Power Systems Engineering Course: Grid Frequency Response Jason MacDowell © 2016 General Electric International, Inc All rights reserved Not for distribution without permission CAISO Frequency Response Study Adopted from CA Stakeholder Conference December 13, 2011 GE Energy Nicholas W Miller Miaolei Shao Sundar Venkataraman For PSEC 2016: This study helped establish some important fundamentals The industry is moving fast on this topic Updates and new notes added since the original work are in GREEN CAISO Clyde Loutan Mark Rothleder © 2016 General Electric International, Inc All rights reserved Not for distribution without permission Frequency Response: Basics • Grid must maintain balance between load and generation • Large disturbances, particularly trips of large generating plants, cause unbalance that must be corrected by “Frequency Response” • Frequency response covers multiple time frames • inertial response (up to a few seconds) • governor response (aka “Primary Response” – 1s to 10s of seconds) • AGC response (re-dispatch) (aka “Secondary Response” - tens of seconds to tens of minutes) • • Committed synchronous generation naturally contributes to system inertia • inertial response for these resources is not controllable, and is not a function of loading level Some synchronous generation provides governor response, if (a) governors are enabled and (b) it has “headroom” to increase output © 2016 General Electric International, Inc All rights reserved Not for distribution without permission Frequency Response: Today’s Reality • System governor response (primary response) has been declining steadily in some parts of the US for many years: Subject of Sept 23, 2010 FERC Technical Conference and NERC Comments of October 14, 2010 Now subject to new (approved on January 14, 2014) NERC Standard BAL-003-1 which establishes “Frequency Response Obligation” • This predates the rapid recent growth of wind generation in North America • Declining response results in deeper frequency (nadir) excursions for system disturbances, increasing the risk of under-frequency load shedding (UFLS) and cascading outages • Concern is most acute at lighter load conditions • economics favor fewer generators committed with less governor response • worst disturbances are not dependent on system load level • The issue is getting considerable attention in many circles © 2016 General Electric International, Inc All rights reserved Not for distribution without permission A Example from WECC • 2010 WECC database • Light summer load (110GW: 65-75% peak) • Full dynamic modeling • data used today to make planning decisions • Disturbance: x Palo Verde NPS (loss of 4132 MW) • really big event • not design basis, but has happened The authors gratefully acknowledge the support of the Western Electricity Coordinating Council (WECC) Renewable Energy Modeling Task Force (REMTF), and its chairman, Abraham Ellis, in supplying access to planning data for this work © 2016 General Electric International, Inc All rights reserved Not for distribution without permission Inertia vs Governor Response • Combination of inertia & governor response dictates minimum frequency (Fmin) (Nadir) • Governor response dominates later (DF), until AGC takes over (not modeled here) Total Generation Inertial Response ~1400 MW/0.1 Hz (~ 0.9 % of peak load) Nominal Frequency Response (NERC guideline is >1%) Governor (Primary) GovernorResponse Response 110,000 Load Response 105,000 100,000 10 15 20 25 30 35 40 Time (seconds) 60.1 Frequency of Various WECC Buses 60 • At 40 seconds: Frequency at Malin 500kV Bus 59.9 Frequency (Hz) ~3800 MW Governor Response ~270 mHz Frequency Deviation Total Load 115,000 Power (MW) • Inertial response dominates initial frequency decline 120,000 DF 59.8 59.7 59.6 Fmin (Nadir) 59.5 59.4 10 15 20 25 Time (seconds) © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 30 35 40 Frequency Response with Wind: What’s Different? Commitment and dispatch • Wind generation causes some synchronous generators to be decommitted, and some to be dispatched down to lower power levels Inertia • Modern variable speed wind turbine-generators not naturally contribute to system inertia • De-commitment of some generators reduces system inertia Governor Response • Wind generation, like many generation resources, including nuclear, boiler follow steam, etc., does not contribute to governor response without incurring significant operational cost penalties • Other resources that have governor response may be • De-committed, removing their contribution • Dispatched down, giving more “headroom” © 2016 General Electric International, Inc All rights reserved Not for distribution without permission Outline • Study Objectives • Development of Study Database and Performance Metrics • Frequency Response of Base Cases • Frequency Response of High Renewable Penetration Cases • Factors Affecting Frequency Response • Mitigation Measures • Conclusions Study Objectives • Frequency response to large generation outages for CAISO as well as the overall WECC, under a variety of system conditions • Impact of unit commitment/dispatch on frequency response • Impact of generator output level on governor response • Potential mitigation measures © 2016 General Electric International, Inc All rights reserved Not for distribution without permission Outline • Study Objectives • Development of Study Database and Performance Metrics • Frequency Response of Base Cases • Frequency Response of High Renewable Penetration Cases • Factors Affecting Frequency Response • Mitigation Measures • Conclusions 10 Comparison of Impact of Increasing Levels of Wind on Frequency Performance to Loss of Two Palo Verde Units More wind has worse but acceptable frequency response California’s frequency response improves (from 287 to 311 MW/0.1 Hz – well above the 205 MW/0.1Hz target) The fractional contribution in California increases greatly, from 20% to 27% The behavior of resources outside of California has impact on the California response © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 21 Outline • Study Objectives • Development of Study Database and Performance Metrics • Frequency Response of Base Cases • Frequency Response of High Renewable Penetration Cases • Factors Affecting Frequency Response • Mitigation Measures • Conclusions 22 Factors Degrading Frequency Response – Fewer Governors in Operation Keep all other factors impacting frequency response fixed: Governor Response (GR) units • 25 GR units, with total dispatch of 3144 MW and rating (MWCAP) of 5189 MW for a total of 2045 MW headroom, were selected to dispatch up 2045 MW and then were set as base load • Another 11 GR units, with total dispatch of 3034MW and rating (MWCAP) of 4165 MW were selected to dispatch down 2045 MW Reduce the count of generators providing response by 25, while holding headroom fixed (i.e Kt is reduced) Headroom not sufficient alone – Kt matters!!! © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 23 Factors Degrading Frequency Response – Reduced Headroom • Small Change in Headroom • Practical Minimum Headroom • Extreme minimum Headroom © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 24 Reduce Headroom - Practical Minimum Headroom GR Pgen (MW) GR MWCAP (MW) GR Headroom (MW) BL Pgen (MW) NG Pgen (MW) Wind Pgen (MW) Solar Pgen (MW) MW Capability CU Pgen (MW) (GR + BL + NG) Total Pgen (MW) Total Load (MW) Wind Pgen/Total Pgen Solar Pgen/Total Pgen Kt GR Pgen/CU Pgen GR Pgen/Total Pgen GR Headroom/CU Pgen GR Headroom/Total Pgen 13640 WECC CA Non-CA # of Units # of Units # of Units 18942 284 5045 92 13897 192 27057 8169 18888 3974 9765 8115 3124 4991 44815 510 12780 168 32035 342 9678 320 2617 99 7060 221 18094 8411 9684 2550 2550 102194 73435 94392 91300 34527 1114 20442 29683 26190 19.2% 2.7% 26.5% 25.8% 20.1% 11.1% 8.6% 28.3% 8.6% 23.7% 25.5% 24.7% 17.0% 15.3% 10.5% 359 25.6% 67667 52992 64710 65111 15.0% 0.0% 27.9% 26.2% 21.5% 9.4% 7.7% Condition in this case was considered to be challenging © 2016 Electric International, Inc All infrequently rights reserved Not for distribution without permission andGeneral might occur relatively 755 25.4% 25 Reduce Headroom - Practical Minimum Headroom © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 26 Generation Summary for Winter Low Load – High WECC Wind Case – Extreme Minimum Headroom GR Pgen (MW) GR MWCAP (MW) GR Headroom (MW) BL Pgen (MW) NG Pgen (MW) Wind Pgen (MW) Solar Pgen (MW) WECC CA # of Units # of Units 23913 284 7018 92 27057 8169 13640 3974 9765 3144 1151 39676 510 11439 168 9678 320 2617 99 18094 8411 2550 2550 MW Capability CU Pgen (MW) (GR + BL + NG) Total Pgen (MW) Total Pload (MW) 97055 73267 94225 91301 33186 1114 21074 30315 26190 Wind Pgen/Total Pgen Solar Pgen/Total Pgen Kt GR Pgen/CU Pgen GR Pgen/Total Pgen GR Headroom/CU Pgen GR Headroom/Total Pgen 19.2% 2.7% 27.9% 32.6% 25.4% 4.3% 3.3% 27.7% 8.4% 24.6% 25.5% 33.3% 23.2% 5.5% 3.8% 359 25.6% Non-CA # of Units 16895 192 18888 1993 28238 342 7060 221 9684 63870 52193 63910 65111 15.2% 0.0% 29.6% 32.4% 26.4% 3.8% 3.1% © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 755 25.4% 27 Impact of Extreme Minimum Headroom and Governor Participation (Kt) on Frequency Performance Winter Low Load – High WECC Wind Case UFLS relay off Kt alone is insufficient to anticipate frequency performance Headroom should be considered – at least when it is in short supply Time and time window for which settling frequency is measured becomes quite important © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 28 Outline • Study Objectives • Development of Study Database and Performance Metrics • Frequency Response of Base Cases • Frequency Response of High Renewable Penetration Cases • Factors Affecting Frequency Response • Mitigation Measures Reduced Governor Withdrawal Inertial Response From Wind Plants Governor Response (Frequency Droop) from Wind Plants Load Control/Fast Energy Storage • Conclusions 29 Mitigation Measures – Reduced Governor Withdrawal Disable load control on the 18 units with lcfb1 model Load control has relatively small impact on the frequency nadir Settling frequency is significantly impacted Withdrawal causes a 20% degradation in NERC frequency response © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 30 Mitigation Measures – Inertial Response From Wind Plant Winter Low Load – High WECC Wind case all of the type wind turbine machines, with a total power output of 14600 MW (out of a total of 18094 MW wind for the case) are assumed to have an inertial control The ability to tune inertial controls presents an opportunity to improve system performance © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 31 Simulation: WindINERTIA vs Conventional 1200 1200 Wind Speed = 14 m/s 1100 Power (MW) Synchronous Machine Without Wind Inertia 1000 1000 900 800 61 Wind Speed = 11 m/s Frequency (Hz) 800 10 61 60 60 59 59 30 seconds 20 30 © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 58 58 10 20 30 Mitigation Measures – Governor Response (Frequency Droop) from Wind Plants Winter Low Load – High WECC Wind Case – Extreme Minimum Spinning Reserves Approximately 41% of all the WTGs in WECC are provided with standard 5% droop, 36mHz deadband governors This condition adds a total of 1812 MW of headroom Primary frequency response from wind generation has the potential to greatly improve system frequency performance of the entire WECC grid The California contribution to frequency response goes from an unacceptable 152 MW/0.1 Hz to a healthy 258 MW/0.1 Hz © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 33 Outline • Study Objectives • Development of Study Database and Performance Metrics • Frequency Response of Base Cases • Frequency Response of High Renewable Penetration Cases • Factors Affecting Frequency Response • Mitigation Measures • Conclusions 34 Conclusions (from California work plus updates) • Secondary reserves need to be adequate • Renewable penetration outside of California (in neighboring systems) is important • Kt is a good primary metric • Kt alone does not give all the necessary information… headroom is important • Speed of primary response is important • Governor Withdrawal has a detrimental impact on frequency response • Impact of reduced System Inertia on initial rate-of-change-of-frequency does not appear to be important • Inertial Fast-Frequency Response controls from Wind Generation help • Participation of renewables (wind and solar) in providing frequency response is beneficial • Load control can be used to improve frequency response • Fast acting Energy Storage will provide significant benefits • Market mechanisms will likely be necessary to assure adequate frequency response in future and under all operating conditions • Control philosophy is more important than the source of energy (e.g wind, solar, storage, load) in determining frequency response © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 35 [...]... of Units # of Units 1 894 2 284 5045 92 13 897 192 27057 81 69 18888 397 4 97 65 8115 3124 499 1 44815 510 12780 168 32035 342 96 78 320 2617 99 7060 221 18 094 8411 96 84 2550 2550 0 102 194 73435 94 392 91 300 34527 1114 20442 296 83 26 190 19. 2% 2.7% 26.5% 25.8% 20.1% 11.1% 8.6% 28.3% 8.6% 23.7% 25.5% 24.7% 17.0% 15.3% 10.5% 3 59 25.6% 67667 5 299 2 64710 65111 15.0% 0.0% 27 .9% 26.2% 21.5% 9. 4% 7.7% Condition in... # of Units 2 391 3 284 7018 92 27057 81 69 13640 397 4 97 65 3144 1151 396 76 510 114 39 168 96 78 320 2617 99 18 094 8411 2550 2550 MW Capability CU Pgen (MW) (GR + BL + NG) Total Pgen (MW) Total Pload (MW) 97 055 73267 94 225 91 301 33186 1114 21074 30315 26 190 Wind Pgen/Total Pgen Solar Pgen/Total Pgen Kt GR Pgen/CU Pgen GR Pgen/Total Pgen GR Headroom/CU Pgen GR Headroom/Total Pgen 19. 2% 2.7% 27 .9% 32.6% 25.4%... Metrics • Frequency Response of Base Cases • Frequency Response of High Renewable Penetration Cases • Factors Affecting Frequency Response • Mitigation Measures • Conclusions 22 Factors Degrading Frequency Response – Fewer Governors in Operation Keep all other factors impacting frequency response fixed: Governor Response (GR) units • 25 GR units, with total dispatch of 3144 MW and rating (MWCAP) of 51 89 MW... reserved Not for distribution without permission 11 Frequency Performance Metrics 20 to 52 sec • Frequency Nadir (Cf) • Frequency Nadir Time (Ct) • LBNL Nadir-Based Frequency Response (MW Loss/Δfc*0.1) • GE-CAISO NadirBased Frequency Response (Δ MW/Δfc *0.1) • Settling Frequency (Bf) • NERC Frequency Response (MW Loss/Δfb*0.1)* • GE-CAISO SettlingBased Frequency Response (Δ MW/Δfb*0.1) • NERC BAL-003-1 finalized... International, Inc CA # of Units 5514 127 97 85 4271 94 78 155 1757 121 8646 6667 127146 85577 114775 110 798 36333 1562 167 49 30525 35155 19. 0% 5 .9% 40.6% 45.1% 33.6% 15.2% 11.3% 28.3% 21.8% 26 .9% 43.4% 32 .9% 18.1% 25.5% 14.0% 403 4% 31.5% Non-CA # of Units 33075 551 41802 8727 2 790 6 276 7845 332 12428 144 90 125 68826 84250 75643 14.8% 0.2% 46.4% 48.1% 39. 3% 12.7% 10.4% 11 59 47.5% Kt is smaller for nonCA here... 8.4% 24.6% 25.5% 33.3% 23.2% 5.5% 3.8% 3 59 25.6% Non-CA # of Units 16 895 192 18888 199 3 28238 342 7060 221 96 84 0 63870 52 193 6 391 0 65111 15.2% 0.0% 29. 6% 32.4% 26.4% 3.8% 3.1% © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 755 25.4% 27 Impact of Extreme Minimum Headroom and Governor Participation (Kt) on Frequency Performance Winter Low Load –... primary response is important • Governor Withdrawal has a detrimental impact on frequency response • Impact of reduced System Inertia on initial rate-of-change-of -frequency does not appear to be important • Inertial Fast -Frequency Response controls from Wind Generation help • Participation of renewables (wind and solar) in providing frequency response is beneficial • Load control can be used to improve frequency. .. Performance Metrics • Frequency Response of Base Cases • Frequency Response of High Renewable Penetration Cases • Factors Affecting Frequency Response • Mitigation Measures Reduced Governor Withdrawal Inertial Response From Wind Plants Governor Response (Frequency Droop) from Wind Plants Load Control/Fast Energy Storage • Conclusions 29 Mitigation Measures – Reduced Governor Withdrawal Disable... MW of governor response is deliberately withdrawn, representing almost 10 percent of total frequency response Load Control should be frequency dependent © 2016 General Electric International, Inc All rights reserved Not for distribution without permission 18 Outline • Study Objectives • Development of Study Database and Performance Metrics • Frequency Response of Base Cases • Frequency Response of Higher... Frequency Response of Base Cases • Frequency Response of Higher Renewable Penetration Cases • Factors Affecting Frequency Response • Mitigation Measures • Conclusions 19 Generation Summary for Weekend Morning – High WECC Wind and Solar Case WECC # of Units 38 590 678 51587 1 299 7 37384 431 96 03 453 21762 6810 GR Pgen (MW) GR MWCAP (MW) GR Headroom (MW) BL Pgen (MW) NG Pgen (MW) Wind Pgen (MW) Solar Pgen ... of Units 1 894 2 284 5045 92 13 897 192 27057 81 69 18888 397 4 97 65 8115 3124 499 1 44815 510 12780 168 32035 342 96 78 320 2617 99 7060 221 18 094 8411 96 84 2550 2550 102 194 73435 94 392 91 300 34527... 2 391 3 284 7018 92 27057 81 69 13640 397 4 97 65 3144 1151 396 76 510 114 39 168 96 78 320 2617 99 18 094 8411 2550 2550 MW Capability CU Pgen (MW) (GR + BL + NG) Total Pgen (MW) Total Pload (MW) 97 055... 27.7% 8.4% 24.6% 25.5% 33.3% 23.2% 5.5% 3.8% 3 59 25.6% Non-CA # of Units 16 895 192 18888 199 3 28238 342 7060 221 96 84 63870 52 193 6 391 0 65111 15.2% 0.0% 29. 6% 32.4% 26.4% 3.8% 3.1% © 2016 General