance and under-volage or under-frequency load shedding. Failure o operae or misoperaion of one or more relays as an even developed was a common facor in several of he disurbances. UFS and UVS proecion schemes resuled from recommendaions made afer previous blackous [1, p. 109]. I appears ha load shedding relays operaed properly, according o heir seings, during he 2003 blackou. Bu such operaion was no adequae o mainain sysem sabiliy, and exising relays and proecion schemes could no be expeced o miigae such a fas-moving cascade. 4.3 UFS Relay Tes Sysem UFS Relay Tes Sysem Overview An exising relay es sysem [28] was upgraded and used for UFS esing a Wichia Sae. Figure 4.1 shows he configuraion of his sysem, and Figure 4.2 shows he acual lab seup. As shown in Figure 4.1, digial signals such as recorded waveforms, simulaed waveforms produced by an elecromagneic ransiens simulaion [29], and arbirary 83

104 programmed signals can be produced and played by a PC worksaion. The digial waveform is convered o analog by a high-resoluion D/A converer. Then he analog signal is sen o a power amplifier o obain he volage applied o he relay. This volage or curren is sen o boh he es relay and a daalogger. The es relay will respond o he volage and send a rip signal o he daalogger when he relay operaes. By analyzing he applied waveform and rip signal, relay performance can be evaluaed. If he es relay is equipped wih a communicaion por, he compuer can read informaion from he relay or modify he relay seings. Figure 4.1: Configuraion of UFS relay es sysem 84

105 Figure 4.2: UFS relay es sysem UFS Relay Tes Sysem Hardware The major componens of his UFS relay es sysem are a deskop compuer PC, digial-o-analog D/A converer, power amplifier, daalogger, and relay under es. The PC is used for producing digial es waveforms, performing resuls analysis, and modifying relays seings for relays wih a communicaion por. Applicaion sofware o generae and analyze waveforms, conrol relays, and perform simulaions, is insalled on his PC. A high-resoluion D/A converer is used for convering he digial signal produced by he PC ino an analog signal. The power amplifier is used for amplifying he analog signal for inpu o he relay. The characerisics of he power amplifiers available a Wichia Sae are shown in Table 4.1. Table 4.1: Characerisic of power amplifiers Three independen curren sources 12 A rms, 10 khz One curren source 80 A rms, 20 khz Three independen volage sources 130 V rms, 10 khz Single- or hree-phase volage source 6 kva, 120 V o 500 V, Full power o 1 khz, Disorion o 20 khz The daalogger is used for recording he signals from power amplifiers as well as he relay ripping signal from relay. Because he volage or curren received by he es relay is idenical o he one received by daalogger, relay performance can be evaluaed by comparing his volage or curren waveform and relay rip signal. 85

106 The relay under es can be an elecromechanical, solid sae, or microprocessor relay. The relay receives he amplified analog signal and rips according o is seing. The relay seing can be modified by he relay panel or by PC for relays wih a communicaion por Sofware Sofware is insalled on he compuer in order o produce he UFS relay es waveforms. As shown in Figure 4.3, he UFS relay es sysem can produce es waveforms from recorded signals, simulaed signals, and arbirary signals produced in sofware. Figure 4.3: UFS relay es sysem sofware This relay es sysem can evaluae he relay performance during a specific even, such as blackou or islanding. Daa of such specific evens come from recorders such as digial faul recorders ha were operaing during he evens, or from power sysem simulaion sofware. Arbirary waveforms sofware is used in his relay es sysem o produce specific waveforms such as pure sine waves, frequency ramping, harmonic disorion, and variable volage magniudes Under-frequency oad Shedding Relays This UFS relay es sysem can es he hree ypes of UFS relays which are available for applicaion in load shedding schemes. These hree ypes of UFS relays are elecromechanical relays, solid-sae saic relays, and digial microprocessor relays. In his projec, wo commonly-used digial UFS relays were provided by heir manufacurers for esing. The specificaions of each relay are shown in Table 4.2 and Table 4.3 respecively. 86

107 Sepoin delay Table 4.2: Relay 1 specificaions Range Hz Accuracy ±0.01 Hz Range cycles Accuracy 0.25 cycles or ±0.1% of seing Sepoin delay Table 4.3: Relay 2 specificaions Range Hz Accuracy ±0.01 Hz Range 3 cycles 990 seconds Accuracy ±1.0 cycle; ±2% of he seing or ±25ms, whichever is greaer 4.4 Under-frequency oad Shedding Relay Tes Scenarios In his projec, wo UFS relay es caegories, conformance ess and applicaion ess, have been designed and implemened. For boh es caegories, differen scenarios are performed o validae wo key seings of UFS relays: pickup frequency and ime delay Conformance Tes Conformance ess verify ha he UFS relay operaes wihin manufacurer s specificaions for various scenarios. Usually he relay s specificaion is given under he assumpion ha his relay is designed o operae wih pure, undisored waveforms. The relay s specificaion under disored waveforms is no usually available, bu his can be imporan o relay applicaion and is included in he es proocol for his projec. Tes waveforms include pure sine waves, frequency ramping, harmonic disorion, and varying volage magniudes Tes Waveforms Tes Waveform Descripion The es waveforms are classified ino he following caegories: Pure sinusoidal waveforms: The UFS relay es sysem generaes pure waveforms like hose used by manufacurers and uiliies in convenional accepance ess. ramping waveform df/d: Waveform signals wih a discree change in frequency are normally used o es UFS relays. However, he discree change canno represen real siuaions where he frequency decays more gradually and coninuously. The UFS relay es sysem allows a user o selec differen values of df/d, he frequency decay rae. Figure 4.4 shows four such values of df/d, 0.1 Hz/sec, 0.2 Hz/sec, 0.4 Hz/sec, and 0.6 Hz/sec. 87

108 Hz df/d = 0.1 Hz/sec df/d = 0.2 Hz/sec df/d = 0.4 Hz/sec df/d = 0.6 Hz/sec Seconds Figure 4.4: decay Harmonic waveform: Volages wih harmonic disorion may cause UFS relay misoperaion. The reason is ha harmonics can cause early, lae, or muliple zero crossings, which can affec he zero-crossing frequency measuremens sill used in some commercial UFS relays. Toal harmonic disorion THD is normally used for measuring harmonic disorion levels, and i is defined as follows [30]: THD = H2 + H H N H 1 where H 2, H 3 H N are he ampliudes of harmonics and H 1 is he ampliude of he fundamenal. The UFS relay es sysem allows a user o choose he values of H 1, H 2, H 3 H N o produce a specified THD. Figure 4.5 shows a combinaion of common harmonic volages 5 h, 7 h, 11 h, and 13 h Magniude Volages Seconds Figure 4.5: Volage wih 5 h, 7 h, 11 h, and 13 h harmonics 88

109 Variable volage magniude waveform: Volages magniudes may change significanly during frequency excursions. UFS relays commonly have an undervolage block funcion, which serves o block load shedding when volage o he relay is los, and o block operaion during faul condiions. Because volage can vary rapidly during cascading ouages, i is sill imporan o evaluae UFS performance under variable magniude volage. The UFS relay es sysem allows a user o specify differen volage magniudes. Figure 4.6 shows a volage waveform wih 6 cycles depressed. 150 Magniude Volages Seconds Figure 4.6: Variable volage magniude Tes Procedure The es procedure for conformance ess is as follows: Pickup frequency es: Tes he pickup frequency a varying pickup frequency seings and raes of change of frequency. The minimum and maximum pickup frequencies are specified by manufacurers. Pickup frequencies ha include he minimum and maximum, and several in beween, are seleced, wih emphasis on hose usually used in pracice. The rae of change of frequency df / d is varied from 0.1Hz/sec o 0.9 Hz/sec in 0.1 Hz/sec incremens. delay es: Pickup frequency ess are performed wih ime delay seings of 6, 16, 36, and 66 cycles, and acual ime delays are recorded. The ess are repeaed a he following specific pickup frequencies and ime delays: 59.3 Hz, 15 second delay 59.5 Hz, 30 second delay Under-volage frequency block es: The pickup frequency es is repeaed a 55.0 Hz, 57.0 Hz, and 59.0 Hz seings wih decay raes of 0.1 Hz/sec and 0.9Hz/sec, a 85%, and 115% volage. Harmonic disorion es: The pickup frequency es is repeaed a 55.0 Hz, 57.0 Hz, and 59.0 seings wih decay raes of 0.1 Hz/s and 0.9Hz/s, wih 5% 5 h harmonic 5% 11 h harmonic 89

110 a combinaion of he mos common harmonic volages: Vdisorion = 1/5 V5 + 1/7 V7 + 1/11 V11+ 1/13 V % 11 h + 5% 13 h harmonics Applicaion Tes Applicaion ess focus on how he UFS relay performs during a specific even, such as cascading blackou or islanding. Daa for applicaion ess come from simulaions of he evens or recorders operaing during he evens. The UFS relay es sysem can uilize hese recorded daa o es a relay. A simulaed 13-bus power sysem has also been developed a Wichia Sae Bus Sysem Descripion A ransien power sysem model has been adaped for applicaion ess. This sysem is an equivalen of a 13 bus sysem [31]. Figure 4.7 shows he single line diagram of he es sysem. The raing of he synchronous machine conneced o bus 3 is 200 MVA. An IEEE ype 1 Auomaic Volage Regulaor AVR is used o represen he exciaion conrol of he generaor, as shown in Figure 4.7. Par of he sysem is represened by is Thevenin equivalen, and bus 13 is he load bus. The ie line beween bus 1 and bus 7 can be designed as a single or double circui ransmission line. The complee model and daa for he 13 bus sysem are given in Appendix C.2. Figure 4.7: Single line diagram of 13-bus equivalen sysem The modeling of loads is a complicaed by he complexiy of aggregaed loads on he sysem. In order o simulae he effecs of load on sysem volage and frequency changes, he load a bus 13 is modeled by differen composiions of resisive and inducive loads differen power facors. Typically, power facor is varied from uniy o 0.6 lagging in incremens of Simulaion wihou UFS Scheme In his scenario, no UFS scheme is implemened on he 13 bus sysem. The single ie line beween bus 1 and bus 7 is opened a 1 second. Figure 4.8 shows he comparison of frequency responses of he generaor for differen composiions of load a bus

111 Simulaion resuls reveal ha he frequency decay rae increases as inducive load is increased. The volage a bus13, however, decreases as he inducive load is increased Hz P.F. = 1.0 P.F. = 0.9 P.F. = 0.8 P.F. = 0.7 P.F. = Seconds Figure 4.8: Generaor frequencies wihou UFS implemenaion Simulaion wih UFS Scheme In his scenario, a UFS scheme is implemened a bus 13. The seings of he UFS relay are shown in Table 4.4 [32]. The single ie line beween bus 1 and bus 7 is opened a 1 second. Figure 4.9 shows he comparison of frequency response of he generaor afer implemening a single-sep UFS scheme a bus 13 10% shedding a 59.3 Hz. Figure 4.10 shows he comparison of frequency responses of he generaor afer implemening a wo-sep UFS scheme a bus 13 10% shedding a 59.3 Hz and 10% shedding a 58.9 Hz. As shown in Figure 4.10, sysem frequencies based on differen composiions of load were recovered afer implemening he UFS scheme. The minimum frequency saddle poin in he curve, however, decreases as he inducive load is increased. For all differen composiions of load, 20% of he load has o be shed in 2 seps in order o recover he frequency. Table 4.5 shows he ime of load shedding for differen composiions of load. Table 4.5 reveals ha he higher he inducance load percenage, he earlier he UFS relay operaes. Table 4.4: Seings of UFS scheme Amoun of oad o be Dropped Minimum Sepoin 10% 59.3 Hz 10% 58.9 Hz 10% 58.5 Hz 91

112 Hz P.F.=1.0 P.F.=0.9 P.F.=0.8 P.F.=0.7 P.F.= s Seconds Figure 4.9: Generaor frequencies wih UFS implemenaion 1 sep Hz P.F. = 1.0 P.F. = 0.9 P.F. = 0.8 P.F. = 0.7 P.F. = Seconds Figure 4.10: Generaor frequencies wih UFS implemenaion 2 sep 92

113 Table 4.5: oad shedding ime oad Composiion for firs 10% for Second 10% oad Shedding oad Shedding P.F. = Second Second P.F. = Second Second P.F. = Second Second P.F. = Second Second P.F. = Second Second Tes Procedure The applicaion es procedure is: Apply differen inpu waveforms from he scenarios described in secion and o he UFS relay being esed. A a seing of 59.3 Hz and 58.5 Hz, he simulaed waveform wihou he UFS scheme Figure 4.8 is applied. A a seing of 58.9 Hz, he simulaed waveform wih he single-sep UFS scheme Figure 4.9 is applied. A a seing of 60.5 Hz, 61.0 Hz and 61.7 Hz, he simulaed waveform wih he wo-sep UFS scheme Figure 4.10 is used. Tes he pickup frequency a he following seings: - Underfrequency seings: 59.3 Hz, 58.9 Hz and 58.5 Hz. - Overfrequency seings: 60.5 Hz, 61.0 Hz and 61.7 Hz Record he acual pickup frequency o verify he relay operaion. 4.5 UFS Relay Tes Resuls The es mehodology presened in he previous secion for conformance and applicaion ess is applied o wo commonly used underfrequency load shedding relays, which were provided by heir manufacurers for use in he projec. One is saic relay and he oher one is digial relay Conformance Tess Conformance ess include pickup frequency and ime delay ess. Waveforms wih differen raes of frequency change, oal harmonic disorion THD, and variable volage magniudes are applied o he relays Pickup Tes The es resuls for pickup frequency esing are shown in Table Table 4.6 presens es resuls of relay 1 when esed a 100% inpu volage and 0% THD. Table 4.7 shows he es resuls of relay 1 esed a 100% inpu volage and 5% THD. Table 4.8 shows he resuls of relay 2 esed a 100% volage and 0% THD. Table 4.9 shows he resuls of relay 2 esed a 100% inpu volage and 5% THD. 93

114 Tes Tes resuls for ime delay ess are shown in Tables Table 4.10 and Table 4.11 show es resuls for relay 1 wih 100% inpu volage and 0% THD a differen raes of frequency change 0.1 and 0.9 Hz/sec. respecively. Table 4.12 and Table 4.13 show es resuls for relay 1 wih 100% inpu volage and 5% THD a 0.1 and 0.9 Hz/s respecively. Table 4.14 and Table 4.15 show he es resuls for relay 2 wih 100% inpu volage and 0% THD a 0.1 and 0.9 Hz/s respecively. Table 4.16 and Table 4.17 show he resuls for relay 2 wih 5% THD a 0.1 and 0.9 Hz/sec respecively Applicaion Tess Tes scenarios oulined in secion 4.4.3, represening realisic condiions, are simulaed and applied o he relays. The acual pickup frequencies are recorded. The applicaion es resuls are shown in Table 4.18 and Table For complee resuls, including esing a differen inpu volages 100%, 85% and 115% of nominal and differen raes of frequency change 0.1, 0.5 and 0.9 Hz/s, please refer o Appendix C.1. 94

115 Table 4.6: Acual pickup frequency in Hz 100% Volage, 0% THD, Relay 1 Sepoin Hz Rae of Change Hz/second

116 Sepoin Hz Table 4.7: Acual pickup frequency in Hz 100% Volage, 5% THD, Relay 1 Rae of Change Hz/s Sepoin Hz Rae of Change Hz/s

117 Sepoin Hz Table 4.8: Acual pickup frequency in Hz 100% Volage, 0% THD, Relay 2 Rae of Change Hz/s Sepoin Hz Rae of Change Hz/s

118 Sepoin Hz Table 4.9: Acual pickup frequency in Hz 100% Volage, 5% THD, Relay 2 Acual Pickup Hz Sepoin Hz Rae of Change Hz/s

119 Sepoin Hz Table 4.10: Acual ime delay 100% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay 1 Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual sec sec sec sec 99

120 Sepoin Hz Table 4.11: Acual ime delay 100% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay 1 Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual sec sec 100

121 Sepoin Hz Table 4.12: Acual ime delay 100% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay 1 Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual sec sec sec sec 101

122 Sepoin Hz Table 4.13: Acual ime delay 100% Volage, 5% THD, 0.9 Hz/sec Rae of Change, Relay 1 Tripped Hz Sepoin Acual Tripped Hz Sepoin Acual Tripped Sepoin Hz Acual Tripped Hz Sepoin Acual sec sec 102

123 Sepoin Hz Table 4.14: Acual ime delay 100% Volage, 0% THD, 0.1 Hz/sec Rae of Change, Relay 2 Acual Pickup Hz Sepoin Acual Acual Pickup Hz Sepoin Acual Acual Pickup Sepoin Hz Acual Acual Pickup Hz Sepoin Acual Sepoin Hz Table 4.15: Acual ime delay 100% Volage, 0% THD, 0.9 Hz/sec Rae of Change, Relay 2 Acual Pickup Hz Sepoin Acual Acual Pickup Hz Sepoin Acual Acual Pickup Sepoin Hz Acual Acual Pickup Hz Sepoin Acual

124 Sepoin Hz Table 4.16: Acual ime delay 100% Volage, 5% THD, 0.1 Hz/sec Rae of Change, Relay 2 Acual Pickup Hz Sepoin Acual Acual Pickup Hz Sepoin Acual Acual Pickup Sepoin Hz Acual Acual Pickup Hz Sepoin Acual Sepoin Hz Table 4.17: Acual ime delay 100% Volage, 5% THD, 0.9Hz/sec Rae of Change, Relay 2 Acual Pickup Hz Sepoin Acual Acual Pickup Hz Sepoin Acual Acual Pickup Sepoin Hz Acual Acual Pickup Hz Sepoin Acual

125 Power Facor Table 4.18: Applicaion es of relay 1 : 2 Cycles Sepoin Hz Acual Pickup Hz Tes no.1 Acual Pickup Hz Tes no.2 Acual Pickup Hz Tes no

126 Power Facor Table 4.19: Applicaion es of relay 2 : 3 Cycles Sepoin Hz Acual Pickup Hz Tes no.1 Acual Pickup Hz Tes no.2 Acual Pickup Hz Tes no

127 4.6 Inerpreaion of he Resuls Conformance Tess The wo relays operaed differenly under conformance ess. In some cases he relays operaed ouside heir specificaions. For relay 1, pickup frequencies deviaed from he sepoin, and he deviaion increased wih increasing frequency decay rae. For he same relay, ime delays were ouside specificaions for high decay raes and long ime delays. For relay 2, acual pickup frequencies deviaed from he sepoin and in some cases, were ou of specificaion. There is no rend in deviaion regarding frequency decay rae for his relay. delays were wihin specificaions excep a 0.9 Hz/sec rae of frequency change and long ime delay seings. Discussion wih uiliy users of hese relays, however, indicae ha he errors, while ouside specificaions, are sill very small, and are inconsequenial for he users Applicaion Tess The specific dynamic es cases are applied o he relays. The acual pickup frequencies are recorded. Boh relays operaed quie accuraely a over-frequency sepoins 60.5, 61.0 and 61.7 Hz. Some deviaions are observed a underfrequency seings 59.3, 58.9 and 58.5 Hz Error Analysis UFS esing requires very high accuracy in boh delay ime and frequency measuremens for accurae resuls. Because measured errors were very small, he es sysem was reevaluaed for is abiliy o discern such small variaions in ime and frequency. The accuracy of he relay es sysem depends upon he accuracy of each componen of he esing environmen, including he waveform generaors and he daalogging equipmen. Accuracy specificaions for he wo relays esed and he daalogger used are: Tes Relay 1 +/-0.01Hz, 0.25 cycle Tes Relay 2 +/-0.01Hz, 1 cycle Daalogger 100ppm or % of he sampling rae is obained by measuring ime a each zero crossing, calculaing he ime difference from he previous zero crossing, and invering o obain frequency. Once he rip frequency is reached, cycles are couned unil he ime delay is reached, a which ime he relay should acually rip. For his mehod, he accuracy of boh he daalogger and relay may conribue error o he resuls. In order o verify he error, he informaion from he daalogger is analyzed. Table 4.20 shows daa recorded during relay esing. The volage 0 column is he inpu volage o he relay, sepped down hrough a volage ransformer. The volage 1 column is he operaion of relay s oupu conac, which goes from approximaely zero o a posiive value when he conac closes. The frequency of each cycle is calculaed using inerpolaion o improve accuracy. The pickup frequency was se a 55Hz. delay was se a 6 cycles. 107

128 Table 4.20: Daa for pickup frequency es 55 Hz Sepoin Row Calculaed Volage 0 Volage Vol 1Vol Hz Remark Zero crossing coninue coninue coninue Above seing frequency coninue coninue coninue Below seing frequency coninue coninue coninue Relay pickup here coninue coninue coninue coninue coninue coninue coninue coninue coninue coninue coninue coninue Conac close here Source: NI VI ogger, Scan rae: second, Number of scans:

129 According o Table 4.20, he las cycle before he frequency decays o he se value of 55.0 Hz ends a row and By inerpolaing beween row and 51648, he zero crossing is esimaed a: = 51647* s s * = s The zero crossing one cycle before his is a: = 51467* s s * = s The period is: T = 1 2 = 0.018s can be calculaed as follows: 1 f = = Hz T A Hz, he relay does no rip because he frequency is sill above he rip frequency. The nex zero crossing is a rows and By inerpolaing beween Row and 51830, he zero crossing is esimaed a: = 51829* s s * = s The zero crossing one cycle before his is a row and which is 1. The period is: T = 3 1 = 0.018s can be calculaed as follows: 1 f = = Hz T The relay may rip here since he frequency is less han he seing. Possible ime error e in one cycle, based on daalogger specs, can be shown as follows: T 6 e= * s* = 1.819*10 s And he possible error in frequency calculaion is: 1 f = error Hz T + e = 1 f = error Hz T e = Wih consideraion of daalogger error, he acual relay rip frequency was in he range from Hz o Hz. The relay could rip wihin specs a his poin, because is specified accuracy of +/- 0.01Hz could allow Hz o be sensed as Hz. There is a corresponding possibiliy ha he relay will no rip even when he measured frequency is calculaed o be slighly wihin 0.01 Hz over he frequency seing. In he nex cycle, he zero crossing is a rows and By inerpolaing beween rows and 52012, he zero crossing is esimaed a: 109

130 = 52011* s s * = s The zero crossing one cycle before his is a Row and which is 3. The period is: T = 4 3 = 0.018s can be calculaed as follows: 1 f = = Hz T The relay rips a his poin. The pickup frequency is recorded a Hz. Wih consideraion of es componen s error, he obained resul may deviae from he acual one. In his specific case of he seing a 55Hz and 6 cycle ime delay, he measured pickup frequency can be eiher 55.00Hz or Hz. while he recorded ime delay could vary beween 6 cycles and 7 cycles. 4.7 Fuure Work While he simulaions provide good daa for applicaion ess, acual field daa would grealy enhance he esing proocol. A library of such recorded daa should be developed. A new IEEE guide [26] addresses issues regarding frequency relay esing. These issues should be considered o improve he es sysem and proocol. Alhough mos of widely used relays oday employ he zero-crossing echnique o measure frequency, some of new frequency relays may apply he oher echnologies. The oher frequency measuring echniques should be invesigaed, and if necessary, algorihms for esing such relays should be developed and incorporaed ino he es proocols. 110

131 5.0 Conclusion 5.1 Disance relays This repor describes a es lab seup developed a Texas A&M Universiy for esing disance relays. The es procedure of relay es implemenaion on he plaform and he use in relay esing are also presened. Three differen disance relays are seleced o implemen relay ess using he proposed mehodology and ess resuls are given a he end. The proposed es mehodology ogeher wih he es ools and es case library composes a comprehensive es environmen for evaluaing he dependabiliy and securiy feaures of proecive relays. In he course of sudy i became apparen ha a differeniaion beween Conformance Tes and Compliance Tes should be made o help focus on differen ypes of design and applicaion ess. The Conformance es objecive is o es he basic funcionaliy of he relays, verify he operaing characerisics, calibrae relay seings and implemen periodic mainenance es. The concern of his es is he saisical performance relaed o he relay operaing characerisic and ripping ime. To fulfill his es, a bach of es cases wih a variey of disurbance condiions including fauls and non-fauls are generaed hrough simulaion. The Compliance es objecive is o verify wheher a relay can operae correcly under peculiar circumsances in power sysem paricularly during abnormal operaing condiions. This ype of es is o invesigae he compliance feaure ha real performance of a proecive relay complies wih is expeced performance. The concern of his es is he rip/no rip response and relay operaing ime performance under specific scenarios. A ypical example is he use of he recorded daa o analyze causes of an unwaned relay operaion in a pos-even analysis. The es resuls have shown ha in he fuure i will be equally imporan o es relays for dependabiliy and securiy of operaion. While he loss of securiy ha resuled in over ripping may have no been a concern in he fuure, due o heavily overloaded lines he unwaned rips can lead o a cascading even ending in a black ou. This repor has shown how he esing for securiy may be implemened. 5.2 Generaor Relays This repor describes he configuraion, simulaion, and insrumenaion requiremens for evaluaing he performance of generaor proecion relays under realisic ransien condiions, as hey may be encounered in a pracical elecric power sysem. As a resul, a comprehensive esing plaform has been buil o reproduce and simulae condiions in he sysem as closely o realiy as possible. The repor presened he esing plaform wih an emphasis on generaor proecive relays. The highlighs of he plaform include a a power sysem simulaor o accuraely compue shor-circui condiions as seen in an acual sysem by he proecive relays; b a signal condiioning uni ha reproduces he simulaed volages and currens a relay insrumenaion volage and curren level, as if 111

132 hey were delivered by acual poenial and curren ransformers; and c a se of procedures o conduc and validae he differen ess of he generaor relay, including relay connecions, sofware configuraion, and he differen es scenarios. An immediae applicaion of he developed mehodology and daa base is o es he seings of specific generaor relays and he degree of coordinaion of he various relay funcions. A fuure research direcion would be o use he developed mehodology in reverse mode, i.e. for he purpose of esimaing he model of he generaor. Accurae generaor modeling remains an issue. Approaches o esimae he generaor model in real ime, while hey exis, have no provided robus performance and he resuling model does no exhibi saisfacory agreemen wih observed generaor response. I is expeced ha he developed generaor model can provide a real ime esimaion mehodology ha will be robus and will resul in an accurae generaor model. The araciveness of he approach is ha he enre procedure can be performed wihin he generaor relay. 5.3 Underfrequency oad Shedding Relays This repor presens a new mehodology specifically designed for UFS relay esing. The ess include conformance and applicaion ess. Philosophies of esing are discussed and es proocols are presened. Tes proocols provide realisic and relevan ess o more accuraely simulae condiions relays may encouner in service. While much relay esing is done wih pure sinusoidal waveforms, he proocols include disored waveforms. Dynamic es cases are also provided o es relays under specific condiions. The cases provided are from simulaions, bu acual recorded daa can also be used when available. Two common UFS digial relays were esed under he new proocol. The resuls show he wo relays operaed differenly during ess. Some small deviaions from manufacurers specificaions were observed. The deviaions recorded in applicaion ess are larger han hose resuling from conformance ess. Discussions wih uiliy users, however, indicae ha he deviaions observed are inconsequenial for he users. The accuracy of esing componens may conribue error o he acquired resuls. The repor analyzes how he error of esing componens can affec he es resuls. Higher accuracy can be achieved by upgrading o higher accuracy hardware, e.g., a daalogger wih higher sampling rae. 112

133 6.0 Projec Publicaions [1] N. Zhang, H. Song and M. Kezunovic, Transien Based Relay Tesing: A New Scope and Mehodology, The 13h IEEE Medierranean Elecroechnical Conference MEECON 06, Torremolinos Málaga, Spain, May [2] M. Kezunovic, X. uo, N. Zhang, and H. Song, Tesing and Evaluaing New Sofware Soluions for Auomaed Analysis of Proecive Relay Operaions, The 7he Inernaional Conference on Power Sysems Transiens, yon, France, June [3] M. Kezunovic, J. Ren, New Tes Mehodology for Evaluaing Proecive Relay Securiy and Dependabiliy, IEEE Power Engineering Sociey General Meeing, Pisburgh, Pennsylvania, July [4] Q. Binh Dam, A. P. Sakis Meliopoulos, Relay Simulaion and Tesing Sofware on he.net Framework Environmen, IEEE Power Engineering Sociey General Meeing, June , Tampa, F USA. [5] Q. Binh Dam, A. P. Sakis Meliopoulos, A Breaker-oriened, Three-phase IEEE 24-subsaion Tes Sysem submied o IEEE Transacions [6] M. Shao, W. Jewell, Analysis of Proecive Relay Performance in he Augus 2003 Norh America Blackou, Froniers of Power Conference Proceedings, Sillwaer, Oklahoma, USA, Ocober [7] M. Shao, P. Poonpun, W. Jewell, An Advanced Mehodology for Underfrequency oad Shedding Relay Tesing, IEEE PES Transmission and Disribuion Conference and Exposiion, Chicago, I, USA, Apr

134 7.0 References [1] U.S.-Canada Power Sysem Ouage Task Force, Final Repor on he Augus 14, 2003 Blackou in he Unied Saes and Canada: Causes and Recommendaions, April 5, [Online]. Available: hp://www. nerc.com [2] M. Kezunovic, Y.Q. Xia, Y. Guo, C.W. Fromen, D.R. Sevcik, An Advanced Mehod for Tesing of Disance Relay Operaing Characerisic, IEEE Transacion on Power Delivery, Vol. 11, No. 1, pp , January [3] M. Kezunovic, B. Kaszenny, Design, Opimizaion and Performance Evaluaion of he Relaying Algorihms, Relays and Proecive Sysem Using Advanced Tesing Tools, IEEE Transacion on Power Delivery, Vol. 15, No. 4, pp , Ocober [4] Power Sysem Relaying Commiee, EMTP Reference Models for Transmission ine Relay Tesing Repor, draf 10a, Dec [Online]. Available: hp:// [5] S. K. M. Kodsi and C. A. Canizares, Modeling and Simulaion of IEEE 14 Bus Sysem wih FACTS Conrollers, Technical Repor , Waerloo, Canada, 2003, [Online]. Available: hp:// [6] CanAm EMTP User Group, Alernaive Transien Program ATP Rule Book, Porland, OR, [Online]. Available: hp:// files/secre/atp_rulebook [7] The MahWorks, Inc., Using MATAB, Naick, MA, July [Online]. Available: hp:// [8] H. Song, M. Kezunovic, A New Analysis Mehod for Early Deecion and Prevenion of Cascading Evens, Elecric Power Sysems Research, Vol. 77, No. 8, pp , June [9] F. Dobraca, M.A. Pai, P.W. Sauer, Relay Margins as a Tool for Dynamical Securiy Analysis, In. J. Elecr. Power Energy Sys., vol. 12 4, pp , Oc [10] X. uo, M. Kezunovic, "Ineracive Proecion Sysem Simulaion Using ATP MODES and C++," IEEE 2005/2006 PES Transmission & Disribuion Conference & Exposiion, Dallas, Texas, May [11] Tes aboraories Inernaional, Inc. Relay Assisan Digial Simulaor Sofware, [Online]. Available: hp:// [12] IEEE Sandard Common Forma for Transien Daa Exchange COMTRADE, IEEE Sandard C , March [13] Schweizer Engineering aboraories, Inc. SE-421, SE Relay, Proecion and Auomaion Sysem, Insrucion Manual User s Guide, Aug. 5, [14] Schweizer Engineering aboraories, Inc. SE-421, SE Relay, Proecion and Auomaion Sysem, Insrucion Manual Applicaions Handbook, Aug. 5, [15] Schweizer Engineering aboraories, Inc. ACSEERATOR QuickSe SE Sofware Version 2.4, Aug. 5, [Online]. Available: hp:// 114

135 [16] Schweizer Engineering aboraories, Inc. SE-321, SE Phase and ground disance relay, direcional over curren relay, faul locaor- Insrucion Manual. Nov. 20, [17] Schweizer Engineering aboraories, Inc. Relay Assisance Sofware SE-5010 Version 4.7, Nov.20, [Online]. Available: hp:// hm [18] GE Mulilin D60 ine Disance Relay Insrucion Manual, [Online]. Available: hp:// [19] GE EnerVisa UR Seup Sofware Version 5.4 July 3, [Online]. Available: hp:// [20] Beckwih M3425-A Generaor Proecion Insrucions Book, [Online]. Available: hp:// [21] IEEE Sandard Common Forma for Transien Daa Exchange COMTRADE for Power Sysems, IEEE Sandard C and C [22] A. P. Meliopoulos, G. J. Cokkinides, A Virual Environmen for Proecive Relaying Evaluaion and Tesing, IEEE Trans. Power Sysems, vol. 19, no. 1, pp , Feb [23] Q. Binh Dam, A. P. Sakis Meliopoulos, Relay Simulaion and Tesing Sofware on he.net Framework Environmen, IEEE Power Engineeering Sociey General Meeing, June , Tampa, F USA. [24] BECO2200-M-3425A, Communicaions Daabase for M3425A Inegraed Proecion Sysem May 2006, [Online]. Available: hp:// [25] BECO2200 Digial Communicaions Proocol, Revision 5 June 1996, [Online]. Available: hp:// [26] IEEE Guide for he Applicaion of Proecive Relays Used for Abnormal oad Shedding and Resoraion, IEEE Sandard C , C , IEEE Sd C [Online]. Available: hp://sandards.ieee.org. [27] M. Shao, P. Poonpun, W. Jewell, An Advanced Mehodology for Underfrequency oad Shedding Relay Tesing, IEEE PES Transmission and Disribuion Conference and Exposiion, Chicago, I, USA, Apr [28] Ward Jewell, R.D.A. Rob, T. K. Wong, and A. Dakkak Resuls of Proecive Relay Harmonic Waveform Disorion Tesing Program, Proceedings of he 7h Inernaional Conference on Harmonics and Qualiy of Power, as Vegas, Ocober [29] Alernaive Transiens Program. [Online]. Available: hp:// [30] G. T. Heyd, Elecric Power Qualiy, Second ediion, Sars in a Circle Publicaions, [31] Elecromagneic Transien Program EMTP Workbook Volume III, Elecric Power Research Insiue, June [32] MAAC Under- oad Shedding Program Requiremens, June 21, [Online]. Available: hp:// [33] S. Mohagheghi, R. H. Alaileh, G. Cokkinides and A. P. Sakis Meliopoulos, A laboraory seup for a subsaion scaled model, Proceedings of Power Tech 2007 Conference, ausanne, Swizerland, July 1 5,

136 [34] S. Mohagheghi, R. H. Alaileh, G. Cokkinides and A. P. Sakis Meliopoulos, Disribued sae esimaion using he SuperCalibraor concep: laboraory implemenaion, Proceedings of Inernaional Insiue for Research and Educaion in Power Sysem Dynamics irep 2007 Symposium, Charleson, SC USA, Aug ,

137 Appendix A: ine Disance Relay Tes A.1 Relay Seings Relays are se according o he es plan discussed in secion 2 and he reference provided by vendors [13], [14], [16], [18]. Table A.1 and Table A.2 are seing ables for SE-421 relay for implemening conformance es and applicaion es respecively. Seing ables for SE-321 relay are negleced because of he same parameers and similar seing names. Figure A.1, A.2 and A.3 are given o presen he seings for GE D60 insead of lising parameers. For SE-421 and SE-321, he seing names come from he SE 5030 and SE 5010 respecively which are sofware used for manage relays [15], [17]. For GE D60, hese figures are generaed by EnerVisa UR Seup sofware [19]. Table A.1: Seing able for SE-421 for Conformance Tes Seing Name Value Seing Name Value SID Saion Idenifier 230kV BUS1 Z1D Zone RID Relay Idenifier SE Z2D Zone NUMBK 1 Z3D Zone BID 1 Breaker 1 Idenifier Breaker 1-ine 1 ESOTF Swich-Ono-Faul Y NFREQ 60 ESPSTF N PHROT ABC EVRST Y ESS N 52AEND 10.0 CTRW 400 SOTFD 10.0 PTRY 2300 CSMON IN102 VNOMY 100 EOOS Ou-of-Sep Y Z1MAG 7.13 OOSB1 Block Zone 1 Y Z1ANG 84.2 OOSB2 Block Zone 2 Y Z0MAG OOSB3 Block Zone 3 N Z0ANG 81.7 OSBD 2.5 EFOC Faul ocaion Y OSBTCH Y ine engh mile 45 EOOST N E21P 3 X1T Z1P Zone 1 Reach 5.71 X1T Z2P Zone 2 Reach 7.13 R1R Z3P Zone 3 Reach R1R6 4.3 Z1PD Zone EOAD Y Z2PD Zone ZF 6.29 Z2PD Zone ZR 6.29 E21MG 3 PAF 45.0 Z1MG Zone 1 Reach 5.71 NAF Z2MG Zone 2 Reach 7.13 PAR Z3MG Zone 3 Reach NAR E21XG N E50P 1 Z1GD Zone P1P evel 1 Pickup 3.23 Z2GD Zone P1D evel Z3GD Zone P1TC 1 k0m DIR3 F k0a TR Trip Z1T OR Z2T OR Z3T ECDTD Y TRSOTF M2P OR Z2G OR M3P OR Z3G 117

138 Table A.2: Seing able for SE-421 for Compliance Tes Seing Name Value Seing Name Value SID Saion Idenifier 138kV BUS2 Z3MG Zone 3 Reach 3.67 RID Relay Idenifier SE E21XG N NUMBK 1 Z1GD Zone BID 1 Breaker 1 Idenifier Breaker 1-Bus 2 Z2GD Zone NFREQ 60 Z1D Zone PHROT ABC Z2D Zone ESS N Z3D Zone CTRW 100 ESOTF Swich-Ono-Faul N PTRY 1380 EOOS Ou-of-Sep Y VNOMY 100 OOSB1 Block Zone 1 Y Z1MAG 1.48 OOSB2 Block Zone 2 Y Z1ANG OOSB3 Block Zone 3 N Z0MAG 3.69 OSBD 3.05 Z0ANG OSBTCH Y EFOC Faul ocaion Y EOOST I ine engh mile 33 OSTD E21P 3 X1T Z1P Zone 1 Reach 1.18 X1T Z2P Zone 2 Reach 1.78 R1R Z3P Zone 3 Reach 3.67 R1R Z1PD Zone EOAD Y Z2PD Zone ZF 1.81 Z2PD Zone ZR 1.81 E21MG 3 PAF 45.0 Z1MG Zone 1 Reach 1.18 NAF Z2MG Zone 2 Reach 1.78 PAR Z3GD Zone NAR k0m DIR3 F k0a TR Trip Z1T OR Z2T OR Z3T ECDTD Y ER Even Repor Trigger M2P OR Z2G OR M3P OR Z3G Figure A.1: Phase disance proecion 118

139 Figure A.2: Power swing proecion Figure A.3: oad encroachmen proecion Noe: Since GE D60 does no have he proecion elemen special for he swich ono faul condiion, a combinaion of ine Pickup and Phase IOC is applied using Flexogic o realize his funcion. 119

140 A.2 Tes Resuls Table A.3: Tes resuls for condiion F1 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y

141 BC Y Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

142 Table A.4: Tes resuls for condiion F2-1 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N BC N BC N BC N BC N BC N BC N BC N BC N BC N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N ABC N ABC N 122

143 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] ABC N ABC N ABC N ABC N ABC N ABC N ABC N Table A.5: Tes resuls for condiion F2-2 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG N AG N AG N BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

144 Table A.6: Tes resuls for condiion F3 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

145 Table A.7: Tes resuls for condiion F4-1 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Uni Trip [s] AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF ABC Y SOTF ABC Y SOTF

146 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Uni Trip [s] ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF Table A.8: Tes resuls for condiion F4-2 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Uni Trip [s] AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF

147 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Uni Trip [s] BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF

148 Table A.9: Tes resuls for condiion F5 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

149 Table A.10: Tes resuls for condiion F6-1 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

150 Table A.11: Tes resuls for condiion F6-2 for SE-421 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y Table A.12: Saisical es resuls for inernal fauls for SE-421 Type oc α Rf Trip No. Mean T Max T Min T Devn [%] [deg] [Ω] Zone T [ms] [ms] [ms] [ms] AG I AG I AG II BC I BC I BC II BCG I BCG I BCG II ABC I ABC I ABC II

151 Table A.13: Tes resuls for no-faul scenarios for SE-421 Type Operaion Trip / Trip Trip NoTrip Zone [s] N1-1 Three phases close afer 2 cycles N N1-1 Phase A close afer 2 cycles N N1-1 Phase B, C close afer 2 cycles N N1-2 Three phases close afer 2 cycles N N1-2 Phase A close afer 2 cycles N N1-2 Phase B, C close afer 2 cycles N N2 Remove S1 afer 2 cycles N N2 Remove S2 afer 2 cycles N N2 Remove S3 afer 2 cycles N N2 Remove S2, S3 afer 2 cycles N N2 Remove S1, S2, S3 simulaneously afer 2 cycles N N2 Remove S1, hen S2 afer 2 cycles, hen S3 afer 2 cycles N N3 Open Bus 2 breaker afer 2 cycles N N3 Open Bus 4 breaker afer 2 cycles N N3 Open SW afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S2, S3 afer 2 cycles N N4 Resore S1, S2, S3 simulaneously afer 2 cycles N N4 Resore S1, hen S2 afer 2 cycles, hen S3 afer 2 cycles N N5 Power swing afer hree-faul occurred on ine1 N N6 Secondary Impedance: N N6 Secondary Impedance: N N6 Secondary Impedance: N N6 Secondary Impedance: 7.90 N 131

152 Type A1 A2 A3 A4 A5 A6 A7 A8 Table A.14: Compliance es resul for SE-421 oc Faul oad Trip / no rip Trip / no rip on Power CCT[s] [%] Type Condiion on Faul Swing or Ou of Sep 10 Y N Single 3-50 Base Y N phase 90 Y N 10 Y N Single 3-50 Over Y N phase 90 Y N 10 Y N Two 3-50 Base Y N phase 90 Y N 10 Y N Two 3-50 Over Y N phase 90 Y N 10 Y Y Single 3-50 Base Y Y phase 90 Y Y 10 Y Y Single 3-50 Over Y Y phase 90 Y Y 10 Y Y Two 3-50 Base Y Y phase 90 Y Y 10 Y Y Two 3-50 Over Y Y phase 90 Y Y 132

153 Table A.15: Tes resuls for condiion F1 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG N AG Y AG Y AG N AG Y AG Y AG N AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y

154 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

155 Table A.16: Tes resuls for condiion F2-1 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N BC N BC N BC N BC N BC N BC N BC N BC N BC N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N ABC N ABC N 135

156 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] ABC N ABC N ABC N ABC N ABC N ABC N ABC N Table A.17: Tes resuls for condiion F2-2 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG N AG N AG N BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

157 Table A.18: Tes resuls for condiion F3 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

158 Table A.19: Tes resuls for condiion F4-1 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Uni Trip [s] AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF ABC Y SOTF ABC Y SOTF

159 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Uni Trip [s] ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF Table A.20: Tes resuls for condiion F4-2 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Uni Trip [s] AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF AG Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BC Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF

160 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Uni Trip [s] BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF BCG Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF ABC Y SOTF

161 Table A.21: Tes resuls for condiion F5 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

162 Table A.22: Tes resuls for condiion F6-1 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

163 Table A.23: Tes resuls for condiion F6-2 for SE-321 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y Table A.24: Saisical es resuls of inernal fauls for SE-321 Type oc α Rf Trip No. Mean T Max T Min T Devn [%] [deg] [Ω] Zone T [ms] [ms] [ms] [ms] AG I AG I AG II BC I BC I BC II BCG I BCG I BCG II ABC I ABC I ABC II

164 Table A.25: Tes resuls of no-faul scenarios for SE-321 Type Operaion Trip / No Trip Trip Trip Zone [s] N1-1 Three phases close afer 2 cycles N N1-1 Phase A close afer 2 cycles N N1-1 Phase B, C close afer 2 cycles N N1-2 Three phases close afer 2 cycles N N1-2 Phase A close afer 2 cycles N N1-2 Phase B, C close afer 2 cycles N N2 Remove S1 afer 2 cycles N N2 Remove S2 afer 2 cycles N N2 Remove S3 afer 2 cycles N N2 Remove S2, S3 afer 2 cycles N N2 Remove S1, S2, S3 simulaneously afer 2 cycles N N2 Remove S1, hen S2 afer 2 cycles, hen S3 afer 2 cycles N N3 Open Bus 2 breaker afer 2 cycles N N3 Open Bus 4 breaker afer 2 cycles N N3 Open SW afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S2, S3 afer 2 cycles N N4 Resore S1, S2, S3 simulaneously afer 2 cycles N N4 Resore S1, hen S2 afer 2 cycles, hen S3 afer 2 cycles N N5 Power swing afer hree-faul occurred on ine1 N N6 Secondary Impedance: N N6 Secondary Impedance: N N6 Secondary Impedance: N N6 Secondary Impedance: 7.90 N 144

165 Type A1 A2 A3 A4 A5 A6 A7 A8 Table A.26: Compliance es resul for SE-321 oc Faul oad Trip / no rip Trip / no rip on Power CCT[s] [%] Type Condiion on Faul Swing or Ou of Sep 10 Y N Single 3-50 Base Y N phase 90 Y N 10 Y N Single 3-50 Over Y N phase 90 Y N 10 Y N Two 3-50 Base Y N phase 90 Y N 10 Y N Two 3-50 Over Y N phase 90 Y N 10 Y Y Single 3-50 Base Y Y phase 90 Y Y 10 Y Y Single 3-50 Over Y Y phase 90 Y Y 10 Y Y Two 3-50 Base Y Y phase 90 Y Y 10 Y Y Two 3-50 Over Y Y phase 90 Y Y 145

166 Table A.27: Tes resuls for condiion F1 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y

167 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

168 Table A.28: Tes resuls for condiion F2-1 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N AG N BC N BC N BC N BC N BC N BC N BC N BC N BC N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N BCG N ABC N ABC N 148

169 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] ABC N ABC N ABC N ABC N ABC N ABC N ABC N Table A.29: Tes resuls for condiion F2-2 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG N AG N AG N BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

170 Table A.30: Tes resuls for condiion F3 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

171 Table A.31: Tes resuls for condiion F4-1 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y

172 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip [s] ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y Table A.32: Tes resuls for condiion F4-2 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y

173 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip [s] BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

174 Table A.33: Tes resuls for condiion F5 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

175 Table A.34: Tes resuls for condiion F6-1 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y

176 Table A.35: Tes resuls for condiion F6-2 for GE D60 Type oc [%] Incepion Angle [deg] Resisance [Ω] Trip / no rip Trip Zone Trip [s] AG Y AG Y AG Y AG Y AG Y AG Y BC Y BC Y BC Y BC Y BC Y BC Y BCG Y BCG Y BCG Y BCG Y BCG Y BCG Y ABC Y ABC Y ABC Y ABC Y ABC Y ABC Y Table A.36: Saisical es resuls for inernal fauls for GE D60 Type oc α Rf Trip No. Mean T Max T Min T Devn [%] [deg] [Ω] Zone T [ms] [ms] [ms] [ms] AG I AG I AG II BC I BC I BC II BCG I BCG I BCG II ABC I ABC I ABC II

177 Table A.37: Tes resuls of no-faul scenarios for GE D60 Type Operaion Trip / No Trip Trip Trip Zone [s] N1-1 Three phases close afer 2 cycles N N1-1 Phase A close afer 2 cycles N N1-1 Phase B, C close afer 2 cycles N N1-2 Three phases close afer 2 cycles N N1-2 Phase A close afer 2 cycles N N1-2 Phase B, C close afer 2 cycles N N2 Remove S1 afer 2 cycles N N2 Remove S2 afer 2 cycles N N2 Remove S3 afer 2 cycles N N2 Remove S2, S3 afer 2 cycles N N2 Remove S1, S2, S3 simulaneously afer 2 cycles N N2 Remove S1, hen S2 afer 2 cycles, hen S3 afer 2 cycles N N3 Open Bus 2 breaker afer 2 cycles N N3 Open Bus 4 breaker afer 2 cycles N N3 Open SW afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S1 afer 2 cycles N N4 Resore S2, S3 afer 2 cycles N N4 Resore S1, S2, S3 simulaneously afer 2 cycles N N4 Resore S1, hen S2 afer 2 cycles, hen S3 afer 2 cycles N N5 Power swing afer hree-faul occurred on ine1 N N6 Secondary Impedance: N N6 Secondary Impedance: N N6 Secondary Impedance: N N6 Secondary Impedance: 7.90 N 157

178 Table A.38: Compliance es resul for GE D60 Type oc [%] Faul Type oad Condiion Trip / no rip CCT[s] Trip / no rip 10 Y N A1 50 Single 3-phase Base Y N 90 Y N 10 Y N A2 50 Single 3-phase Over Y N 90 Y N 10 Y N A3 50 Two 3-phase Base Y N 90 Y N 10 Y N A4 50 Two 3-phase Over Y N 90 Y N 10 Y Y A5 50 Single 3-phase Base Y Y 90 Y Y 10 Y Y A6 50 Single 3-phase Over Y Y 90 Y Y 10 Y Y A7 50 Two 3-phase Base Y Y 90 Y Y 10 Y Y A8 50 Two 3-phase Over Y Y 90 Y Y 158

179 Appendix B: Generaor Relay Tes B.1 Generaor Relay Proecion Scheme and Connecions The proecion scheme o be reproduced o es he generaor proecion relays are shown in Figure B.1. The figure shows wha measuremens he relay acceps, namely high-side volage, high-side and low-side currens, neural currens and volages, and zero sequence currens and volages. All measuremens are conneced o a specific measuremen channel of he relay. The relay has 12 inpus oal, and all 12 inpus are uilized o es he differen proecion schemes suppored by he relay. A deailed schemaic of he same proecion scheme is shown in Figure B.2 for he M-3425A relay and Figure B.3 for he 300G relay. The figures all come from he daa shees available from he relay manufacurers. The connecions for he ypical proecion scheme shown in hese figures enable a number of funcions ha are idenified in Figure B.4 by heir number. The developed laboraory seup does no use measuremens from acual CTs or VTs. Insead, he CT and VT signals are simulaed using he sofware plaform and recreaed using a waveform generaor. Par of he waveform generaor is a heaer amplifier ha scales he oupu of he D/A converer from 10 V o 30 V. For volage measuremens, he oupu of he amplifier is brough o he nominal volage of he relay 69 V using a booser ransformer bench. The final cabling is shown in Figure B

180 Generaor Proecion Relay Figure B.1: Insrumenaion connecions of he generaor proecion relays 160

181 Figure B.2: M-3425A deailed connecions of measuremen channels o relay inpus for a ypical proecion scheme aken from [20] page

182 Figure B.3: M-3425 funcions available from ypical volage and curren wirings o he relay aken from [20], page

183 Figure B.4: Typical connecion diagram for he 300G relay available from he manufacurer daa shee 163

184 Figure B.5: 300G funcions available from ypical volage and curren wirings o he relay available from he manufacurer daa shee 164

185 Figure B.6: Connecions beween signal amplifiers and he esed generaor relays B.2 is of Generaor Evens for Relay Tesing This Appendix describes he suggesed procedures for reproducing specific evens o es he generaor relay. In realiy, several of hese evens may happen simulaneously as a resul of a larger even in he power sysem, such as a faul or a large-scale acion on he sysem. The presen capabiliies of he proposed generaor model see Appendix B.3 for ransien evens simulaion are also lised in his appendix. 165