RAN Reliability Requirements RASC June 10, 2020 RASC010, RASC011, - PowerPoint PPT Presentation

RAN Reliability Requirements RASC June 10, 2020 RASC010, RASC011, RASC012

Purpose & Purpose: Present draft framing, discuss preliminary Key analysis results, review industry trends and Takeaways benchmark Key Takeaways: • How we define system reliability needs must change, recognizing an evolving risk profile • Analysis indicates shift of risk patterns outside typical summer peak load periods and growing flexibility needs with the changing resource mix • Resource adequacy approaches vary across regions including pros and cons of different metrics

MISO’s product development process moves from problem definition to buildable solutions Explore Decide Do We are here Problem Exploration Proposal & Software, Deploy to Review Definition & of Options Business Rules BPM updates production outcomes Education & Training 3 Selects an option and details specific design. Defines important design elements and derives options via higher fidelity modeling. Defines needs to address through additional steps. Can iterate as move through the process. 3

Problem Definition • Today’s analysis does not sufficiently capture risk across the year • The analysis should better reflect patterns across the year and not just summer • The analysis should better reflect the magnitude of risks • Today’s approach to setting requirements does not sufficiently mitigate risk • The approach to setting targets should leverage enhanced risk calculation • The approach to setting targets should sufficiently mitigate risk under current and future portfolios • System risk profiles will continue to change with evolution of the resource mix Risks could emerge in time periods other than summer peak • The margin between available resources and total obligations will be more • impacted by uncertainty and variability 4

Operating margins impact reliability • The margin between supply resources and obligations is an indicator of how close the system is to emergency or loss of load.* • It is influenced by a number of factors, some of which are highly variable and uncertain • Outages • Intermittent generation • Net scheduled interchange Margin = Available non-intermittent generation + intermittent generation + RDT limit + Net Scheduled Interchange + Load Resources (BTMG + LMR + EDR) - Load - Operating Reserve * Emergencies include alerts through to load shed RDT = Regional Dispatch Transfer Limit | BTMG = Behind the Meter Generation 5 LMR = Load Modifying Resources | EDR = Emergency Demand Response

A historical look indicates that periods of tight operating margin aligns with low efficiency / high risk Systemwide Historical Margins 2018 Time Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 Month  1  Hour 2 3 4 Occurrence Margin Risk 5 Risks exist outside 6 None Mild of summer months 7 8 9 Moderate Moderate 10 11 Severe High 12 13 14 15 16 17 18 19 20 21 22 23 Occurrences are defined as Max Gen, Alert, or RSG greater than three times average 6 Severity is color-coded by margin size

Adjustment to modeling assumptions can better capture risk across the year Adjustments to summer-Focused LOLE Modeling Assumptions Assumption Current Approach Trial Analysis Assumptions • Intermittent Flat capacity throughout the year Using 8760 profiles corresponding to resource based on summer performance. 2018 weather year. capacity • Non-firm Adjustment to the PRM based on Using monthly average NSI from the external support imports during summer peak . last 3 years, assuming a perfect unit. Modeled with adders / subtractors at Modelled as a single average • Forced outage different date-hour based on forced outage rate for the entire rates (FOR) temperature correlation model using 3- year. years of historical data. Optimized to avoid outages during Scheduled using a 90% optimality • Planned outages peak summer load periods . (“best behavior”) assumption. Margin = Available non-intermittent generation + intermittent generation + RDT limit + Load Resources (BTMG + LMR + EDR) + Net Scheduled Interchange - Load - Cleared Operating Reserve 7

Trial inputs were modified to better reflect system risk across the year Forced Outages Planned Maintenance 16 16 GW GW 14 14 12 12 10 10 8 8 6 6 4 4 2 2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Month Month 16 16 GW GW Non-Firm External Support Intermittent Production 14 14 12 12 10 10 8 8 6 6 4 4 2 2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Month Month 8 Draft analysis of 2018 Current Trial

Modifying Loss of Load Expectation inputs to reflect seasonality can better reflect risk throughout the year Loss of Load Expectation* Modeled 2018 Current Method Trial Method Time Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Time Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 0 Month   Hour 1 1 2 2 3 3 Risks shift outside of 4 4 summer months 2018 EUE 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 9 *0.1 LOLE Target No risk Moderate Risk High Risk

The team also ran trials on future sample future scenarios • The futures reflect many dynamics in the region Changing fuel costs, clean energy commitments, aging fleet, electrification • • The analysis leveraged draft MTEPs as a starting point MTEP 2019 AFC 2033 Future III, 2030** Future 1 2040 ** (760TWh) (954 TWh) (825 TWh) • Less coal retirement • Gas in lieu of coal • Higher solar • Reflects member plans* • High load • Middle load growth * As announced plans submitted to commissions 10 ** Does not include very recent change in MTEP F3 that adjusts load growth from 60% to 50%.

Growth in intermittent resources could increase variability and overall impact of uncertainty Scenario Input Size Average Variability + (Range) (GW) (GW) 6 (0 to 16) 3 Wind Current Solar NA NA 13 (0.3 to 31 ) 7 Wind FutureI-2040 14 (0 to 52 ) 16 Solar 37 (0.8 to 85) 20 Wind FutureIII-2030 2 (0 to 6) 2 Solar 18 (0.4 to 43) 10 Wind MTEP19AFC + Standard deviation ++ Reflects changes between Day Ahead and Real Time, including forecast error 11 2033 6 (0 to 23) 7 Solar 23 (0.3 85) 17 Wind Total 7 (0 to 52) 11 Solar 11

Future resource evolution and total capacity shape pattern of need & non-summer peak Future 1 2040 only focus is still relevant Draft results 0.1 Loss of Load Expectation Proxy Operator Experience* Time Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Time Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 0 Month  1 1 Risks shift outside of  Hour 2 2 3 3 summer months & 4 4 5 traditional peak hours 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 * 0.6 LOLE Target 12 No risk Moderate Risk High Risk

Alternative measures may help differentiate system impacts Portfolios with similar LOLE result in a wide range of EUE values Draft results MTEP 2019 AFC 2033 Interval LOLE (day) LOLP (%) EUE (MWh) EDNS (MW) Annual 0.10 0.03 1,765 700 Future III, 2030 ** Interval LOLE (day) LOLP (%) EUE (MWh) EDNS (MW) Annual 0.10 0.03 3,021 1,550 Future 1 2040 Interval LOLE (day) LOLP (%) EUE (MWh) EDNS (MW) Annual 0.10 0.03 3,550 2,400 LOLE = Loss of Load Expectation EUE = Expected Unserved Energy 13 EDNS = Expected Demand Not Supplied

Stakeholder Feedback Request MISO is requesting feedback by June 24, 2020 on • The proposed problem definition • Priority adjustments to inputs for higher fidelity analysis in • the evaluation of options Issue Tracking ID#: RASC010, RASC011, RASC012 • Feedback requests and responses are managed • through the Feedback Tool on the MISO website: https://www.misoenergy.org/stakeholder- engagement/stakeholder-feedback/ 14 |

Alternative RA Metrics, and RA Construct Elements June 10, 2020 PRESENTED TO PREPARED BY MISO Resource Adequacy Sam Newell Michael Hagerty Subcommittee Hannes Pfeifenberger Walter Graf

Agenda • RA Metric Benchmark • RA Construct Elements Benchmark Brattle.com | 2