Response_Of_Power_System_MATTEI_PRESENTATION

Response of Power System Protective Relays to Solar and HEMP MHD-E3 GIC Andrew K Mattei, Baylor University Brazos Electric Cooperative Dr W Mack Grady, Baylor University Texas A&M Conference for Protective Relay Engineers, 2019 Outline • Introduction to the phenomenon • Harmonic Distortion and Currents • Solar Storms • Time-Overcurrent Relay Testing • High Altitude Nuclear Bursts and Results • The Magnetosphere and the Electric Field • Differential Relay Testing and • Solar vs HEMP GIC Results • The Waveforms of DC Injection • Detection and (Possible) • Waveform 1: Test Bench Data Mitigation of GIC Interference • Waveform 2: NERC Example • Waveform 3: DTRA Field Test Data Solar Storms • Coronal Mass Ejection (CME) events send large quantities of magnetized gas into space • If the ejection occurs on the proper trajectory, these particles may disturb the Earth’s magnetosphere • NOAA Space Weather Prediction Center https://www.swpc.noaa.gov https://photojournal.jpl.nasa.gov/catalog/?IDNumber=PIA03149 High Altitude Nuclear Burst • >30 km above the Earth’s surface, up to several hundred km • Thin atmosphere allows wide spread of gamma particles, whose collisions facilitate the ionization of the more dense atmospheric layer below • The ionization and subsequent fireball rise from atmospheric heating result in distinct periods of slow-moving electric fields – “Blast” and “Heave” https://en.wikipedia.org/wiki/File:SunBurst10.jpg – modified version of https://commons.wikimedia.org/wiki/File:Robot_Arm_Over_Earth_with_Sunburst_-_GPN-2000-001097.jpg Characteristics of a High-Altitude Nuclear EMP Burst • E1 – Fast, strong EMP (nsec rise, 10’s of nsec in duration) • E2 – Not quite as strong, slower (lightning-esque), ~1 sec • E3 – separated into E3A “Blast” and E3B “Heave”, 300 seconds in duration Source: Gilbert et.al., [3] HEMP Footprint • Pictured, altitude of 150 km • Affected area depends on altitude, weapon design, and Earth’s field state Source: Gilbert et.al., [3] The Magnetosphere of the Earth • Earth’s orbit distorts the magnetic fields • Asymmetrical field lines between daylight and night sides • Creates natural areas for common solar disturbance / Auroras (poles) https://commons.wikimedia.org/wiki/File:Structure_of_the_magnetosphere-en.svg Electric Fields • Slow disturbance of geomagnetic field lines result in slow-moving electric fields • Elevated V/km fields, coupled with low impedance lines with ground connections at either end (GY transformers) forms a loop • Loop allows for quasi-DC flow Photo by Matthias Cooper from Pexels Solar GIC Event vs Single HEMP Event • Solar events can be hours or days in duration, but of lower magnitude (8-12 V/km for NERC Reference and Supplemental events per TPL-007-2) • 10-second data points in upper graph, 11,200 points total • MHD-E3 event lasts 300-400 seconds (single burst) • Stronger electric fields (10-50 V/km) – possible thousand+ amps per phase • Either may result in hot-spots in transformers, cap bank / SVC protection operations due to harmonics, relay mis-operations Derived from NERC Reference Event based on custom calculation Rackliffe, Crouse, Legro, Kruse “Simulation of Geomagnetic Currents Induced in a Power System by Magnetohydrodynamic Electromagnetic Pulses”, IEEE Transactions on Power Delivery, Vol 3, No 1, January 1988 What are we going to see with DC in the system? • IEEE Std C57.163-2015: IEEE Guide for Establishing Power Transformer Capability while under Geomagnetic Disturbances • The Late-Time (E3) High-Altitude Electromagnetic Pulse (HEMP) and Its Impact on the U.S Power Grid – Metatech, 2010 Source: IEEE Std C57.163-2015 [6] Source: IEEE Std C57.163-2015 [6] Source: Gilbert et.al., [3] DTRA DC Injection Test Bed @ Idaho National Labs “Waveform 3” Source: [11] In a Normal World (50 or 60 Hz) 𝐼𝐼𝑅𝑅𝑅𝑅𝑅𝑅 ≅ 𝐼𝐼𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹 ≅ 𝐼𝐼𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹 ≅ 𝐼𝐼𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 In a Distorted World 𝐼𝐼𝑅𝑅𝑅𝑅𝑅𝑅 ≇ 𝐼𝐼𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹 ≇ Test Waveform Waveform Waveform DC 0mA DC 40mA DC 68mA DC 107mA IRMS 10.25 10.25 10.25 6.73 6.93 7.31 8.18 IFILTERED/FUNDAMENTAL 9.06 5.03 9.46 6.70 6.85 7.13 7.72 𝐼𝐼𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 IPEAK÷√2 17.37 16.79 12.81 6.24 6.75 8.26 11.48 Time-overcurrent relays tested Westinghouse CO-9 GE IAC-53B SEL-421 • All set to same pickup / time dial - obviously curves / times will be different between them • The Question: Can we predict EM relay operation based on RMS current? Waveform CO9 THD Time (s) 10.25A RMS 53% 3.08 177.5% 4.994 41.8% 2.738 IAC53B Time (s) 2.445 2.281 2.425 SEL421 Time (s) 3.587 480+ 3.2 Waveform – Varying RMS Magnitude • Concern becomes unpredictable behavior near pick-up point when compared to fundamental component Differential Relay Testing (Simplified) Westinghouse HU, GE STD, GE BDD, SEL-487E Test (baseline): Does relay trip for loss of secondary side current? • All current flowing into primary side winding when secondary current drops out • Of course it trips, as it should All relays tested - tripped Test 2: Does relay trip during Waveform on Primary winding? • Drop secondary current again, just like Test • None of the tested relays tripped Second harmonic blocking / restraint inhibits tripping • Unrestrained tripping (instantaneous) is not affected Relay will still trip on high current levels Observations • Time-overcurrent protection can become a bit unpredictable with harmonics Even though relays may feature similar construction (induction-disc), the way they handle harmonics may be different • With a microprocessor-based relay, severe harmonics (Waveform 2) may not result in protection activating even though peak current is well above pickup (operating only on fundamental) • Differential relays may not operate during GIC events until the instantaneous pickup is reached This may slow tripping and result in more damage than a ‘fast trip’ • Two types of relays? That’s all you’ve got? Selection - So what relays we focus on? • Overheard within ERCOT: How we know what relays to look at? Where can we start? • Start with the TPL-007-2 GIC model for the Benchmark Event • Evaluate the higher-magnitude GIC flows within your area of responsibility • Determine transformers, capacitor banks, and SVC’s near the elevated-GIC area • Examine the protection on those devices Run some harmoniclaced COMTRADE files if necessary Detection – When is it happening? • Borrowing from Zweigle, Pope, & Whitehead [12] – use relays to measure harmonics and feed back via SCADA or synchrophasor • Upper graph – September 7, 2017, shell form autotransformer G3 event No increase in second harmonic found 2nd, 4th, 5th harmonic being logged every minute – have over years of data No coincidental GIC harmonics ever found (here in Texas) • Lower graph – Two seconds of synchrophasor data based on each waveform from Baylor Test Bench (varying DC injection levels) Second harmonic percentage of Operate Current A good indicator that something is ‘happening’ in the grid Mitigation – what can relays do? • Harmonic detection can send alarms to System Operators so that mitigating actions can be taken • Use a timer on second harmonic blocking pick-up If duration is greater than 120 cycles (or ?), assert an alarm to SCADA • Plan ahead with alternate settings groups for high-impact areas Consider reduced safety margin to keep equipment – especially reactive equipment – in service Transformers in deep saturation absorb VARs and cap banks / SVCs become very important for stability Conclusion • EM relays may be unpredictable Examine the GIC model and determine impacted areas In these areas, strongly consider replacing EM relays with uP • Be considerate with uP relays and harmonic distortion Will distortion be greater than the fundamental (like Waveform 2)? • Consider alarming on long-duration harmonic pickup This will alert System Operator that differential protection may be disabled, or a time-overcurrent element may not have enough fundamental to assert Concern Areas that Need Testing • Electromechanical Negative-Sequence Protection • Voltage waveforms are different than current waveforms – EM cap bank protection • Looking for relays… • FFT’s for waveforms are included in appendix of paper COMTRADE files for everyone… Questions Look for me during the conference – comments and suggestions for research are appreciated! Andrew_Mattei@baylor.edu Thank you to Gene Corpuz @ LCRA and Mark Chronister @ Oncor for letting us borrow EM relays, to Brazos Electric Cooperative for the test lab, Schweitzer Engineering Labs for their support, and to the Defense Threat Reduction Agency (DTRA) for data, questions, and funding the research https://en.wikipedia.org/wiki/File:SunBurst10.jpg – modified version of https://commons.wikimedia.org/wiki/File:Robot_Arm_Over_Earth_with_Sunburst_-_GPN-2000-001097.jpg